20.02.2013 Views

Diploma thesis in Physics submitted by Florian Freundt born in ...

Diploma thesis in Physics submitted by Florian Freundt born in ...

Diploma thesis in Physics submitted by Florian Freundt born in ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Faculty of <strong>Physics</strong> and Astronomy<br />

Heidelberg University<br />

<strong>Diploma</strong> <strong>thesis</strong><br />

<strong>in</strong> <strong>Physics</strong><br />

<strong>submitted</strong> <strong>by</strong><br />

<strong>Florian</strong> <strong>Freundt</strong><br />

<strong>born</strong> <strong>in</strong> Lemgo<br />

2011


Measur<strong>in</strong>g annual variation of soil<br />

atmosphere composition focus<strong>in</strong>g on<br />

the effect of oxygen depletion on<br />

noble gas partial pressures<br />

This diploma <strong>thesis</strong> was carried out <strong>by</strong> <strong>Florian</strong> <strong>Freundt</strong> at the<br />

Institute of Environmental <strong>Physics</strong><br />

under the supervision of<br />

Prof. Dr. Werner Aeschbach-Hertig


Measur<strong>in</strong>g annual variation of soil atmosphere composition<br />

focus<strong>in</strong>g on the effect of oxygen depletion on noble gas partial<br />

pressures<br />

Abstract<br />

It is known that the partial pressures of soil atmosphere components like oxygen, carbon dioxide and<br />

nitrous oxides show fluctuations on both temporal and spatial scales. These are primarily caused <strong>by</strong><br />

microbiologic activities <strong>in</strong> the soil. However, the use of noble gases dissolved <strong>in</strong> ground water as a climate<br />

proxy utilizes the basic assumption that the ground air equilibrat<strong>in</strong>g with water dur<strong>in</strong>g recharge is of<br />

atmospheric composition with regard to noble gases. This assumption has been questioned to account<br />

for lower than expected noble gas temperatures, suggest<strong>in</strong>g a rise of noble gas partial pressures <strong>in</strong> the<br />

soil atmosphere caused <strong>by</strong> removal of CO2 (produced <strong>by</strong> O2 deplet<strong>in</strong>g soil respiration) due to its high<br />

solubility <strong>in</strong> water [Hall et al., 2005]. To test this proposition three permanent sampl<strong>in</strong>g sites were built<br />

<strong>in</strong> clay dom<strong>in</strong>ated soil, allow<strong>in</strong>g for sampl<strong>in</strong>g of ground air <strong>in</strong> regular <strong>in</strong>tervals and depths up to 6 meters.<br />

O2 and CO2 concentrations were measured on site while the noble gases helium, neon, argon, krypton<br />

and xenon were sampled and measured <strong>in</strong> the laboratory us<strong>in</strong>g mass spectrometry.<br />

It was confirmed that O2 and CO2 concentrations with<strong>in</strong> the soil atmosphere fluctuate strongly, the sum<br />

of O2 and CO2 reached a m<strong>in</strong>imum of 16.5 Vol%. Soil atmosphere noble gas composition deviated from<br />

atmospheric composition, i.e. their concentrations <strong>in</strong>creased when O2+CO2 concentrations decreased.<br />

The highest observed <strong>in</strong>crease <strong>in</strong> noble gas concentrations was 106 % of atmospheric air concentrations.<br />

Based on an actual soil temperature of 12 ◦ C, this would cause an underestimation of temperature <strong>by</strong><br />

the currently employed CE model <strong>by</strong> 1.5 ◦ C.<br />

Zusammenfassung<br />

Die Partialdrücke von Gasen wie Sauerstoff, Kohlendioxid und Stickoxiden <strong>in</strong> der Bodenluft schwanken<br />

sowohl zeitlich als auch räumlich, angetrieben durch mikrobiologische Aktivität im Boden. Dennoch wird<br />

bei der Rekonstruktion von Paläotemperaturen aus <strong>in</strong> Grundwasser gelösten Edelgasen angenommen,<br />

dass die Edelgasanteile der Bodenluft identisch mit denen atmosphärischer Luft s<strong>in</strong>d. Diese Annahme<br />

wurde von Hall et al. [2005] <strong>in</strong> Frage gestellt, um Edelgastemperaturen unterhalb der erwarteten Werte<br />

zu erklären. Hall et al. [2005] schlagen vor, dass der Anstieg der Edelgaspartialdrücke durch e<strong>in</strong> Defizit<br />

der Summe von O2 und CO2 <strong>in</strong> der Bodenluft verursacht wird. Dieses Defizit entsteht, wenn mikrobiologische<br />

Prozesse O2 <strong>in</strong> CO2 umwandeln und das CO2 durch se<strong>in</strong>e hohe Löslichkeit bei Niederschlag aus<br />

der Bodenluft entfernt wird. Um dies zu überprüfen wurden drei permanente Messstellen <strong>in</strong> Lehmböden<br />

e<strong>in</strong>gerichtet, die e<strong>in</strong>e regelmäßige Beprobung der Bodenluft bis <strong>in</strong> 6 m Tiefe erlaubten. O2 und CO2<br />

wurden vor Ort gemessen, Proben mit Bodenluft wurden im Labor am Massenspektrometer auf ihre<br />

Edelgaszusammensetzung h<strong>in</strong> untersucht.<br />

Die Schwankung der O2- und CO2-Konzentrationen <strong>in</strong> der Bodenluft konnte bestätigt werden, der niedrigste<br />

gemessene Wert der Summe von O2 und CO2 war 16.5 Vol%. Die Edelgaszusammensetzung zeigte<br />

Abweichungen von der atmosphärischen Zusammensetzung: Steigende Edelgaskonzentrationen korrelierten<br />

mit der Abnahme der Summe von O2 und CO2. Die höchste gemessene Edelgaskonzentration<br />

betrug 106 % der atmosphärischen Konzentration. Bei e<strong>in</strong>er Bodentemperatur von 12 ◦ C würde e<strong>in</strong> derartiger<br />

Anstieg bei dem derzeit verwendeten CE-Modell zu e<strong>in</strong>er Unterschätzung der Temperatur um<br />

1.5 ◦ C führen.


Contents<br />

Abstract 4<br />

Contents 6<br />

1 Introduction 11<br />

2 Theory 15<br />

2.1 Noble gas temperatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15<br />

2.1.1 Solubility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16<br />

2.1.2 Noble gas fractions <strong>in</strong> ground water . . . . . . . . . . . . . . . . . . . . . 17<br />

2.1.3 Excess air model<strong>in</strong>g approaches . . . . . . . . . . . . . . . . . . . . . . . . 19<br />

2.2 Physical processes and properties of soils . . . . . . . . . . . . . . . . . . . . . . . 24<br />

2.2.1 Subsurface thermal regime . . . . . . . . . . . . . . . . . . . . . . . . . . . 24<br />

2.2.2 Soil structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26<br />

2.2.3 Gas transport processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26<br />

2.3 Soil atmosphere composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28<br />

2.3.1 Sources, s<strong>in</strong>ks and profiles of O2 and CO2 . . . . . . . . . . . . . . . . . . 28<br />

2.3.2 Variability of soil respiration . . . . . . . . . . . . . . . . . . . . . . . . . 31<br />

2.3.3 Molecular nitrogen and nitrogenous gases . . . . . . . . . . . . . . . . . . 32<br />

2.3.4 Radon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33<br />

3 Sampl<strong>in</strong>g sites and methods 35<br />

7


CONTENTS CONTENTS<br />

3.1 Setup of the sampl<strong>in</strong>g sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35<br />

3.1.1 Drill<strong>in</strong>g and <strong>in</strong>strumentation . . . . . . . . . . . . . . . . . . . . . . . . . 35<br />

3.1.2 Locations and soil properties . . . . . . . . . . . . . . . . . . . . . . . . . 36<br />

3.1.3 Development of the sites dur<strong>in</strong>g the sampl<strong>in</strong>g period . . . . . . . . . . . . 39<br />

3.2 Sampl<strong>in</strong>g methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40<br />

3.2.1 Noble gas samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40<br />

3.2.2 O2, CO2 and CH4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42<br />

3.2.3 Radon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42<br />

3.2.4 Temperatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42<br />

4 Measur<strong>in</strong>g methods 43<br />

4.1 Mass spectrometry of He, Ne, Ar, Kr and Xe . . . . . . . . . . . . . . . . . . . . 43<br />

4.1.1 Sample preparation and measur<strong>in</strong>g procedure . . . . . . . . . . . . . . . . 43<br />

4.1.2 Determ<strong>in</strong>ation of sample gas amount . . . . . . . . . . . . . . . . . . . . . 44<br />

4.1.3 Estimation of relative humidity with<strong>in</strong> the <strong>in</strong>let section . . . . . . . . . . 45<br />

4.1.4 Result<strong>in</strong>g data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46<br />

4.2 On site measurement of O2, CO2, CH4 and Radon . . . . . . . . . . . . . . . . . 46<br />

4.2.1 Geotech BM2000 Biogas Monitor . . . . . . . . . . . . . . . . . . . . . . . 46<br />

4.2.2 Durridge RAD7 Radon Detector . . . . . . . . . . . . . . . . . . . . . . . 47<br />

5 Results 49<br />

5.1 Radon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49<br />

5.2 Temperature profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49<br />

5.3 O2 and CO2 profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51<br />

5.4 Borehole seal<strong>in</strong>g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54<br />

5.4.1 Site A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54<br />

5.4.2 Site B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57<br />

5.5 Noble gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57<br />

8


CONTENTS CONTENTS<br />

5.5.1 Effects of storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57<br />

5.5.2 Atmospheric air samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58<br />

5.5.3 Soil atmosphere samples from Site A . . . . . . . . . . . . . . . . . . . . 60<br />

5.5.4 Soil atmosphere samples from Site B . . . . . . . . . . . . . . . . . . . . 62<br />

6 Discussion 63<br />

7 Summary 71<br />

8 Outlook 73<br />

A Calculations 75<br />

A.1 Gas sample size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75<br />

A.1.1 Vapor pressure of water . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75<br />

A.1.2 Calculation of sample size us<strong>in</strong>g the <strong>in</strong>let pressure . . . . . . . . . . . . . 75<br />

A.1.3 Calculation of sample size us<strong>in</strong>g the sample tube length . . . . . . . . . . 76<br />

A.2 Mass spectrometer <strong>in</strong>let volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77<br />

A.3 Fractionation-Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78<br />

B Data 79<br />

C Additional plots and figures 89<br />

D Datasheets 107<br />

Bibliography 111<br />

Acknowledegment 117<br />

Deposition 119<br />

9


Chapter 1<br />

Introduction<br />

Understand<strong>in</strong>g the Earth’s climate and the processes affect<strong>in</strong>g it has become very important <strong>in</strong><br />

the past decades to comprehend how anthropogenic factors <strong>in</strong>fluence the climate. Predictions of<br />

climate change rely on model<strong>in</strong>g approaches, which <strong>in</strong> turn require a large data basis of past and<br />

present climate records. Data on climate parameters from direct human records extend only a<br />

few hundred years back <strong>in</strong>to the past at best. It is possible to reconstruct older data <strong>by</strong> us<strong>in</strong>g so<br />

called climate proxies which are essentially physical characteristics of past climate parameters,<br />

stored <strong>in</strong> natural archives. Various k<strong>in</strong>ds of archives exist, conta<strong>in</strong><strong>in</strong>g paleoclimate tracers of<br />

different type and quality. A few examples are tree r<strong>in</strong>gs, ice cores and ocean sediments. The<br />

general concept is that a tracer dependent on a climate parameter, like temperature, precipitation<br />

or CO2 concentrations <strong>in</strong> atmospheric air to name a few, is embedded <strong>in</strong>to the archive at a<br />

certa<strong>in</strong> time and is ideally conserved, unaffected <strong>by</strong> any processes afterwards. As such archives<br />

display a chronological structure due to their formation processes, dat<strong>in</strong>g of the embedded <strong>in</strong>formation<br />

results <strong>in</strong> a paleoclimate record extend<strong>in</strong>g way beyond human record<strong>in</strong>gs. The type of<br />

climate parameter, its accuracy and the chronological range and resolution of the record depend<br />

on the type of tracer and archive.<br />

One of these archives is ground water. The analyzed tracers used to reconstruct paleotemperatures<br />

from ground water are dissolved noble gases. Ground water makes up 30.1 % of the<br />

Earth’s fresh water [Hölt<strong>in</strong>g and Coldewey, 2009], found with<strong>in</strong> the Earth’s crust below the vadose<br />

(unsaturated) soil zone. A water permeable ground layer conta<strong>in</strong><strong>in</strong>g ground water is called<br />

an aquifer, while a layer restrict<strong>in</strong>g water flow is called aquitard. In an unconf<strong>in</strong>ed aquifer, where<br />

the ground layer above the water saturated zone is water-permeable, the water reaches the water<br />

table where the water pressure equals the ambient atmospheric pressure. The shallowest aquifer<br />

<strong>in</strong> a given soil structure is usually unconf<strong>in</strong>ed. A conf<strong>in</strong>ed aquifer is characterized <strong>by</strong> an aquitard<br />

layer above the aquifer, restrict<strong>in</strong>g upwards movement of the ground water and there<strong>by</strong> caus<strong>in</strong>g<br />

the water pressure to be higher than the atmospheric pressure at the upper aquifer boundary.<br />

A conf<strong>in</strong>ed aquifer isolates the stored ground water from <strong>in</strong>teraction with the soil atmosphere of<br />

the overly<strong>in</strong>g vadose zone.<br />

Ground water flow with<strong>in</strong> aquifers is dom<strong>in</strong>ated <strong>by</strong> gravitative forc<strong>in</strong>g, volume flow rates are described<br />

<strong>by</strong> Darcy’s law (see Hölt<strong>in</strong>g and Coldewey [2009] for a detailed description) and depend<br />

11


1 Introduction<br />

on the pressure gradient and a parameter describ<strong>in</strong>g the hydraulic conductivity of the given<br />

soil. Flow rates and residence time of water with<strong>in</strong> aquifers therefore vary depend<strong>in</strong>g on the<br />

soil’s hydraulic properties an recharge rates. Water <strong>in</strong> conf<strong>in</strong>ed aquifers can reach an age of up<br />

to several 10 5 years [Sturchio et al., 2004]. Radiocarbon dat<strong>in</strong>g of dissolved <strong>in</strong>organic carbon<br />

<strong>in</strong> ground water is used to provide the age <strong>in</strong>formation for up to 50,000 year old water while<br />

different isotopic dat<strong>in</strong>g methods have to be used for older ground waters.<br />

Reconstruct<strong>in</strong>g paleotemperatures from ground water utilizes the dissolved noble gases neon,<br />

argon, krypton and xenon and requires conf<strong>in</strong>ed aquifers that limit gas exchange. The idea beh<strong>in</strong>d<br />

this paleorecord is the temperature dependency of gas solubility: the <strong>in</strong>filtrat<strong>in</strong>g meteoric<br />

water lead<strong>in</strong>g to ground water recharge is <strong>in</strong> contact with the soil atmosphere before enter<strong>in</strong>g<br />

the aquifer. Dur<strong>in</strong>g this contact the gaseous and aqueous phases equilibrate. This equilibration<br />

is dependent on soil temperature and the composition of the soil atmosphere. Due to temperature<br />

damp<strong>in</strong>g with<strong>in</strong> the soil, its temperature is closely related to annual mean atmospheric<br />

temperatures [Hillel, 1980]. The standard assumption on soil atmosphere noble gas composition<br />

is that it is close or equal to the atmospheric composition, show<strong>in</strong>g only <strong>in</strong>significant<br />

fluctuations [Stute and Schlosser, 1993]. This however neglects the fact that soil atmospheres<br />

greatly vary <strong>in</strong> O2 and CO2 composition spatially as well as temporally [Yamaguchi et al., 1967;<br />

Dowdell and Smith, 1974; Amundson and Davidson, 1990; Magnusson, 1992]. This likely affects<br />

the partial pressures of its rema<strong>in</strong><strong>in</strong>g components. These changes <strong>in</strong> soil air composition are<br />

ma<strong>in</strong>ly caused <strong>by</strong> microbiological activities and are <strong>in</strong>fluenced <strong>by</strong> a multitude of soil properties<br />

like temperature, precipitation, soil hydraulic properties and others <strong>in</strong> a complex relationship<br />

[Suarez and ˇ Sim˚unek, 1993; Welsch and Hornberger, 2004; Riveros-Iregui et al., 2011]. While<br />

paleotemperature studies have largely neglected this possibility so far and have nonetheless successfully<br />

employed model<strong>in</strong>g approaches to account for other factors affect<strong>in</strong>g the dissolved noble<br />

gases [Aeschbach-Hertig et al., 1999b; Kipfer et al., 2002; Peeters et al., 2003], studies on young<br />

ground water <strong>in</strong> recharge areas [Stute and Sonntag, 1992; Ma et al., 2004; Castro et al., 2007]<br />

have led to noble gas temperatures a few degrees below measured soil temperatures. Hall et al.<br />

[2005] proposed that this shift could be caused <strong>by</strong> a deviation of noble gas partial pressures <strong>in</strong><br />

the soil atmosphere from atmospheric values. As the cause for this change they suggest the<br />

process of oxygen depletion, mean<strong>in</strong>g that the O2 removed from the soil atmosphere <strong>by</strong> microorganisms<br />

is not replaced <strong>by</strong> an equimolar amount of CO2 produced <strong>by</strong> these organisms, lead<strong>in</strong>g<br />

to a pressure deficit affect<strong>in</strong>g the partial pressures of the rema<strong>in</strong><strong>in</strong>g gases. While their model<strong>in</strong>g<br />

approach <strong>in</strong>clud<strong>in</strong>g the proposed oxygen depletion effect successfully leads to match<strong>in</strong>g noble<br />

gas and mean atmospheric temperatures for their study area, an actual measurement of soil air<br />

composition was not executed to prove oxygen depletion has the suggested effect on noble gas<br />

partial pressures and could <strong>in</strong>deed be the ma<strong>in</strong> factor <strong>in</strong> caus<strong>in</strong>g the noble gas temperature shift.<br />

Extensive research has been done on the composition of soil atmospheres for various reasons,<br />

ma<strong>in</strong>ly ow<strong>in</strong>g to its importance <strong>in</strong> agricultural contexts. Modern climate research has also been<br />

<strong>in</strong>terested <strong>in</strong> soil atmospheres because of their part <strong>in</strong> the global atmospheric gas balance. Soils<br />

provide both s<strong>in</strong>ks and sources for CO2, CH4, N2O and various other gases relevant to the<br />

Earth’s radiation balance. These studies ma<strong>in</strong>ly focused on O2, CO2 and nitrogenous gases and<br />

usually their flux from the soil rather than <strong>in</strong> situ concentrations. Data on noble gases <strong>in</strong> the<br />

soil atmospheres relevant to noble gas temperatures is sparse at best, as little research on soil<br />

atmosphere composition focus<strong>in</strong>g on the noble gas components has been done so far. While<br />

12


1 Introduction<br />

Mol<strong>in</strong>s and Mayer [2007] have noted a relative enrichment of N2 and Ar partial pressures <strong>in</strong><br />

the presence of O2 depletion at an exam<strong>in</strong>ation of the soil atmosphere at a crude oil spill site,<br />

their study neither did focus on noble gases nor did it conta<strong>in</strong> any soils likely to be relevant<br />

to paleoclimate research. Previous one-time sampl<strong>in</strong>gs of soil atmospheres focused on noble<br />

gas measurements <strong>in</strong> the region around Heidelberg (various permanent and s<strong>in</strong>gular sampl<strong>in</strong>g<br />

sites), executed <strong>by</strong> Schneider [2010], could <strong>in</strong> fact not provide any data capable of support<strong>in</strong>g<br />

the oxygen depletion model.<br />

The objective of this study is to create an annual record of soil atmosphere composition at<br />

various depths and locations, focus<strong>in</strong>g on the stable noble gases He, Ne, Ar, Kr and Xe as well<br />

as O2 and CO2 to confirm or refute the existence of changes <strong>in</strong> noble gas composition <strong>in</strong> soil<br />

atmospheres and to quantify the possible annual variation and effect on noble gas temperatures<br />

should such changes occur. Three permanent soil atmosphere sampl<strong>in</strong>g sites were created,<br />

allow<strong>in</strong>g for depth profiles to be taken. Noble gas samples were taken over a span of ten months<br />

and the concentrations were measured us<strong>in</strong>g mass spectroscopy. O2 and CO2 concentrations<br />

were measured over a span of five months.<br />

13


Chapter 2<br />

Theory<br />

2.1 Noble gas temperatures<br />

The concept of us<strong>in</strong>g dissolved noble gases <strong>in</strong> ground water as tracers for paleotemperature is<br />

based on the observed temperature dependency of gas solubility <strong>in</strong> water, the ma<strong>in</strong>ly atmospherical<br />

source of noble gases <strong>in</strong> ground water and the chemically <strong>in</strong>ert behavior of noble gases.<br />

With<strong>in</strong> the relevant temperature <strong>in</strong>terval found at ground water recharge areas the solubility<br />

decreases with <strong>in</strong>creas<strong>in</strong>g temperature, this effect is more dist<strong>in</strong>ctive for heavier noble gases.<br />

The ma<strong>in</strong> source of all stable noble gases found <strong>in</strong> meteoric and ground water is the Earth’s<br />

atmosphere. Radiogenic and terrigenic noble gas sources are of little <strong>in</strong>fluence, only 3 He, 4 He<br />

and sometimes 40 Ar are <strong>in</strong>fluenced <strong>by</strong> non-atmospheric sources [Kipfer et al., 2002]. Due to this<br />

<strong>in</strong>fluence and the small temperature dependance of its solubility, helium is not used for model<strong>in</strong>g<br />

paleotemperatures.<br />

F<strong>in</strong>ally, the chemical <strong>in</strong>ertness of the stable noble gases ensures the absence of chemical s<strong>in</strong>ks,<br />

mak<strong>in</strong>g dissolved noble gases a conservative tracer and lead<strong>in</strong>g to a preservation of the temperature<br />

<strong>in</strong>formation. This requires a conf<strong>in</strong>ed aquifer however, s<strong>in</strong>ce depressurization would cause<br />

degass<strong>in</strong>g and there<strong>by</strong> loss of the temperature <strong>in</strong>formation.<br />

When water from a conf<strong>in</strong>ed aquifer is sampled and analyzed for its noble gas concentrations,<br />

the result<strong>in</strong>g noble gas temperatures therefore represent mean annual soil temperatures at the<br />

ground water table of the recharge area. The temporal resolution of the noble gas temperatures<br />

depends on mix<strong>in</strong>g, dispersion and transport processes with<strong>in</strong> the aquifer and the accuracy of<br />

the water sample dat<strong>in</strong>g.<br />

The follow<strong>in</strong>g description of gas solubility and noble gas temperature calculation is largely based<br />

on Kipfer et al. [2002], Aeschbach-Hertig et al. [2008] and Schneider [2010].<br />

15


2.1. Noble gas temperatures 2 Theory<br />

� � � � � � � � � � � �� � � �<br />

Figure 2.1: Temperature dependency of the Ostwald solubility L(T, S = 0) for the noble<br />

gases He, Ne, Ar, Kr and Xe, at atmospheric pressure. Plot data was calculated us<strong>in</strong>g the<br />

fit equation and parameters provided <strong>by</strong> Benson and Krause [1976].<br />

2.1.1 Solubility<br />

The partition<strong>in</strong>g of gas at the phase boundary between water and air at a constant temperature<br />

is described <strong>by</strong> Henry’s Law, formulated <strong>in</strong> 1803 <strong>by</strong> William Henry:<br />

C g<br />

i = Hi(T, S) · C w i (2.1)<br />

where C g<br />

i and Cw i are the concentrations of gas i <strong>in</strong> the gas and the water phase respectively.<br />

Hi is the gas specific Henry constant, a dimensionless constant that is a function of temperature<br />

T and sal<strong>in</strong>ity S. Influences on Hi <strong>by</strong> chemical <strong>in</strong>teractions of solutes found <strong>in</strong> water can be<br />

neglected for meteoric and ground waters [Kipfer et al., 2002]. S<strong>in</strong>ce the Henry constant is<br />

dimensionless, its numerical value is def<strong>in</strong>ed <strong>by</strong> the units chosen for the gas concentrations C.<br />

The reciprocal of the Henry constant, called Ostwald solubility L, is a measure of solubility.<br />

Li(T, S) =<br />

1<br />

Hi(T, S) = Cw i<br />

C g<br />

i<br />

The temperature dependency of the solubility is described <strong>by</strong> a numerical approximation <strong>by</strong><br />

fitt<strong>in</strong>g measured data to an equation like<br />

ln � Li(T, S = 0) � 1 1<br />

= a0 + a1 + a2<br />

T<br />

16<br />

T 2<br />

(2.2)<br />

(2.3)


2 Theory 2.1. Noble gas temperatures<br />

where T is the temperature <strong>in</strong> Kelv<strong>in</strong>, lead<strong>in</strong>g to parameters ai as given Table B.1 [Benson and Krause,<br />

1976]. The result<strong>in</strong>g functions for <strong>in</strong>dividual noble gases are shown <strong>in</strong> Figure 2.1 where the Ostwald<br />

solubility Li(T, S = 0) is plotted versus temperature.<br />

In ground water research it is usually more convenient to work with partial pressures pi of gas<br />

i. As described <strong>by</strong> Kipfer et al. [2002], us<strong>in</strong>g the equilibrium concentration Ci,eq, Henry’s Law<br />

is formulated as<br />

pi = Hi(T, S) · Ci,eq<br />

(2.4)<br />

The partial pressures of noble gases <strong>in</strong> atmospheric air, based on the assumption that the<br />

atmospheric noble gas composition is constant and usable as a standard [Porcelli et al., 2002],<br />

is calculated from ambient atmospheric pressure ptot:<br />

pi = zi · � ptot − p 0 �<br />

H2O<br />

with the volume fraction zi of the gas i <strong>in</strong> air (see Table B.3) and the saturated water vapor<br />

pressure p0 H2O as def<strong>in</strong>ed <strong>in</strong> Appendix A.1.1. Local conditions <strong>in</strong> the recharge area (e.g. elevation)<br />

may require some corrections of the total pressure as well as the equilibrium concentration due<br />

to the atmosphere’s barometric pressure profile:<br />

�<br />

ptot(h) = p0 · exp − h<br />

�<br />

(2.6)<br />

h0<br />

�<br />

ptot(h) − p<br />

Ci,eq(T, S, ptot(h)) = Ci,eq(T, S, p0) ·<br />

0 H2O<br />

p0 − p0 �<br />

(2.7)<br />

H2O<br />

with h be<strong>in</strong>g the local height above sea level and h0 the local scale height, typically at around<br />

8000 – 8300 m [Kipfer et al., 2002]. Henry’s Law is then expressed as<br />

Ci,eq(T, S, ptot(h)) = zi · (ptot(h) − pH2O)<br />

Hi(T, S)<br />

describ<strong>in</strong>g the gas concentration <strong>in</strong> water due to atmospheric equilibrium at the local site’s<br />

meteorologic parameters.<br />

2.1.2 Noble gas fractions <strong>in</strong> ground water<br />

As shown above, noble gases enter the water phase <strong>by</strong> equilibration with air. In the case<br />

of ground water recharge this equilibration takes place with<strong>in</strong> the vadose zone between the<br />

percolat<strong>in</strong>g meteoric water and the local soil atmosphere. The composition and similarity of soil<br />

air compared to atmospheric air is the ma<strong>in</strong> objective of this study. Isotopic fractionation occurs<br />

when this equilibration takes place due to mass differences of the different noble gas isotopes.<br />

Benson and Krause [1980] found that the solubility of 3 He and 4 He deviates <strong>by</strong> up to 1.8 %.<br />

Other studies [Aeschbach-Hertig, 1994; Beyerle et al., 2000] showed similar effects for Ne and<br />

Ar <strong>in</strong> an order of magnitude of per mil.<br />

The noble gas concentrations found <strong>in</strong> actual ground waters display additional <strong>in</strong>fluences as<br />

shown <strong>in</strong> Figure 2.2, orig<strong>in</strong>at<strong>in</strong>g from various other sources which shall be discussed below.<br />

17<br />

(2.5)<br />

(2.8)


2.1. Noble gas temperatures 2 Theory<br />

Figure 2.2: Approximate representation of components of noble gas concentrations usually<br />

found <strong>in</strong> ground water, standardized to the atmospheric equilibration component. Adapted<br />

from Wieser [2011].<br />

Radiogenic sources<br />

Radioactive decay processes directly and <strong>in</strong>directly produce noble gas isotopes. The yields for<br />

all noble gases except He are so low that possible effects on the noble gas concentrations <strong>in</strong><br />

ground water can be ignored 1 [Stute, 1989]. The 4 He yield is large enough to <strong>in</strong>fluence the 4 He<br />

concentration <strong>in</strong> ground water as 4 He is produced <strong>in</strong> α decays of ma<strong>in</strong>ly primordial radionuclides<br />

found <strong>in</strong> the bedrock. The magnitude of this production depends on the bedrock type, as the<br />

specific activity of various rocks differ [Scheffer and Schachtschabel, 2010]. The particle emitted<br />

<strong>in</strong> α decays ionizes surround<strong>in</strong>g atoms <strong>by</strong> captur<strong>in</strong>g electrons and there<strong>by</strong> transforms to 4 He<br />

α + 2e − −→ 4 He (2.9)<br />

The term terrigenic noble gas describes radiogenic noble gases produced with<strong>in</strong> the Earth’s soil,<br />

bedrock, mantle and crust. Radiogenic 4 He <strong>in</strong> ground water is of terrigenic orig<strong>in</strong>, as well as a<br />

fraction of 3 He produced <strong>in</strong> processes related to the 4 He production [Ballent<strong>in</strong>e and Burnard,<br />

2002]. The part of helium orig<strong>in</strong>at<strong>in</strong>g from Earth’s mantle is characterized <strong>by</strong> a 3 He/ 4 He ratio<br />

significantly higher than the ratio associated to crustal 4 He production [Aeschbach-Hertig et al.,<br />

1999a].<br />

1 The argon isotope 40 Ar is produced <strong>by</strong> the decay of 40 K. With a half-life τ1/2 = 1.25 × 10 9 a, this is only<br />

relevant for very old ground waters.<br />

18


2 Theory 2.1. Noble gas temperatures<br />

The other fraction of 3He <strong>in</strong> ground water is tritiogenic 3He produced <strong>by</strong> the β− decay of the<br />

hydrogen isotope tritium:<br />

3 3 −<br />

H −→ He + e + ¯νe<br />

(2.10)<br />

Tritium is produced <strong>in</strong> the upper lithosphere through a reaction of lithium with cosmogenic<br />

neutrons and <strong>in</strong> the upper atmosphere through nuclear spallation of nitrogen atoms, <strong>in</strong>duced <strong>by</strong><br />

cosmogenic neutrons as well. An anthropogenic fraction was <strong>in</strong>duced <strong>in</strong>to the atmosphere dur<strong>in</strong>g<br />

the 1960s <strong>by</strong> atmospheric hydrogen bomb tests. Atmospheric tritium enters the aquifers bound<br />

<strong>in</strong> water molecules, where its decay leads to a rise <strong>in</strong> 3 He concentrations. Due to these additional<br />

sources, He is usually not used to calculate noble gas temperatures [Stute et al., 1995].<br />

Excess air<br />

Additionally to the aforementioned additional noble gas sources, ground water noble gas concentrations<br />

usually display a surplus that is air derived. In some cases this surplus is fractionated<br />

relative to atmospheric air with the heavy noble gases more enriched than the lighter<br />

ones [Stute et al., 1995; Aeschbach-Hertig et al., 1999b, 2000]. The orig<strong>in</strong> of this surplus, called<br />

excess air <strong>by</strong> Heaton and Vogel [1981], lies <strong>in</strong> gas bubbles caught <strong>in</strong> the pore space that get<br />

trapped <strong>by</strong> ground water table fluctuations and are transported <strong>in</strong>to the saturated soil zone.<br />

There, <strong>in</strong>creased hydrostatic pressure causes partial or complete dissolution of the gas <strong>in</strong>to the<br />

ground water. The amount of excess air <strong>in</strong>troduced is highly variable and dependent on the soil<br />

structure of the base of the vadose zone and precipitation patterns [Heaton and Vogel, 1981].<br />

This allows for excess air to be used as proxy for past environmental conditions like water table<br />

fluctuation <strong>in</strong> arid and semi-arid areas [Aeschbach-Hertig et al., 2002; Beyerle et al., 2003;<br />

Wieser, 2011].<br />

2.1.3 Excess air model<strong>in</strong>g approaches<br />

Results of measurements of dissolved noble gases <strong>in</strong> ground waters are total gas concentrations.<br />

As outl<strong>in</strong>ed above, the measured total concentrations are the sum of several different mechanisms<br />

that <strong>in</strong>troduce noble gases <strong>in</strong>to the water:<br />

Ci,m = Ci,eq + Ci,ex + Ci,rad + Ci,ter + Ci,tri<br />

(2.11)<br />

where m stands for measured, eq for atmospheric equilibration, ex for excess air, rad for radiogenic,<br />

ter for terrigenic and tri for tritiogenic. This relation can be simplified for certa<strong>in</strong> noble<br />

gas isotopes as several of them ( 20 Ne, 36 Ar and virtually all Kr and Xe isotopes [Kipfer et al.,<br />

2002]) are only <strong>in</strong>fluenced <strong>by</strong> atmospheric equilibrium and excess air. Model<strong>in</strong>g the excess air<br />

fraction is therefore usually sufficient to calculate noble gas temperatures from measured total<br />

gas concentrations. To describe the effect of excess air on measured noble gas concentrations,<br />

various models were created and applied to ground water records. Some of them will be outl<strong>in</strong>ed<br />

below.<br />

The actual separation of components and calculation of noble gas temperatures is done <strong>by</strong> <strong>in</strong>verse<br />

model<strong>in</strong>g, optimiz<strong>in</strong>g various model specific free parameters to achieve the best agreement of<br />

19


2.1. Noble gas temperatures 2 Theory<br />

modeled and measured data. The quality of this match<strong>in</strong>g is assessed <strong>by</strong> us<strong>in</strong>g the χ 2 method<br />

that weights the deviation between the measured and the predicted concentrations with the<br />

measurement accuracy σi:<br />

UA, PR and MR model<br />

χ 2 = �<br />

i<br />

�<br />

Ci,m − Cmodel �2 i<br />

σ2 i<br />

(2.12)<br />

The first model created to account for the presence of an excess air component was the unfractionated<br />

air model, or UA model. It is based on the assumption that enclosed air bubbles are<br />

completely dissolved <strong>in</strong> the water at or below the ground water table, lead<strong>in</strong>g to an atmospheric<br />

composition of the excess air component of the noble gas concentration without any signs of<br />

fractionation, as described <strong>by</strong><br />

C UA<br />

i = Ci,eq + A · Ci,atm (2.13)<br />

where Ci,atm is the unfractioned excess air component and A is the fraction of air entrapped <strong>in</strong><br />

the water volume, a scal<strong>in</strong>g parameter of this model. Us<strong>in</strong>g Henry’s law (Equation 2.1), this<br />

relation is expressed as<br />

C UA<br />

i = Ci,eq(1 + A · Hi) (2.14)<br />

This basic assumption of complete dissolution did not prove to be realistic though, as Stute et al.<br />

[1995] showed excess air components <strong>in</strong> ground water to be enriched <strong>in</strong> the heavier noble gases.<br />

They <strong>in</strong>troduced a modification of the UA model, called partial re-equilibration model, or PR<br />

model, assum<strong>in</strong>g diffusive gas loss affect<strong>in</strong>g the completely dissolved excess air component. The<br />

mass and temperature dependent diffusion coefficients of the <strong>in</strong>dividual noble gases [Jähne et al.,<br />

1987] would then account for the occurr<strong>in</strong>g fractionation. Inter-isotopic fractionation effects are<br />

not observed however [Peeters et al., 2003], on which the PR model’s applicability was criticized.<br />

Whether this effect should actually be expected at all to occur <strong>in</strong> a manner similar to <strong>in</strong>ternoble-gas<br />

fractionation has been questioned though [Bourg and Sposito, 2008]. The PR model<br />

can be expressed as [Aeschbach-Hertig et al., 2008]:<br />

�<br />

�<br />

� ��<br />

β<br />

C PR<br />

i<br />

= Ci,eq<br />

1 + A · Hi · exp<br />

−FPR<br />

� Di<br />

DNe<br />

(2.15)<br />

where FPR is a parameter characteriz<strong>in</strong>g the amount of excess air loss, Di are the noble gas<br />

diffusion coefficients and 0.5 ≤ β ≤ 1 is a model parameter from gas transfer theory. S<strong>in</strong>ce<br />

this s<strong>in</strong>gle step excess air <strong>in</strong>trusion and degass<strong>in</strong>g leads to <strong>in</strong>itial Ne excess amounts requir<strong>in</strong>g<br />

unrealistic physical conditions <strong>in</strong> some studies [Kipfer et al., 2002] and failed to describe certa<strong>in</strong><br />

noble gas records [Ballent<strong>in</strong>e and Hall, 1999], a variation of the PR model was <strong>in</strong>troduced as the<br />

multi-step partial re-equilibration model, or MR model, <strong>by</strong> Kipfer et al. [2002]. The MR model<br />

assumes a physically more realistic occurrence of multiple (n) excess air <strong>in</strong>trusions and degass<strong>in</strong>g<br />

steps, conta<strong>in</strong><strong>in</strong>g the PR model as the special case n = 1:<br />

�<br />

�<br />

� ��<br />

β<br />

C MR<br />

i<br />

= Ci,eq<br />

1 + A · Hi ·<br />

n�<br />

exp<br />

k=1<br />

20<br />

−k · FPR<br />

� Di<br />

DNe<br />

(2.16)


2 Theory 2.1. Noble gas temperatures<br />

Unfractionated excess air (UA)<br />

AEW & air air entrapment complete dissolution<br />

Oxygen depletion (OD)<br />

AEW & O2 depl. air air entrapment complete dissolution<br />

Partial re-equilibration (PR)<br />

AEW & air air entrapment complete dissolution di�usive gas loss<br />

Closed-system equilibration (CE)<br />

Multi-step partial re-equilibration (MR)<br />

AEW & air air entrapment partial dissolution,<br />

equilibration<br />

water-gas separation<br />

Figure 2.3: Schematic illustration of the mechanisms lead<strong>in</strong>g to excess air as proposed <strong>by</strong><br />

some of the different presented models. Adapted from Wieser [2011].<br />

21


2.1. Noble gas temperatures 2 Theory<br />

CE model<br />

The closed-system equilibration model, or CE model, was formulated <strong>by</strong> Aeschbach-Hertig et al.<br />

[2002] and tries to expla<strong>in</strong> the noble gas fractionation of the excess air component without<br />

rely<strong>in</strong>g on diffusive degass<strong>in</strong>g effects as the PR and MR do. Assum<strong>in</strong>g smaller hydrostatic<br />

pressure <strong>in</strong>creases dur<strong>in</strong>g ground water table fluctuations, <strong>in</strong>capable of caus<strong>in</strong>g the complete<br />

dissolution of entrapped air bubbles leads to rema<strong>in</strong><strong>in</strong>g bubbles below the water table. These<br />

bubbles, represent<strong>in</strong>g limited gas reservoirs, then equilibrate with the surround<strong>in</strong>g water lead<strong>in</strong>g<br />

to an <strong>in</strong>ter-nobel-gas fractionation caused <strong>by</strong> the differ<strong>in</strong>g solubilities, but not lead<strong>in</strong>g to <strong>in</strong>terisotopic<br />

fractionation possibly caused <strong>by</strong> diffusive processes. A mathematical description can be<br />

formulated as<br />

�<br />

C CE<br />

i<br />

�<br />

1 + A ′<br />

Hi<br />

= Ci,eq ·<br />

1 + BHi<br />

(2.17)<br />

with A ′ be<strong>in</strong>g the <strong>in</strong>itial ratio of entrapped air volume to ground water volume and B be<strong>in</strong>g<br />

the f<strong>in</strong>al ratio of rema<strong>in</strong><strong>in</strong>g air volume to water volume. This allows for the def<strong>in</strong>ition of a<br />

fractionation factor FCE = B/A ′ describ<strong>in</strong>g the magnitude of excess air <strong>in</strong>fluence, FCE < 1<br />

<strong>in</strong>dicates excess air <strong>in</strong>fluence, for FCE = 1 the CE model transforms to the UA model and for<br />

FCE > 1 it describes degassed ground water.<br />

OD and GR model<br />

Studies on recent ground waters <strong>by</strong> Ma et al. [2004] and Hall et al. [2005] showed that modern<br />

noble gas temperatures showed a systematical underestimation of the mean annual air temperatures<br />

(MAAT) <strong>by</strong> several degrees Celsius. This prompted Hall et al. [2005] to <strong>in</strong>troduce the oxygen<br />

depletion model, or OD model, provid<strong>in</strong>g a different explanation for the excess air component<br />

as it moved away from the assumption of Stute and Schlosser [1993] that equilibration with<strong>in</strong><br />

the soil takes place between water and atmospherically composed air. While Stute and Schlosser<br />

[1993] were aware of fluctuations <strong>in</strong> soil atmospheres <strong>in</strong> O2 and CO2 composition, they estimated<br />

the effect on noble gas partial pressures to be negligible based on Brook et al. [1983], giv<strong>in</strong>g an<br />

upper limit of 2 Vol% CO2 <strong>in</strong> usual soil atmospheres dur<strong>in</strong>g grow<strong>in</strong>g season. A compilation<br />

of worldwide data on soil atmosphere CO2 concentrations <strong>by</strong> Amundson and Davidson [1990]<br />

shows that CO2 concentrations, while fluctuat<strong>in</strong>g strongly, generally <strong>in</strong>crease with depth and<br />

were observed to range between 0.04 and 13.0 Vol%. Hall et al. [2005] argue that biological<br />

processes deplet<strong>in</strong>g 2 O2 and produc<strong>in</strong>g CO2 (see Section 2.3), comb<strong>in</strong>ed with CO2 removal from<br />

the soil air due to its high solubility <strong>in</strong> water may lead to a pressure deficit that is compensated<br />

<strong>by</strong> ris<strong>in</strong>g partial pressures of the rema<strong>in</strong><strong>in</strong>g gases.<br />

The magnitude of this proposed effect of O2 depletion with an accompany<strong>in</strong>g CO2 deficit on<br />

the noble gas partial pressures is expressed <strong>by</strong> the parameter POD [Schneider, 2010]: The<br />

parameter 0 ≤ α ≤ 1 describes the degree of O2 depletion, while Z is the amount of O2<br />

given <strong>in</strong> Vol%, Zatm = 20.9 % [Porcelli et al., 2002]. The amount of O2 <strong>in</strong> soil air is given <strong>by</strong><br />

2 In the context of the OD model, the term oxygen depletion is often used synonymously with a noble gas partial<br />

pressure <strong>in</strong>crease. In this study however, the term shall only denote a decrease <strong>in</strong> O2 levels without imply<strong>in</strong>g<br />

further changes of the soil atmosphere.<br />

22


2 Theory 2.1. Noble gas temperatures<br />

Figure 2.4: Theoretical <strong>in</strong>crease of partial pressures of soil gases other than O2 if O2 depletion<br />

isn’t fully compensated <strong>by</strong> CO2 <strong>in</strong>crease. The expected realistically relevant <strong>in</strong>terval of<br />

O2+CO2 at the sampled sites is 20.9 – 16.9 Vol%, lead<strong>in</strong>g to theoretically expected noble<br />

gas partial pressure <strong>in</strong>creases of up to 5 % relative to atmospheric noble gas partial pressures.<br />

From Schneider [2010].<br />

Zsoil,O2 = α · Zatm. Correspond<strong>in</strong>gly, the partial pressures of the rema<strong>in</strong><strong>in</strong>g gases Zsoil,i change<br />

<strong>by</strong> a factor POD = Zsoil,i<br />

, related to Zatm,O2 <strong>by</strong><br />

Zatm,i<br />

α · Zatm,O2 + POD · �<br />

Zatm,i = 1 (2.18)<br />

i<br />

For atmospheric air without O2 with �<br />

i Zatm,i = 79.1 % follows<br />

POD =<br />

1 − 0.209 · α<br />

0.791<br />

=⇒ 1 ≤ POD ≤ 1.264 (2.19)<br />

The theoretical maximum <strong>in</strong>crease of the sum of all partial pressures of the rema<strong>in</strong><strong>in</strong>g gases<br />

i would therefore be 26.4 % as shown <strong>in</strong> Figure 2.4 though such a scenario, requir<strong>in</strong>g full O2<br />

consumption without any CO2 production (or complete CO2 removal), is hardly realistic. The<br />

previously observed CO2 deficits at one of this study’s sampl<strong>in</strong>g sites, result<strong>in</strong>g from O2 depletion,<br />

reached up to 4 % [Schneider, 2010], lead<strong>in</strong>g to a theoretically expected <strong>in</strong>crease of noble<br />

gas partial pressures of around 5 %.<br />

23


2.2. Physical processes and properties of soils 2 Theory<br />

The OD model resembles the UA model <strong>in</strong> formulation [Aeschbach-Hertig et al., 2008] and can<br />

be <strong>in</strong>corporated <strong>in</strong>to the exist<strong>in</strong>g models:<br />

C OD<br />

i = Ci,eq(POD + A · Hi) (2.20)<br />

A modification of the OD model add<strong>in</strong>g partial re-equilibration at the air-water boundary layer,<br />

the gas diffusion relaxation model, or GR model, was <strong>in</strong>troduced <strong>by</strong> Sun et al. [2008] to describe<br />

a noble gas record from a Michigan aquifer:<br />

C GR<br />

i<br />

= Ci,eq<br />

�<br />

POD + A · Hi · exp<br />

�<br />

−FGRD β<br />

i<br />

��<br />

(2.21)<br />

where FGR is a parameter that depends on the time taken for the gas transfer as well as on the<br />

length scale of the boundary layer.<br />

Compar<strong>in</strong>g the different models<br />

Sun et al. [2010] showed <strong>in</strong> a recent comparison of the UA, the CE, the OD and the GR model<br />

that, while the absolute noble gas temperatures deviate between the different models, temperature<br />

differences are reproduced quite consistently <strong>by</strong> all models. While the accuracy of the fits<br />

achieved <strong>by</strong> the different models and the respective χ 2 values give an <strong>in</strong>dication of which model<br />

might describe the given dataset best, the decisive criteria which model to employ should be the<br />

physical relevance of the model concepts and of the parameters result<strong>in</strong>g from the optimization<br />

process. Therefore a more detailed analysis of the processes <strong>in</strong>volved <strong>in</strong> noble gas dissolution<br />

is called for, like the observation of excess air formation done <strong>by</strong> Holocher et al. [2002] and<br />

Klump et al. [2007, 2008].<br />

This study’s <strong>in</strong>tention is to test the physical foundation of the OD model’s assumption of <strong>in</strong>creased<br />

noble gas partial pressures <strong>in</strong> soil atmospheres to better assess the significance of the<br />

good χ 2 values found <strong>by</strong> Hall et al. [2005] and Castro et al. [2007] us<strong>in</strong>g the OD model approach.<br />

2.2 Physical processes and properties of soils<br />

Understand<strong>in</strong>g the soil regime is essential, as it provides the environment <strong>in</strong> which the components<br />

of meteoric water relevant to noble gas paleoclimatology <strong>in</strong>teract (ideally) for the last<br />

time with the atmosphere before the water enters the aquifer. Both soil temperature and soil<br />

atmosphere composition are related to conditions above the surface, but are not necessarily<br />

identical and therefore an understand<strong>in</strong>g of these relations is required to be able to <strong>in</strong>terpret the<br />

measured noble gas concentrations. The follow<strong>in</strong>g summary of the soil regime’s properties and<br />

processes is largely based on Scheffer and Schachtschabel [2010].<br />

2.2.1 Subsurface thermal regime<br />

The ma<strong>in</strong> heat source dom<strong>in</strong>at<strong>in</strong>g the upper soil’s temperature profile is the radiation of the sun.<br />

Geothermal heat flux can generally be neglected for depths relevant to ground water recharge<br />

24


2 Theory 2.2. Physical processes and properties of soils<br />

Temperature [°C]<br />

50<br />

40<br />

30<br />

20<br />

10<br />

0 0<br />

0 4 8 12 16 20 24 J F M A M J J A S O N D<br />

Time [h]<br />

(a) (b)<br />

30<br />

Temperature [°C]<br />

35<br />

25<br />

20<br />

15<br />

10<br />

5<br />

Month<br />

Figure 2.5: Temperature variations <strong>in</strong> soil regimes. (a) Diurnal variations at different depths<br />

at a loam soil. (b) Annual variation. From Hillel [1980].<br />

as it usually takes place too close to the soil surface to be affected <strong>by</strong> geothermal heat. Heat<br />

loss from the soil is caused <strong>by</strong> emission of <strong>in</strong>frared radiation from the soil and evaporation.<br />

The driv<strong>in</strong>g force of conductive and convective heat transfer with<strong>in</strong> the soil are up- and downward<br />

temperature and pressure gradients. Convective heat transfer is usually mediated <strong>by</strong> the<br />

movement of liquid water and water vapor through the soil structure. The importance of water<br />

<strong>in</strong> the soil thermal regime is also visible <strong>in</strong> the fact that dry soils heat up more quickly and<br />

to a higher temperature than wet soils, as water acts damp<strong>in</strong>g on heat transfer due to its high<br />

heat capacity compared to air [Scheffer and Schachtschabel, 2010; Wieser, 2011]. Infiltrat<strong>in</strong>g<br />

meteoric water does, aside from convectively <strong>in</strong>troduc<strong>in</strong>g heat, cause significant temperature<br />

changes depend<strong>in</strong>g on <strong>in</strong>itial water content of the soil due to release of thermal energy dur<strong>in</strong>g<br />

adsorption, called heat of wett<strong>in</strong>g [Prunty and Bell, 2005].<br />

However, as a rough approximation, the soil temperature at a certa<strong>in</strong> depth can be described as<br />

oscillat<strong>in</strong>g s<strong>in</strong>usoidally around an average value, driven <strong>by</strong> atmospheric temperature changes (see<br />

Figure 2.5). S<strong>in</strong>usoidal approximation of the soil temperature as a function of depth and time<br />

T (z, t) with the boundary conditions T (0, t) = ¯ T + A0 s<strong>in</strong> ωt and T (z = ∞, t) = ¯ T , with z = 0<br />

be<strong>in</strong>g the soil surface and ¯ T the diurnal average surface temperature, leads to the <strong>in</strong>troduction<br />

of a damp<strong>in</strong>g depth d and a phase shift −z/d [Hillel, 1980]:<br />

T (z, t) = ¯ �<br />

T + A0 · exp − z<br />

� �<br />

· s<strong>in</strong> ωt −<br />

d<br />

z<br />

�<br />

(2.22)<br />

d<br />

The damp<strong>in</strong>g depth d is related to the thermal properties of the soil and is therefore <strong>in</strong>dividual<br />

for any given soil regime. While the <strong>in</strong>clusion of the annual temperature fluctuations complicates<br />

the situation, the basic characterization of soil temperature at fixed depths follow<strong>in</strong>g a harmonic<br />

oscillation around an average value as a crude description of soil temperature regimes rema<strong>in</strong>s<br />

viable. An detailed description of temperature model<strong>in</strong>g is given <strong>by</strong> Saito and ˇ Sim˚unek [2009].<br />

25


2.2. Physical processes and properties of soils 2 Theory<br />

The temperatures calculated from dissolved noble gases <strong>in</strong> ground water ideally represent the<br />

mean annual air temperature of the <strong>in</strong>filtration region. Yet this is not necessarily the case, as<br />

the previously described propagation of heat with<strong>in</strong> the soil is modulated <strong>by</strong> various mechanisms<br />

lead<strong>in</strong>g to a decoupl<strong>in</strong>g of MAAT and mean annual soil temperatures. Some of these mechanisms<br />

are meteorological and biological <strong>in</strong>fluences like cloud cover, precipitation, evaporation and<br />

changes <strong>in</strong> albedo due to snow cover and vegetation and there<strong>by</strong> affect the soil temperature<br />

regime on various spatial and time scales.<br />

Soil temperatures are also <strong>in</strong>directly affect<strong>in</strong>g noble gas temperatures as they do <strong>in</strong>fluence the<br />

magnitude of microbiological activity with<strong>in</strong> the soil [Ratkowsky et al., 1982], there<strong>by</strong> modulat<strong>in</strong>g<br />

the soil atmosphere composition <strong>in</strong> regard to O2 and CO2 [Fang and Moncrieff, 1999;<br />

Howard and Howard, 1993], and possibly the noble gas partial pressures.<br />

2.2.2 Soil structure<br />

Air space with<strong>in</strong> the soil matrix is def<strong>in</strong>ed <strong>by</strong> the available pore space, which is dependent on the<br />

type of soil, its deposition as well as gra<strong>in</strong> size and type. The porosity of a soil gives the fraction<br />

of pore space volume <strong>in</strong> the total soil volume. Scheffer and Schachtschabel [2010] estimate the<br />

porosity of silts as 55 – 30 % and that of clays as 65 – 35 %.<br />

S<strong>in</strong>ce soils always conta<strong>in</strong> a certa<strong>in</strong> amount of water, even <strong>in</strong> arid climates, the pore space<br />

available to gases with<strong>in</strong> the vadose (unsaturated) zone is limited <strong>by</strong> the water content. Water<br />

with<strong>in</strong> the soil can be divided <strong>in</strong>to two fractions, one be<strong>in</strong>g connate water that is held aga<strong>in</strong>st<br />

gravity <strong>by</strong> adsorption and capillary forces, fill<strong>in</strong>g the smaller pores. The other fraction is seep<strong>in</strong>g<br />

water mov<strong>in</strong>g downwards <strong>by</strong> gravitative forc<strong>in</strong>g [Hölt<strong>in</strong>g and Coldewey, 2009].<br />

The result<strong>in</strong>g available air space at maximum field capacity 3 for clay soils is given as 10 – 25 %<br />

<strong>by</strong> Scheffer and Schachtschabel [2010], it is usually higher though s<strong>in</strong>ce the water content of<br />

maximum field capacity is rarely realized.<br />

2.2.3 Gas transport processes<br />

Convective mass transport of gases with<strong>in</strong> the soil requires pressure gradients as the driv<strong>in</strong>g force,<br />

like changes <strong>in</strong> barometric pressure or temperature. As shown above, anaerobic production of<br />

methane is able to produce such high amounts of additional gas that convective transport of<br />

CH4 towards the soil surface occurs. Dur<strong>in</strong>g rapid <strong>in</strong>filtration, water can displace gas from small<br />

pore spaces, lead<strong>in</strong>g to fast convective degass<strong>in</strong>g of air [Liu et al., 2002], as well as transport<br />

of dissolved gas. Fluctuations of the ground water table also <strong>in</strong>duce convective gas transport.<br />

However, the sum of these effects is of little relevance, Amundson and Davidson [1990] estimate<br />

that less than 10 % of the total loss of CO2 from soils are caused <strong>by</strong> convective mass transport.<br />

Diffusive gas transport requires only partial pressure (concentration) gradients <strong>in</strong>stead of total<br />

pressure gradients as a driv<strong>in</strong>g force and is therefore much more prevalent than convective mass<br />

3 Measure of soil water content that is reta<strong>in</strong>ed aga<strong>in</strong>st gravitative dra<strong>in</strong>age [Hölt<strong>in</strong>g and Coldewey, 2009].<br />

26


2 Theory 2.2. Physical processes and properties of soils<br />

Di�usion coe�. D S [cm2 s-1]<br />

0.015<br />

0.010<br />

0.005<br />

Sand<br />

Clay<br />

Silt<br />

waterlogged soils<br />

0 5 10 15 20 25<br />

Air content n [%]<br />

Figure 2.6: CO2 diffusion coefficient DS of various soils <strong>in</strong> relation to their air content nA.<br />

From Richter and Großgebauer [1978].<br />

transport. The limit<strong>in</strong>g factor of diffusive transport <strong>in</strong> the soil is its water content s<strong>in</strong>ce diffusion<br />

with<strong>in</strong> water is 10 −4 times smaller than <strong>in</strong> air. Water mass and distribution with<strong>in</strong> the pore<br />

space therefore has a large <strong>in</strong>fluence on the diffusion dynamics, as shown <strong>in</strong> Figure 2.6. Diffusive<br />

transport <strong>in</strong> a stationary system is described <strong>by</strong> Fick’s first law:<br />

dc<br />

I = −DS<br />

dx<br />

A<br />

(2.23)<br />

where I is the flow of gas <strong>in</strong> units of mol s −1 cm −2 , dc/dx is the concentration gradient and<br />

DS is the diffusion coefficient of the respective gas <strong>in</strong> the soil, <strong>in</strong> units of cm 2 s −1 . Changes <strong>in</strong><br />

concentration C due to this diffusive flow are accounted for <strong>by</strong> us<strong>in</strong>g Fick’s second law, describ<strong>in</strong>g<br />

a non-stationary situation:<br />

∂C<br />

∂t<br />

∂<br />

= DS<br />

2C ∂x2 (2.24)<br />

Because of the <strong>in</strong>teraction of the gas molecules with the medium <strong>in</strong> which they are mov<strong>in</strong>g,<br />

the diffusion coefficient is not only dependent on the molecule size but on various other factors.<br />

These factors are accounted for <strong>by</strong> def<strong>in</strong><strong>in</strong>g the diffusion coefficient <strong>in</strong> soil DS <strong>by</strong> weight<strong>in</strong>g the<br />

diffusion coefficient of air DA with the available air filled pore space nA and the turtuosity τ,<br />

which is an approximation of the <strong>in</strong>fluence of geometric structure of the pore space on diffusion:<br />

DS = − 1<br />

τ · nA · DA<br />

(2.25)<br />

Various properties of the soil can therefore <strong>in</strong>hibit diffusive flow, most importantly the amount<br />

and spatial distribution of water with<strong>in</strong> the soil matrix as well as its structure.<br />

27


2.3. Soil atmosphere composition 2 Theory<br />

2.3 Soil atmosphere composition<br />

2.3.1 Sources, s<strong>in</strong>ks and profiles of O2 and CO2<br />

The most important subsurface source of CO2 and s<strong>in</strong>k of O2 is soil respiration, which summarizes<br />

the production of CO2 caused <strong>by</strong> root respiration and microbiological processes. How<br />

important each process is <strong>in</strong> a given soil is hard to estimate and varies with many parameters.<br />

Usually the fraction of CO2 production caused <strong>by</strong> root respiration is given as rang<strong>in</strong>g from 0<br />

– 50 % [Amundson and Davidson, 1990; Scheffer and Schachtschabel, 2010]. Tang et al. [2005]<br />

give an annual mean of 44 % and a grow<strong>in</strong>g season average of 56 % <strong>in</strong> a forest soil.<br />

In the presence of O2, CO2 is produced <strong>by</strong> aerobic bacteria ga<strong>in</strong><strong>in</strong>g energy from the decomposition<br />

of glucose:<br />

C6H12O6 + 6 O2 −→ 6 CO2 + 6 H2O + 2800 kJ/mol (2.26)<br />

This reaction produces equimolar amounts of gas and therefore does not <strong>in</strong>fluence the partial<br />

pressure equilibrium of the soil atmosphere directly as the sum of O2 and CO2 should still be<br />

20.9 Vol%. The produced CO2 has a much higher solubility <strong>in</strong> water than O2, lead<strong>in</strong>g to a<br />

reduction of CO2 partial pressure. In acidic soils (pH < 5) this removal of CO2 from the gaseous<br />

phase is generally described <strong>by</strong> the follow<strong>in</strong>g equilibrium reaction [Yamaguchi et al., 1967]:<br />

CO2(g) + H2O ⇋ H2CO3 ⇋ HCO − 3 + H+ ⇋ CO 2−<br />

3<br />

+ 2 H+<br />

(2.27)<br />

Oxygen depletion <strong>by</strong> microbial activity can therefore lead to a deficit <strong>in</strong> partial pressure, which<br />

is the basis for the proposed OD model presented <strong>in</strong> Section 2.1.3.<br />

When O2 is not available (close to the water table or under waterlogged conditions), anaerobic<br />

<strong>in</strong>stead of aerobic bacteria flourish, lead<strong>in</strong>g to different carbohydrate decomposition processes.<br />

In the absence of O2 the oxidation of glucose<br />

C6H12O6 −→ 2 CH3 CO COOH + 4 H + + 4 e −<br />

(2.28)<br />

requires a match<strong>in</strong>g reduction reaction to be provided <strong>by</strong> the microorganism, lead<strong>in</strong>g to various<br />

possible reactions [Rowell, 1997] such as denitrification or the production of methane:<br />

2 NO − 3 + 12 H+ + 10 e − −→ N2 + 6 H2O (2.29)<br />

CO2 + 8 H + + 8 e − −→ CH4 + 2 H2O (2.30)<br />

Another anaerobic process is given <strong>by</strong> Scheffer and Schachtschabel [2010] as<br />

C6H12O6 −→ 3 CO2 + 3 CH4 + 188 kJ/mol (2.31)<br />

Anaerobic processes are mostly non-equimolar regard<strong>in</strong>g the gas phase, lead<strong>in</strong>g to an <strong>in</strong>crease of<br />

gas concentration and therefore a partial pressure gradient. In the case of waterlogged marshes<br />

this can cause significant flow and emissions of CH4 from the soil.<br />

In general, microbial activity reduces the amount of O2 <strong>in</strong> the soil and <strong>in</strong>creases the amount of<br />

CO2, CH4 and of some nitrogenous gases. The ma<strong>in</strong> source for O2 is the atmosphere, result<strong>in</strong>g <strong>in</strong><br />

28


2 Theory 2.3. Soil atmosphere composition<br />

Gas concentrations [%]<br />

Depth [cm]<br />

0<br />

10<br />

20<br />

30<br />

40<br />

CO2 production [mg m-2 cm-1 h-1]<br />

10 20<br />

30 40 50 60<br />

Autumn day<br />

Summer day<br />

Figure 2.7: Depth dependency of the production of CO2 <strong>in</strong> a loess luvisol soil. From<br />

Richter and Großgebauer [1978].<br />

20<br />

15<br />

10<br />

5<br />

0<br />

30 cm depth<br />

O2<br />

CO2<br />

Mar Jun<br />

Sep Dec<br />

20<br />

15<br />

10<br />

5<br />

0<br />

90 cm depth<br />

O2<br />

CO2<br />

Mar Jun Sep Dec<br />

Figure 2.8: O2 and CO2 concentrations of soil atmospheres at 30 and 90 cm depth.<br />

The dotted l<strong>in</strong>e <strong>in</strong>dicates a sandy silt while the solid l<strong>in</strong>e represents a silty clay. From<br />

Boynton and Compton [1944].<br />

29


2.3. Soil atmosphere composition 2 Theory<br />

Depth [m]<br />

0.0<br />

0<br />

0.2 0.4 0.6 0.8<br />

1<br />

2<br />

Jan.<br />

Feb.<br />

Mar.<br />

Apr.<br />

CO2 concentration [Vol%]<br />

0.0 0.2 0.4 0.6 0.8<br />

May<br />

Jun.<br />

Jul.<br />

Aug.<br />

0.0 0.2 0.4 0.6 0.8<br />

Sep.<br />

Oct.<br />

Nov.<br />

Dec.<br />

Figure 2.9: Monthly averaged CO2 concentration <strong>in</strong> relation to depth and its seasonal development<br />

at a mounta<strong>in</strong> forest site <strong>in</strong> Japan. From Hamada and Tanaka [2001].<br />

decreas<strong>in</strong>g O2 concentrations with<strong>in</strong> the soil with <strong>in</strong>creas<strong>in</strong>g depth, start<strong>in</strong>g at the atmospheric<br />

20.9 Vol%. The ma<strong>in</strong> source of CO2 lies with<strong>in</strong> the soil, its spatial distribution and strength<br />

dependent on microbial diversity, organic nutrient reservoirs and temperature. The highest<br />

CO2 production takes place with<strong>in</strong> the first 10 – 50 cm of soil dur<strong>in</strong>g grow<strong>in</strong>g season, as shown<br />

<strong>in</strong> Figure 2.7. This corresponds to the fact that microbial biomass is one to two orders of<br />

magnitude higher close to the surface than at 2 m depth and is likely caused <strong>by</strong> the availability<br />

of nutrients [Fierer et al., 2003]. The ma<strong>in</strong> s<strong>in</strong>k of CO2 is the atmosphere. CO2 concentrations<br />

<strong>in</strong> soil atmospheres were observed to reach up to 13.0 Vol% [Amundson and Davidson, 1990],<br />

<strong>in</strong> laboratory experiments on soil columns even up to 17.0 Vol% [Yamaguchi et al., 1967]. Due<br />

to the production zone located close to the surface, diffusive transport of CO2 exists directed<br />

both <strong>in</strong>to the atmosphere as well as <strong>in</strong>to deeper soil dur<strong>in</strong>g spr<strong>in</strong>g and early summer, lead<strong>in</strong>g to<br />

<strong>in</strong>creased CO2 concentrations there. The diffusive flow with<strong>in</strong> the deeper soil regions changes<br />

direction <strong>in</strong> late summer and w<strong>in</strong>ter when microbial activity ceases. The CO2 concentration<br />

profile therefore fluctuates heavily on diurnal as well as annual scales and is closely related to<br />

the O2 concentrations, as shown <strong>in</strong> Figures 2.8 and 2.9.<br />

30


2 Theory 2.3. Soil atmosphere composition<br />

Respiration rate K [g CO2 m-2 d-1]<br />

8<br />

6<br />

4<br />

2<br />

0<br />

0<br />

Jan<br />

Mar Apr<br />

Feb<br />

Dec<br />

Nov<br />

5 10<br />

K0 = 1.2<br />

May<br />

Oct<br />

Aug<br />

Jun<br />

K0 = 0.9<br />

average soil temperature at 10 cm depth [°C]<br />

Figure 2.10: Relation between average soil temperature and daily respiration rate K of CO2<br />

from a fallow soil. From Rowell [1997].<br />

2.3.2 Variability of soil respiration<br />

The dynamics of subsurface CO2 concentration can be described for model<strong>in</strong>g purposes as <strong>by</strong><br />

Riveros-Iregui et al. [2011] based on production terms stemm<strong>in</strong>g from autotrophic (root respiration<br />

<strong>by</strong> carbon fixat<strong>in</strong>g organisms) and heterotrophic (mircobiotic organisms rely<strong>in</strong>g on the<br />

availability of organic carbon) activities and a diffusive transport term:<br />

∂CCO2<br />

nA<br />

∂t<br />

= − ∂<br />

∂z<br />

�<br />

DS<br />

∂CCO2<br />

∂z<br />

15<br />

Sep<br />

Jul<br />

20<br />

4<br />

3<br />

2<br />

1<br />

0<br />

Respiration rate K [10-3 m3 CO2 m-2 d-1]<br />

�<br />

+ γA(B, P AR, Θ) + γH(M, TS, Θ) (2.32)<br />

where γA describes the autotrophic and γH the heterotrophic component. B parameterizes root<br />

biomass, P AR is photosynthetically active radiation, M is soil organic matter, Θ is soil water<br />

content and TS is soil temperature.<br />

The driv<strong>in</strong>g force of soil respiration is microbial activity, which is highly dependent on soil<br />

temperatures. The temperature dependency of the rates K of most chemical and biological<br />

reactions is calculated <strong>by</strong> us<strong>in</strong>g the Arrhenius equation<br />

�<br />

K = K0 · exp − E<br />

�<br />

(2.33)<br />

RT<br />

31


2.3. Soil atmosphere composition 2 Theory<br />

where K0 is a reaction specific constant, E the reaction specific activation energy and R the<br />

universal gas constant. However, due to denaturation of essential prote<strong>in</strong>s at high temperatures,<br />

this approximation is only valid over a limited temperature <strong>in</strong>terval. W<strong>in</strong>kler et al. [1996] found<br />

that respiration rates of CO2 from soils generally followed the exponential predictions of the<br />

Arrhenius equation between soil temperatures of 4 – 38 ◦ C, see also Figure 2.10. The Q10 value<br />

describes the <strong>in</strong>crease of respiration rate per 10 ◦ C and was found to be between 1.7 – 1.9,<br />

decreas<strong>in</strong>g l<strong>in</strong>early with <strong>in</strong>creas<strong>in</strong>g temperature. This means that for every 10 ◦ C <strong>in</strong>crease <strong>in</strong> soil<br />

temperature with<strong>in</strong> the 4 – 38 ◦ C range, the soil respiration would nearly double, if not limited<br />

<strong>by</strong> other parameters.<br />

This is the reason temperature dependency is the primary controll<strong>in</strong>g factor on soil respiration,<br />

<strong>in</strong>troduc<strong>in</strong>g an annual variability. Temperature also <strong>in</strong>fluences the diurnal cycle of CO2<br />

production, however under certa<strong>in</strong> conditions (high soil water content, production larger than<br />

transport) this <strong>in</strong>fluence can be quite complicated, result<strong>in</strong>g <strong>in</strong> a hysteresis between production<br />

rate and temperature [Riveros-Iregui et al., 2007].<br />

Soil water content becomes the controll<strong>in</strong>g factor of CO2 production and concentrations when soil<br />

temperatures are ideal dur<strong>in</strong>g spr<strong>in</strong>g and summer [Buyanovsky and Wagner, 1983; Yuste et al.,<br />

2003] as high water availability enhances the biological activity and <strong>in</strong>hibits diffusive transport.<br />

The lack of water reduces microbial activity, the more prolonged a drought, the greater the<br />

adverse effect on size and functionality of microbial life with<strong>in</strong> the soil [Schimel et al., 1999].<br />

Both field [Liu et al., 2002; Tang et al., 2005] as well as laboratory [Orchard and Cook, 1983]<br />

experiments showed a rapid and strong response of CO2 production after the rewett<strong>in</strong>g of dry<br />

soils.<br />

Smaller, but still notable modulations of soil respiration are caused <strong>by</strong> soil type, nutrient availability<br />

and vegetation above ground [Buyanovsky and Wagner, 1983] to name a few.<br />

2.3.3 Molecular nitrogen and nitrogenous gases<br />

Molecular nitrogen (N2) is a major component of soil air, as it is also the most abundant gas <strong>in</strong><br />

atmospheric air. With<strong>in</strong> the soil regime, N2 is <strong>in</strong>fluenced <strong>by</strong> a complex system of processes both<br />

consum<strong>in</strong>g as well as produc<strong>in</strong>g N2 and nitrogenous gases. Under aerobic conditions nitrification<br />

takes place, produc<strong>in</strong>g nitric oxide NO and nitrous oxide N2O. Under anaerobic conditions, N2<br />

an N2O are produced <strong>by</strong> denitrification [Nieder and Benbi, 2008].<br />

While rarely measured <strong>in</strong> studies of soil atmosphere composition, Magnusson [1994] found N2<br />

concentrations to fluctuate between 78 – 95 % <strong>in</strong> water saturated soils, where CO2 is removed<br />

from the soil atmosphere <strong>by</strong> dissolution <strong>in</strong> water. Based on its high abundance and the existence<br />

of a subsurface source additional to the atmospheric reservoir it shares with the noble gases,<br />

N2 was speculated to be able to restore equilibrium conditions when O2 depletion and CO2<br />

dissolution lead to a partial pressure deficit <strong>in</strong> the soil atmosphere <strong>by</strong> Schneider [2010].<br />

32


2 Theory 2.3. Soil atmosphere composition<br />

2.3.4 Radon<br />

Even though it is a nobel gas, radon is irrelevant to noble gas paleotemperature studies as all<br />

of its isotopes are radioactive and therefore not conserved <strong>in</strong> ground water.<br />

222 Rn is produced with<strong>in</strong> the soil as a member of the radioactive decay cha<strong>in</strong> of 238 U, 220 Rn<br />

with<strong>in</strong> the 232 Th decay cha<strong>in</strong>. The source strength for radon isotopes is dependent on the<br />

uranium and thorium concentrations of the given soil and will therefore fluctuate depend<strong>in</strong>g on<br />

the soil constituent’s chemical composition and their spatial allocation.<br />

S<strong>in</strong>ce atmospheric radon concentrations are negligible [Porcelli et al., 2002], 222 Rn and 220 Rn<br />

are used as tracers to mark the soil atmosphere orig<strong>in</strong> of sampled air, as is be<strong>in</strong>g done <strong>in</strong> this<br />

study.<br />

33


Chapter 3<br />

Sampl<strong>in</strong>g sites and methods<br />

3.1 Setup of the sampl<strong>in</strong>g sites<br />

The sampl<strong>in</strong>g sites were chosen based on prelim<strong>in</strong>ary data <strong>by</strong> Schneider [2010] suggest<strong>in</strong>g high<br />

probabilities of oxygen depletion, there<strong>by</strong> be<strong>in</strong>g well suited to study whether oxygen depletion<br />

has any <strong>in</strong>fluence on noble gas partial pressures. Additionally, all sampl<strong>in</strong>g sites are <strong>in</strong> a reasonably<br />

close proximity to the Institute of Environmental <strong>Physics</strong>’ weather station as well as a<br />

weather station owned <strong>by</strong> the company meteomedia.<br />

3.1.1 Drill<strong>in</strong>g and <strong>in</strong>strumentation<br />

Drill<strong>in</strong>g the sample sites was done us<strong>in</strong>g a pneumatic pile hammer driven <strong>by</strong> a gas compressor.<br />

The pneumatic hammer was used to drive a hollow rod of 6 cm diameter and 1 m length <strong>in</strong>to<br />

the ground, allow<strong>in</strong>g for 1 m long, partly compressed core samples to be extracted, from which<br />

the soil layer structures were deduced. Us<strong>in</strong>g 1 m long solid extension rods the maximum depth<br />

reachable with the rented equipment was 9 m, however the maximum achieved depth was 5.9 m<br />

at Site A, 4.9 m at Site B and 2.9 m at Site C due to <strong>in</strong>creas<strong>in</strong>g friction at larger depths.<br />

Ground water tables were not reached at any of the sites. For simplification all soil air sampl<strong>in</strong>g<br />

depths are hereafter referred to <strong>by</strong> values rounded to the full meter.<br />

Sampl<strong>in</strong>g equipment, consist<strong>in</strong>g of three bundled TYGON R○ R-3603 Vacuum Tub<strong>in</strong>g tubes of<br />

different lengths and two LogTag TREX-8 temperature sensors, was lowered <strong>in</strong>to the boreholes.<br />

The tubes have an <strong>in</strong>ner diameter of 8 mm and were closed at their tips and perforated over a<br />

length of 10 cm <strong>by</strong> holes of 1 mm diameter (see Figure C.1). At 15 cm depth the tubes and sensor<br />

cables ran horizontally <strong>in</strong>to an underground box which was placed at a distance of about 20 cm<br />

to the hole, with a footpr<strong>in</strong>t of 0.03 m 2 , m<strong>in</strong>imally disturb<strong>in</strong>g the immediate site surround<strong>in</strong>gs.<br />

The boxes were covered <strong>by</strong> removable grass sods and conta<strong>in</strong>ed one LogTag TRIX-8 and two<br />

LogTag TREX-8 temperature data loggers as well as connectors and valves to access the different<br />

depth tubes. The sampl<strong>in</strong>g tubes were bundled <strong>in</strong>to a s<strong>in</strong>gle access tube, but rema<strong>in</strong>ed isolated<br />

from each other <strong>by</strong> valves, prevent<strong>in</strong>g contam<strong>in</strong>ation <strong>by</strong> backflow of atmospheric air.<br />

35


3.1. Setup of the sampl<strong>in</strong>g sites 3 Sampl<strong>in</strong>g sites and methods<br />

Figure 3.1: Equipment box at Site B before bury<strong>in</strong>g, with sampl<strong>in</strong>g tubes and temperature<br />

sensor cables runn<strong>in</strong>g to the open borehole at the left.<br />

The <strong>in</strong>strumented boreholes were refilled with the excavated material from the respective sites,<br />

mixed with water for better fluidity. The sample sites were drilled and refilled between 19 th and<br />

21 st of June, 2010. First samples were taken one month later to allow for the refilled material<br />

to settle down and solidify. Interconnectivity between the different depths through the refilled<br />

borehole was tested for <strong>by</strong> measur<strong>in</strong>g Radon profiles preced<strong>in</strong>g each noble gas sampl<strong>in</strong>g. See<br />

Section 5.1 for a detailed discussion of the results.<br />

3.1.2 Locations and soil properties<br />

Site A<br />

Sampl<strong>in</strong>g Site A (see Figure C.2) was located a few meters north of the Institute of Environmental<br />

<strong>Physics</strong>, at the coord<strong>in</strong>ates N49 ◦ 25.050 ′′ E8 ◦ 40.496 ′′ at 112 m above sea level, see Figure<br />

3.2. The borehole sits with<strong>in</strong> an area of about 50 m 2 of lawn. To the east and south the ground<br />

is sealed <strong>by</strong> paved roads, each at a distance of about 5 m to the site. Southeast, at a distance of<br />

1.5 m to the borehole, grows a s<strong>in</strong>gle tree.<br />

The soil at the site predom<strong>in</strong>antly consists of clay. The upper layers are most likely of an artificial<br />

orig<strong>in</strong>, as they were created or at least partly disturbed <strong>in</strong> 1998 when the Institute was<br />

built. The first 10 cm of turf are followed <strong>by</strong> 20 – 30 cm of clay, a 10 cm th<strong>in</strong> layer of a mixture<br />

36


3 Sampl<strong>in</strong>g sites and methods 3.1. Setup of the sampl<strong>in</strong>g sites<br />

SITE B<br />

SITE A<br />

SITE C<br />

Figure 3.2: Location of the sites created and sampled for this study [Source: Google Maps,<br />

map data provided <strong>by</strong> Tele Atlas].<br />

of sand and gravel and another layer of about 450 cm of clay (see Figure 3.3). Below a depth of<br />

5 m the ground primarily consists of sand and gravel.<br />

The site was equipped with three soil air sampl<strong>in</strong>g tubes, the <strong>in</strong>let screens located at depths 6 m,<br />

4 m and 2 m. Temperature sensors were <strong>in</strong>stalled at 2.5 m, 1.0 m and 0.1 m, the latter located<br />

<strong>in</strong>side the buried equipment box.<br />

Site B<br />

Sampl<strong>in</strong>g Site B (see Figures 3.4 and C.3) was located <strong>in</strong> an agricultural environment near<br />

Handschuhsheim, at the coord<strong>in</strong>ates N49 ◦ 25.775 ′′ E8 ◦ 39.760 ′′ at 109 m above sea level, see Figure<br />

3.2. The borehole was placed very close to the end of a hedge, partly covered <strong>by</strong> its leaves. A<br />

paved road of approximately 2 m width runs from north to south, at a distance of 1.5 m to the<br />

borehole. The immediate surround<strong>in</strong>g consists of a small meadow with apple trees, enclosed <strong>by</strong><br />

agriculturally used land.<br />

The soil at the site consists predom<strong>in</strong>antly of clay as well: The first 5 cm of turf are followed <strong>by</strong><br />

320 cm of clay, <strong>in</strong>creas<strong>in</strong>gly humid with depth. The follow<strong>in</strong>g layer consisted of 25 cm of humid,<br />

sandy clay. Below 3.5 m the clay displays a reddish gravel fraction.<br />

37


3.1. Setup of the sampl<strong>in</strong>g sites 3 Sampl<strong>in</strong>g sites and methods<br />

Figure 3.3: Detailed view of a core sample from a depth of 2 m at Site A, show<strong>in</strong>g the soil’s<br />

dense clay composition.<br />

The site is equipped with three soil air sampl<strong>in</strong>g tubes, at depths 5 m, 3 m and 2 m. Temperature<br />

sensors are <strong>in</strong>stalled at 2.5 m, 1.0 m and 0.1 m, the latter located <strong>in</strong>side the buried equipment<br />

box.<br />

Site C<br />

Sampl<strong>in</strong>g Site C is located with<strong>in</strong> an orchard on a hill slope above Dossenheim, at the coord<strong>in</strong>ates<br />

N49 ◦ 26.698 ′′ E8 ◦ 40.806 ′′ at 136 m above sea level, see Figure 3.2. The borehole was placed<br />

on ivy-covered but otherwise bare ground, sloped to the west.<br />

The soil at the site consists solely of clay down to the maximum depth of 3 m, a turf layer was<br />

not discernible at this site.<br />

The site is equipped with three soil air sampl<strong>in</strong>g tubes, at depths 3 m, 2 m and 1 m. No temperature<br />

sensors were <strong>in</strong>stalled.<br />

38


3 Sampl<strong>in</strong>g sites and methods 3.1. Setup of the sampl<strong>in</strong>g sites<br />

3.1.3 Development of the sites dur<strong>in</strong>g the sampl<strong>in</strong>g period<br />

Site A<br />

The radon measurements at Site A suggest that atmospheric seal<strong>in</strong>g of the borehole was ma<strong>in</strong>ta<strong>in</strong>ed<br />

dur<strong>in</strong>g the entire sampl<strong>in</strong>g period from late August 2010 to May 2011. However, cont<strong>in</strong>uous<br />

O2 and CO2 measurements <strong>in</strong> May 2011, exceed<strong>in</strong>g 10 m<strong>in</strong>utes, showed a change <strong>in</strong><br />

air composition after 200 seconds at 2 m depth. For a more detailed discussion of the borehole<br />

seal<strong>in</strong>g quality see Section 5.4.1.<br />

The extensive and unusual dry period dur<strong>in</strong>g spr<strong>in</strong>g 2011 (Table B.8 and Figure C.9) led to the<br />

decision to artificially irrigate Site A <strong>in</strong> May 2011. Between May 9 th and May 13 th , the site was<br />

irrigated four times. The amount of water was estimated to be similar to 30 mm of precipitation<br />

per irrigation event. It was artificially created <strong>by</strong> irrigat<strong>in</strong>g an area with a radius of 3 m around<br />

the borehole with tap water over one hour. A fifth irrigation of 30 m<strong>in</strong>utes length was done on<br />

May 18 th , lead<strong>in</strong>g to about 15 mm of artificial precipitation.<br />

The water was sprayed to emulate a natural ra<strong>in</strong> event as accurately as possible, the amount<br />

of 30 mm/h is comparable with heavy ra<strong>in</strong> [Baumgartner, 1996]. If evaporation dur<strong>in</strong>g spray<strong>in</strong>g<br />

and horizontal spread<strong>in</strong>g of the water with<strong>in</strong> the soil <strong>in</strong>to the surround<strong>in</strong>g dry soil is accounted<br />

for, this value likely represents a generous upper limit estimate.<br />

Dur<strong>in</strong>g the irrigation period O2 and CO2 measurements as well as noble gas samples were taken<br />

at an <strong>in</strong>creased rate (Sample IDs A17 to A23), document<strong>in</strong>g the effect of the <strong>in</strong>creased abundance<br />

of water with<strong>in</strong> the soil. Site A represents the largest and most detailed set of samples taken<br />

dur<strong>in</strong>g this study.<br />

Site B<br />

Site B was sampled regularly throughout the sampl<strong>in</strong>g period with at least one sample per<br />

month. However, the radon data suggests problems with the borehole seal<strong>in</strong>g <strong>in</strong> November<br />

2010 and from April 2011 onward, see Section 5.4.2 for a detailed discussion of seal<strong>in</strong>g quality<br />

of this particular site. While noble gas samples were taken dur<strong>in</strong>g the entire sampl<strong>in</strong>g time,<br />

measurement of most of the samples taken dur<strong>in</strong>g the time of verified atmospheric contam<strong>in</strong>ation<br />

had low priority due to limitations <strong>in</strong> measur<strong>in</strong>g time on the mass spectrometer. Therefore most<br />

of the noble gas samples taken <strong>in</strong> 2011 were not evaluated and rema<strong>in</strong> <strong>in</strong> storage at the time<br />

this <strong>thesis</strong> is f<strong>in</strong>ished.<br />

Site C<br />

The sampl<strong>in</strong>g Site C was abandoned after problems had arisen early on caused <strong>by</strong> heavy ra<strong>in</strong>falls<br />

the day before the first sampl<strong>in</strong>g attempt (August 2010, sample ID C01). The precipitation<br />

created a water saturated layer <strong>in</strong> the soil. While pump<strong>in</strong>g from the depths 1 m and 3 m led to<br />

normal airflow, pump<strong>in</strong>g at a depth of 2 m caused highly liquid sludge to be transported up the<br />

39


3.2. Sampl<strong>in</strong>g methods 3 Sampl<strong>in</strong>g sites and methods<br />

sampl<strong>in</strong>g tube. Sampl<strong>in</strong>g was <strong>in</strong>terrupted and could not be cont<strong>in</strong>ued at this depth until the<br />

end of September 2010 (Sample ID C04). Additionally, several extractions of Site C samples<br />

with<strong>in</strong> the mass spectrometer failed due to mechanical damage to the sample seal<strong>in</strong>g and led<br />

to an even more <strong>in</strong>complete record. Sporadic sampl<strong>in</strong>g dur<strong>in</strong>g w<strong>in</strong>ter and limited access to the<br />

mass spectrometer led to the site f<strong>in</strong>ally be<strong>in</strong>g omitted from sampl<strong>in</strong>g. Furthermore, an O2 and<br />

CO2 measurement <strong>in</strong> May 2011 proved this site to suffer from atmospheric air contam<strong>in</strong>ation <strong>in</strong><br />

all three depths as well.<br />

3.2 Sampl<strong>in</strong>g methods<br />

Gas samples were extracted from the sites <strong>by</strong> attach<strong>in</strong>g a Durridge RAD7 Radon Detector as<br />

a pump to the buried sampl<strong>in</strong>g tubes via the connector tube <strong>in</strong>side the equipment boxes. The<br />

perforated tips of the sampl<strong>in</strong>g tubes allow for soil air to be sampled from a screen of 10 cm height,<br />

however the actual ground layer be<strong>in</strong>g sampled was probably larger. The RAD7 -pump has a<br />

nom<strong>in</strong>al maximum pump rate of 1 l/m<strong>in</strong>, measurements of the pump rate resulted <strong>in</strong> an actual<br />

maximum pump rate of only 0.45 l/m<strong>in</strong> though [T. Reichel, personal note]. Pressures dur<strong>in</strong>g<br />

sample pump<strong>in</strong>g were seldom more than 10 mbar below local atmospheric pressure, <strong>in</strong>dicat<strong>in</strong>g<br />

low resistance allow<strong>in</strong>g the pump to reach its actual maximum pump rate. Each sampl<strong>in</strong>g depth<br />

was pumped for 30 m<strong>in</strong>utes before the actual noble gas sample was taken. The samples A17<br />

to A23 were taken us<strong>in</strong>g the Geotech BM2000 Biogas Monitor pump <strong>in</strong>stead of the RAD7 ’s,<br />

omitt<strong>in</strong>g accompany<strong>in</strong>g Radon measurements but lead<strong>in</strong>g to reduced pump times per sample of<br />

approximately 2 m<strong>in</strong> <strong>in</strong>stead of 30 m<strong>in</strong> 1 . The air with<strong>in</strong> the largest occurr<strong>in</strong>g dead space (the 6 m<br />

sampl<strong>in</strong>g tube at Site A) is moved away with<strong>in</strong> the first 1.5 m<strong>in</strong>utes of pump<strong>in</strong>g at a pump rate<br />

of 0.5 l/m<strong>in</strong>. As Figure C.6 shows, O2 and CO2 measurements reached stable read<strong>in</strong>g with<strong>in</strong> the<br />

first 60 seconds of pump<strong>in</strong>g with the BM2000. Scheffer and Schachtschabel [2010] gives a range<br />

of 65 – 35 % for pore space fraction of clay soils, assum<strong>in</strong>g a worst case of 35 % leads to an upper<br />

estimate of evacuated soil volume (at 0.5 l/m<strong>in</strong> pump rate and 30 m<strong>in</strong> of pump<strong>in</strong>g) of 0.043 m 3<br />

per sampl<strong>in</strong>g depth. Assum<strong>in</strong>g ideally homogenous soil structure lead<strong>in</strong>g to an isotropic, circular<br />

air <strong>in</strong>flow, this would correspond to an sampl<strong>in</strong>g radius of 22 cm.<br />

Therefore the depth accuracies for noble gas, O2 and CO2 measurements are given as ±0.2 m<br />

assum<strong>in</strong>g <strong>in</strong>tact seal<strong>in</strong>g, while the depth uncerta<strong>in</strong>ties for the temperature data should not<br />

exceed ±0.1 m.<br />

3.2.1 Noble gas samples<br />

The soil air was pumped through a 6x1 mm tube made of deoxidized copper (Wieland cuprofrio R○<br />

Cu-DHP), see Figure C.4. After 30 m<strong>in</strong>utes of pump<strong>in</strong>g, the pump-ward end of the copper tube<br />

was squeezed shut airtight [Wieser, 2006] us<strong>in</strong>g a pneumatic plier. Pressure was measured<br />

<strong>in</strong>l<strong>in</strong>e after clos<strong>in</strong>g the pump-ward end of the copper tube: once the connection to the pump<br />

1 Pump rates for the BM2000 were specified <strong>by</strong> the manufacturer as approximately 0.5 l/m<strong>in</strong>. Comparison of<br />

the achieved pump<strong>in</strong>g pressures of the RAD7 and the BM2000 dur<strong>in</strong>g sampl<strong>in</strong>g led to the conclusion that the<br />

BM2000 ’s pump rate is likely higher than specified and lies somewhere <strong>in</strong> the range of 0.5 to 1.0 l/m<strong>in</strong>.<br />

40


3 Sampl<strong>in</strong>g sites and methods 3.2. Sampl<strong>in</strong>g methods<br />

Figure 3.4: Typical sampl<strong>in</strong>g setup us<strong>in</strong>g the RAD7 as a pump, pictur<strong>in</strong>g Site B. (a)<br />

Location of the refilled borehole. (b) Equipment box with access valves to the different<br />

sampl<strong>in</strong>g screens and conta<strong>in</strong><strong>in</strong>g the LogTag dataloggers. (c) Inl<strong>in</strong>e pressure measurement<br />

<strong>in</strong>stalled between borehole and sample tube. (d) O2 and CO2 measur<strong>in</strong>g device <strong>in</strong>stalled<br />

between borehole and sample tube. (e) Copper sample tube and temperature sensor. (f)<br />

Drierite desiccant tube and RAD7 radon monitor.<br />

was separated the <strong>in</strong>l<strong>in</strong>e pressure rose from the slightly lower pump<strong>in</strong>g pressure to a value<br />

correspond<strong>in</strong>g to the current atmospheric pressure. Therefore all sampl<strong>in</strong>g tubes conta<strong>in</strong> gas at<br />

atmospheric pressure. From there the copper tube was divided <strong>in</strong>to three parts of roughly the<br />

same length, result<strong>in</strong>g <strong>in</strong> three samples per depth and sampl<strong>in</strong>g event. Sampl<strong>in</strong>g was always<br />

done from top to bottom, start<strong>in</strong>g with the shallowest available depth. The copper tubes were<br />

marked <strong>by</strong> carv<strong>in</strong>g <strong>in</strong> the IDs rather than mark<strong>in</strong>g them with tape or a pen, to ensure m<strong>in</strong>imal<br />

contam<strong>in</strong>ation when <strong>in</strong>stalled <strong>in</strong> the mass spectrometer. The sample ID consists of the site name<br />

(A, B or C), the two digit successive number of the sampl<strong>in</strong>g event and the depth given <strong>in</strong> roman<br />

numerals, ascend<strong>in</strong>g with depth (at Site A, I, II and III referred to 2 m, IV, V and VI to 4 m<br />

and VII, VIII and IX to 6 m). The ID A04-V therefore refers to a sample taken at Site B on<br />

the fourth sampl<strong>in</strong>g event of that site, at a depth of 4 m.<br />

The sample tube length was <strong>in</strong>creased from ∼6 cm to ∼12 cm (correspond<strong>in</strong>g to a volume of<br />

∼0.7 cc and ∼1.5 cc) after the first batch of samples was analyzed <strong>in</strong> the mass spectrometer<br />

show<strong>in</strong>g that the amount of 84 Kr <strong>in</strong> the small samples was technically impractical for the utilized<br />

mass spectrometer preparation l<strong>in</strong>e. Samples A01 – A08, B01 – B07 and C01 – C05 are of small<br />

size, all other samples are of the larger variant.<br />

41


3.2. Sampl<strong>in</strong>g methods 3 Sampl<strong>in</strong>g sites and methods<br />

Over 350 noble gas samples were taken dur<strong>in</strong>g the sampl<strong>in</strong>g period, 118 of which were successfully<br />

evaluated <strong>in</strong> the mass spectrometer. The rema<strong>in</strong><strong>in</strong>g samples either proved leaky or were not<br />

analyzed <strong>in</strong> the mass spectrometer due to time constra<strong>in</strong>ts.<br />

3.2.2 O2, CO2 and CH4<br />

O2, CO2 and CH4 were measured on site us<strong>in</strong>g a Geotech BM2000 Biogas Monitor with a<br />

built-<strong>in</strong> pump. Unfortunately the device was available only as of February 2011, leav<strong>in</strong>g noble<br />

gas samples A01 – A10, B01 – B09 and C01 – C05 without correspond<strong>in</strong>g O2, CO2 and CH4<br />

measurements.<br />

The BM2000 was connected to the sampl<strong>in</strong>g tube parallel to the RAD7, measurements were<br />

taken dur<strong>in</strong>g the first two m<strong>in</strong>utes as well as dur<strong>in</strong>g the last ten m<strong>in</strong>utes of the 30 m<strong>in</strong>ute<br />

pump<strong>in</strong>g period at each depth. S<strong>in</strong>ce variations between the two measurements never exceeded<br />

the nom<strong>in</strong>al accuracy of the sensors, the data from the last 10 m<strong>in</strong>utes of pump<strong>in</strong>g was used for<br />

its closer proximity to the noble gas sample.<br />

3.2.3 Radon<br />

A Durridge RAD7 Radon Detector was used as the primary sampl<strong>in</strong>g pump as well as for<br />

measur<strong>in</strong>g 222 Rn and 220 Rn concentrations, given <strong>in</strong> Bq/m 3 . The RAD7 was set up beh<strong>in</strong>d the<br />

noble gas sampl<strong>in</strong>g tubes and shielded from humidity <strong>by</strong> a Drierite desiccant tube. The RAD7<br />

was set for three 10 m<strong>in</strong>ute cycles, only 222 Rn values from the last cycle were used.<br />

3.2.4 Temperatures<br />

Atmospheric ambient temperatures at 2 m height above ground were provided <strong>by</strong> a weather<br />

station run <strong>by</strong> meteomedia close to Site A. The sampl<strong>in</strong>g sites were equipped each with LogTag<br />

TREX-8 and TRIX-8 data loggers (see Figure C.5), the latter featur<strong>in</strong>g build-<strong>in</strong> resistive temperature<br />

sensors, the former attached to external resistive temperature sensors. The loggers were<br />

set to record the temperature at an hourly <strong>in</strong>terval. Accuracy of the sensors between −10 ◦ C<br />

and +40 ◦ C is accord<strong>in</strong>g to the specifications (see Datasheet D.2) ±0.5 ◦ C. While the TRIX-8<br />

was placed <strong>in</strong>side the <strong>in</strong>strumentation box shielded from moisture (but also from fast temperature<br />

changes due to the <strong>in</strong>sulat<strong>in</strong>g air), none of the external sensors attached to the TREX-8 s<br />

were rated for deployment <strong>in</strong> moist to wet surround<strong>in</strong>gs. They were therefore <strong>in</strong>sulated us<strong>in</strong>g<br />

shr<strong>in</strong>k-sleeves with glue and tested <strong>in</strong> water beforehand. Test<strong>in</strong>g showed that moisture <strong>in</strong>trusion<br />

leads to temporary offsets <strong>in</strong> temperature read<strong>in</strong>gs rather than complete failure.<br />

42


Chapter 4<br />

Measur<strong>in</strong>g methods<br />

4.1 Mass spectrometry of He, Ne, Ar, Kr and Xe<br />

Measurements of He, Ne, Ar, Kr and Xe were conducted us<strong>in</strong>g a GV Instruments MM5400<br />

mass spectrometer and a preparation l<strong>in</strong>e built and set up <strong>by</strong> Friedrich [2007]. The samples<br />

were measured <strong>in</strong> four dist<strong>in</strong>ct runs, the first three dur<strong>in</strong>g September to November 2010 and the<br />

last one <strong>in</strong> May 2011. The first three runs utilized the same setup and sample size, while some<br />

changes were made for the fourth run, <strong>in</strong>clud<strong>in</strong>g a slightly different <strong>in</strong>let setup, different sample<br />

size and modified measur<strong>in</strong>g processes for He, Ar, Kr and Xe.<br />

4.1.1 Sample preparation and measur<strong>in</strong>g procedure<br />

The copper sample tubes were <strong>in</strong>stalled <strong>in</strong> an <strong>in</strong>let volume of known size which was evacuated<br />

to less than 5.0 × 10 −6 mbar. The process of <strong>in</strong>stall<strong>in</strong>g the sample tubes <strong>in</strong>side the <strong>in</strong>let was<br />

responsible for most sample losses s<strong>in</strong>ce mechanical stress had to be exerted on the fragile seal<strong>in</strong>g<br />

before the surround<strong>in</strong>g volume was evacuated. The sample was vented <strong>in</strong>to the evacuated <strong>in</strong>let<br />

volume <strong>by</strong> apply<strong>in</strong>g pressure perpendicular to one of the p<strong>in</strong>ch po<strong>in</strong>ts us<strong>in</strong>g a modified valve,<br />

there<strong>by</strong> crack<strong>in</strong>g the seal<strong>in</strong>g and releas<strong>in</strong>g the gas <strong>in</strong>to the <strong>in</strong>let volume almost <strong>in</strong>stantaneously.<br />

In rare cases the crack<strong>in</strong>g was so m<strong>in</strong>ute that the vent<strong>in</strong>g happened slowly over a timespan of 3<br />

to 10 m<strong>in</strong>utes.<br />

After tak<strong>in</strong>g a pressure read<strong>in</strong>g of the gas <strong>in</strong>side the <strong>in</strong>let volume, the gas was transferred to<br />

the preparation l<strong>in</strong>e where gas separation was done: The different noble gases cannot be measured<br />

side <strong>by</strong> side <strong>in</strong> a s<strong>in</strong>gle step but have to be separated. This is due to the differences<br />

<strong>in</strong> concentration magnitudes as well as to prevent <strong>in</strong>terference effects: The mass spectrometer<br />

discrim<strong>in</strong>ates mass-to-charge ratios, therefore a twice ionized 40 Ar atom will provide the same<br />

result as a s<strong>in</strong>gle-ionized 20 Ne atom. The gas separation was realized <strong>by</strong> an adsorption refrigeration<br />

system, adsorb<strong>in</strong>g the noble gases Ar, Kr and Xe, as well as all other permanent gases, at<br />

25 K <strong>in</strong> a permanent gas trap while He and Ne were adsorbed at temperatures below 12 K <strong>in</strong> an<br />

activated carbon reservoir. The gases were then released separately through <strong>in</strong>cremental heat<strong>in</strong>g<br />

43


4.1. Mass spectrometry of He, Ne, Ar, Kr and Xe 4 Measur<strong>in</strong>g methods<br />

to their respective desorption temperatures. Depend<strong>in</strong>g on the gas amounts present with<strong>in</strong> a<br />

sample, a splitt<strong>in</strong>g step is required and was <strong>in</strong>itiated if a specified limit was exceeded. This was<br />

determ<strong>in</strong>ed after separation us<strong>in</strong>g a quadrupole mass spectrometer to estimate the abundances<br />

of each noble gas.<br />

The gases were detected us<strong>in</strong>g a Faraday cup ( 4 He, 20 Ne, 22 Ne, 36 Ar, 40 Ar) and an electron<br />

multiplier ( 3 He, 84 Kr, 132 Xe). A specified volume of gas (atmospheric air of known noble gas<br />

composition) was measured at least once a day as a calibration standard and to characterize the<br />

long term stability of the measurement process, short term variations were characterized <strong>by</strong> fast<br />

calibrations executed before every sample measurement, us<strong>in</strong>g pure He, Ne and Kr-Xe gases.<br />

4.1.2 Determ<strong>in</strong>ation of sample gas amount<br />

The amount of gas <strong>in</strong>side a sample tube was calculated us<strong>in</strong>g two different methods, both based<br />

on the ideal gas law but rely<strong>in</strong>g on <strong>in</strong>dependently measured parameters. The first method<br />

utilizes the empirical formula A.6 created <strong>by</strong> Wieser [2006] that relates the sample tube length<br />

(at its closed state with squeezed ends) to its volume. The <strong>in</strong>l<strong>in</strong>e pressure dur<strong>in</strong>g sample shutoff<br />

and ambient temperature were measured dur<strong>in</strong>g sampl<strong>in</strong>g. This method isn’t very accurate<br />

though s<strong>in</strong>ce both volume calculation and temperature measurement are prone to large errors,<br />

result<strong>in</strong>g <strong>in</strong> relative errors <strong>in</strong> gas amount of 2.5 – 4.5 %. Details of the calculation can be found<br />

<strong>in</strong> equation A.5.<br />

The second method uses the gas pressure <strong>in</strong>side the <strong>in</strong>let volume p<strong>in</strong>let of the mass spectrometer,<br />

as well as the laboratory temperature TLab after expand<strong>in</strong>g the gas sample <strong>in</strong>to the evacuated<br />

<strong>in</strong>let volume. The <strong>in</strong>let pressure was measured us<strong>in</strong>g a Keller Lex 1 pressure gauge with a<br />

nom<strong>in</strong>al accuracy of 0.5 mbar (Datasheet D.3). The <strong>in</strong>let volume was calculated from volumetric<br />

measurements, <strong>by</strong> expand<strong>in</strong>g air at known pressure from a known volume of a different part of the<br />

mass spectrometer preparation l<strong>in</strong>e <strong>in</strong>to the <strong>in</strong>let volume. Details of the <strong>in</strong>let volume calculation<br />

can be found <strong>in</strong> Appendix A.2.<br />

S<strong>in</strong>ce the most important parameters of this method were measured with better precision and<br />

accuracy than those of the the other method, the accuracy of the calculated gas amount was<br />

superior. Average relative errors achieved <strong>by</strong> this gas amount measurement are at 1 %. Therefore<br />

this method, utiliz<strong>in</strong>g the <strong>in</strong>let pressure, was chosen over the other for all further calculations<br />

and data evaluations.<br />

The equation (detailed <strong>in</strong> Appendix A.1.2) derived from the ideal gas law <strong>in</strong>cludes a correction<br />

for the sample tube volume derived from its mass mCu and density ρCu as well as a correction<br />

for the water vapor pressure pH2O due to humidity with<strong>in</strong> the sample gas:<br />

�<br />

(p<strong>in</strong>let − pH2O) · VC −<br />

n =<br />

mCu<br />

�<br />

ρCu<br />

TLab · R<br />

The vapor pressure of water pH2O is part of the total pressure <strong>in</strong> the sample tube and therefore<br />

must be <strong>in</strong>cluded <strong>in</strong> the calculation of the gas sample size. Neither water vapor pressure nor<br />

the relative humidity were measured directly dur<strong>in</strong>g or after sampl<strong>in</strong>g. Instead it was assumed<br />

44<br />

(4.1)


4 Measur<strong>in</strong>g methods 4.1. Mass spectrometry of He, Ne, Ar, Kr and Xe<br />

that the relative humidity of soil air is always higher than 95 % <strong>in</strong> non-arid regions [Hillel, 1980].<br />

The vapor pressure of water was thus calculated from this assumption of relative humidity, as<br />

detailed <strong>in</strong> Appendix A.1.1.<br />

4.1.3 Estimation of relative humidity with<strong>in</strong> the <strong>in</strong>let section<br />

Expand<strong>in</strong>g the gas sample <strong>in</strong>to the <strong>in</strong>let volume of the mass spectrometer results <strong>in</strong> a change of<br />

its relative humidity φ due to the <strong>in</strong>crease <strong>in</strong> volume, φ<strong>in</strong>let �= φsample. Therefore the calculation<br />

of the water vapor pressure has to be based on this new relative humidity φ<strong>in</strong>let, which is<br />

calculated from the actual mass of the water mW,sample conta<strong>in</strong>ed <strong>in</strong> a sample tube us<strong>in</strong>g the<br />

absolute humidity ρW:<br />

ρW = mW,sample<br />

Vsample<br />

and ρW = φsample · ρW,max(T ) (4.2)<br />

⇒ mW,sample = φsample · ρW,max(T ) · Vsample (4.3)<br />

The maximum vapor pressure above water ρW,max(T ) at a given temperature T is calculated<br />

us<strong>in</strong>g the Magnus Formula [WMO, 2008]<br />

ρW,max(T ) = EW(t)<br />

RW · T<br />

EW(t) = 6.112 hPa · exp<br />

�<br />

17.62 · t<br />

243.12 ◦C + t<br />

�<br />

with the water vapor content EW(t), the specific gas constant of water RW = 461.52 J<br />

kg K and<br />

the temperature T [K], t [ ◦ C]. Assum<strong>in</strong>g φsample ≈ 100 % for soil air [Hillel, 1980] leads to<br />

mW,sample = EW(t)<br />

RW · T<br />

· Vsample<br />

(4.6)<br />

The sample volume is calculated us<strong>in</strong>g formula A.6, the sampl<strong>in</strong>g temperature however is of very<br />

low accuracy. Soil air temperatures were not measured <strong>in</strong> situ dur<strong>in</strong>g sampl<strong>in</strong>g at the respective<br />

sampl<strong>in</strong>g depths. In addition, the sample gas temperature was <strong>in</strong>fluenced dur<strong>in</strong>g sampl<strong>in</strong>g <strong>by</strong><br />

the process of pump<strong>in</strong>g the soil air from the ground and through a copper tube at ambient<br />

atmospheric temperature as well as <strong>by</strong> chang<strong>in</strong>g conditions (w<strong>in</strong>d, direct sunlight versus cloud<br />

cover) dur<strong>in</strong>g sampl<strong>in</strong>g. Therefore the relative error of sampl<strong>in</strong>g temperature measurements is<br />

estimated to be 5 %. The relative humidity with<strong>in</strong> the mass spectrometer <strong>in</strong>let volume is given<br />

<strong>by</strong><br />

φ<strong>in</strong>let =<br />

mW,sample<br />

V<strong>in</strong>let · ρW,max(TLab)<br />

where TLab is the average temperature of 24 ◦ C with<strong>in</strong> the laboratory, and V<strong>in</strong>let = (26.832 ±<br />

0.252) ml for runs 1 – 3 and V<strong>in</strong>let = (27.212 ± 0.153) ml for run 4, see Tables B.4 and B.5 for<br />

the result<strong>in</strong>g relative humidities.<br />

45<br />

(4.4)<br />

(4.5)<br />

(4.7)


4.2. On site measurement of O2, CO2, CH4 and Radon 4 Measur<strong>in</strong>g methods<br />

Mass Spectrometer<br />

noble gas<br />

[ peak heights ]<br />

Mass Spectrometer<br />

calibration gas<br />

[ peak heights ]<br />

Sample<br />

total gas amount<br />

[ mol ]<br />

4.1.4 Result<strong>in</strong>g data<br />

WuCEM<br />

data evaluation<br />

software<br />

Sample<br />

noble gas amount<br />

[ ccSTP ]<br />

Sample<br />

noble gas mix<strong>in</strong>g ratios<br />

[ Vol% ]<br />

Sample<br />

deviation from atm. air<br />

[ % atm. air ]<br />

Figure 4.1: Schematic diagram show<strong>in</strong>g how the raw data was processed.<br />

Volume of 1 mol<br />

of gas at STP<br />

[ ccSTP/mol ]<br />

Atmospheric air<br />

noble gas mix<strong>in</strong>g ratios<br />

[ Vol% ]<br />

The detected signals are converted to gas amounts <strong>in</strong> units ccSTP 1 <strong>by</strong> compar<strong>in</strong>g peak heights<br />

of the calibration gas measurements with those of the measured sample gas us<strong>in</strong>g a specialized<br />

evaluation software called WuCEM. The result<strong>in</strong>g gas amounts are therefore limited <strong>in</strong> accuracy<br />

<strong>by</strong> statistical fluctuations, variations of the spectrometer’s sensitivity and the accuracy of the<br />

calibration gas composition. Data characteriz<strong>in</strong>g the accuracy of the <strong>in</strong>dividual isotope read<strong>in</strong>gs<br />

result<strong>in</strong>g from these measurements can be found <strong>in</strong> Table B.7. Gas concentrations were calculated<br />

<strong>by</strong> divid<strong>in</strong>g the gas amounts <strong>by</strong> the molar volume at STP and <strong>by</strong> the measured total gas<br />

amount of the sample to deliver standardized and comparable gas concentrations.<br />

S<strong>in</strong>ce the <strong>in</strong>terest of this study is to show whether the soil air composition deviates from atmospheric<br />

air composition, the noble gas concentrations were divided <strong>by</strong> the mix<strong>in</strong>g ratio literature<br />

data of the respective noble gas isotopes [Porcelli et al., 2002], result<strong>in</strong>g <strong>in</strong> a relative <strong>in</strong>dication<br />

of deviation, expressed <strong>in</strong> percent relative to atmospheric air. A schematic representation of the<br />

process<strong>in</strong>g can be found <strong>in</strong> Figure 4.1.<br />

4.2 On site measurement of O2, CO2, CH4 and Radon<br />

4.2.1 Geotech BM2000 Biogas Monitor<br />

The Geotech BM2000 Biogas Monitor is a portable device capable of measur<strong>in</strong>g O2, CO2, CH4<br />

and CO, see Datasheet D.1. The O2 concentration is measured electrochemically, provid<strong>in</strong>g a<br />

measur<strong>in</strong>g range of 0 to 25 Vol% O2 with an error of ±1 Vol%. CO2 and CH4 are measured <strong>by</strong><br />

1 cm 3 gas at standard conditions for temperature and pressure, i.e. T = 273.15 K, p0 = 101.325 kPa.<br />

46


4 Measur<strong>in</strong>g methods 4.2. On site measurement of O2, CO2, CH4 and Radon<br />

<strong>in</strong>frared spectroscopy. The range of the CO2 and of the CH4 sensor is 0 – 100 Vol% with an<br />

accuracy of ±0.5 Vol% at 0 – 5, ±1.0 Vol% at 5 – 15 and ±3.0 Vol% above 15 Vol%. Calibration<br />

of the device was verified <strong>by</strong> us<strong>in</strong>g a certified test gas, confirm<strong>in</strong>g the provided accuracies but<br />

show<strong>in</strong>g a slight overestimation of O2, well with<strong>in</strong> the given error range though (see Table B.2).<br />

Atmospheric air was regularly sampled before and after soil air sampl<strong>in</strong>gs (see Figure C.7), the<br />

mean O2 value of (20.89±0.04) Vol% is <strong>in</strong> good agreement with the literature value of 20.9 Vol%<br />

[Porcelli et al., 2002].<br />

4.2.2 Durridge RAD7 Radon Detector<br />

The Durridge RAD7 Radon Detector detects α-particles of characteristic energies created with<strong>in</strong><br />

the decay cha<strong>in</strong>s of 222 Rn and 220 Rn, us<strong>in</strong>g a solid state detector. Solid state detectors consist of<br />

semiconductor material, a PN junction diode <strong>in</strong> reverse bias. An α-particle hitt<strong>in</strong>g the depletion<br />

layer creates electron-hole pairs that get separated <strong>by</strong> the electrical field spann<strong>in</strong>g the junction,<br />

lead<strong>in</strong>g to a measurable current. The energy resolution depends on the semiconductor’s bandgap,<br />

allow<strong>in</strong>g to differentiate between the different decay<strong>in</strong>g Radon isotopes: The RAD7 detects<br />

the 6.00 MeV α-decay of 218 Po, which itself is the product of 222 Rn decay, and the 6.78 MeV<br />

α-decay of 216 Po, belong<strong>in</strong>g to the 220 Rn decay cha<strong>in</strong>. This allows for the parallel monitor<strong>in</strong>g<br />

of 222 Rn and 220 Rn, lead<strong>in</strong>g to a more detailed source of <strong>in</strong>formation: while the read<strong>in</strong>gs of the<br />

long-lived 222 Rn (t 1/2 = 3.82 days) will usually <strong>in</strong>crease <strong>in</strong> every one of the three 10 m<strong>in</strong>ute<br />

cycles due to its long lifetime, read<strong>in</strong>gs of the short-lived 220 Rn (t 1/2 = 55.6 sec) should be<br />

constant <strong>in</strong> all three cycles. A change <strong>in</strong> 220 Rn read<strong>in</strong>gs <strong>in</strong> one of the three cycles therefore<br />

<strong>in</strong>dicates an <strong>in</strong>flow of differently composed air caus<strong>in</strong>g dilution or enrichment depend<strong>in</strong>g on its<br />

220 Rn content.<br />

47


Chapter 5<br />

Results<br />

5.1 Radon<br />

222 Rn profiles were generally similar for Site A and Site B, the shallowest depths show<strong>in</strong>g<br />

the least amount of Radon, the medium depths the highest and the deepest depths a medium<br />

concentration. This decrease of 222 Rn concentration is likely connected to the soil profile. Both<br />

sampl<strong>in</strong>g sites showed a less dense soil structure at their respective deepest sampl<strong>in</strong>g depths<br />

with <strong>in</strong>creas<strong>in</strong>g gravel and sand fractions. This may result <strong>in</strong> a better aeration compared to the<br />

dense clay soil at medium depths, likely lead<strong>in</strong>g to the lower 222 Rn concentrations. In addition,<br />

the radon source strength also differs due to variations <strong>in</strong> soil composition.<br />

The 222 Rn measurements taken on Site A (Figure 5.4) show little variation. A closer look<br />

reveals a slight <strong>in</strong>crease of the 222 Rn concentration at 2 m depth <strong>by</strong> 10 %, start<strong>in</strong>g <strong>in</strong> January<br />

2011. The elevated 222 Rn concentrations co<strong>in</strong>cide with an <strong>in</strong>crease <strong>in</strong> the concentration of the<br />

short-lived 220 Rn from the first to the second measur<strong>in</strong>g cycle at each sampl<strong>in</strong>g of 2 m depth.<br />

However, these 220 Rn <strong>in</strong>creases are with<strong>in</strong> the marg<strong>in</strong> of error of the measurements.<br />

The 222 Rn data at Site B shown <strong>in</strong> Figure 5.5, clearly display a steep dip <strong>in</strong> December 2010.<br />

The concentrations recovered dur<strong>in</strong>g January and February 2011. As of March 2011 the 222 Rn<br />

concentrations recessed back to a decreased level.<br />

5.2 Temperature profiles<br />

The data loggers on Site B delivered a nearly complete record cover<strong>in</strong>g the entire sampl<strong>in</strong>g<br />

period, as shown <strong>in</strong> Figure 5.1. At Site A the record (see Figure C.8) is not as complete,<br />

show<strong>in</strong>g gaps caused <strong>by</strong> logger failures as well as offsets caused <strong>by</strong> moisture <strong>in</strong>trusions <strong>in</strong>to the<br />

sensors. While the logged temperatures at both sites <strong>in</strong> general deviate very little from each<br />

other, the diurnal variations at 0.1 m depth at Site B are more pronounced <strong>in</strong> August to October<br />

2010. This is likely caused <strong>by</strong> the differ<strong>in</strong>g vegetation cover at the two sites, lead<strong>in</strong>g to different<br />

durations of direct sunlight exposure.<br />

49


5.2. Temperature profiles 5 Results<br />

� � � � � � � � � � � ��� � �<br />

Figure 5.1: Temperature measurements taken at Site B at 0.1 m, 1.0 m and 2.5 m depth<br />

and s<strong>in</strong>usoidal fits. Depth placement accuracy is ±0.1 m, temperature accuracy is ±0.5 ◦ C.<br />

The atmospheric daily mean temperatures were provided <strong>by</strong> meteomedia, measured <strong>by</strong> a<br />

weather station close to Site A.<br />

The damp<strong>in</strong>g of the atmospheric temperature signal is clearly visible and <strong>in</strong>creas<strong>in</strong>g <strong>in</strong> magnitude<br />

with depth. The <strong>in</strong>sulat<strong>in</strong>g effect of snow cover is apparent dur<strong>in</strong>g December and January,<br />

lead<strong>in</strong>g to nearly constant temperatures above zero at 0.1 m depth while atmospheric temperature<br />

above ground was well below 0 ◦ C. Neither sampl<strong>in</strong>g site experienced ground frost at 0.1 m<br />

depth and lower dur<strong>in</strong>g the W<strong>in</strong>ter 2010/2011.<br />

Comparison of the m<strong>in</strong>ima of the s<strong>in</strong>usoidal fits of the temperature curves (exclud<strong>in</strong>g known<br />

erroneous read<strong>in</strong>gs) show the retardation of temperature penetration <strong>in</strong>to the soil. At 0.1 m<br />

the retardation is about ten days, at 1.0 m it is one month and two months at 2.5 m. The<br />

atmospheric average temperature based on the fit (to account for gaps <strong>in</strong> the datasets) of the<br />

meteomedia data from August 2010 to May 2011 was (12.5 ± 0.5) ◦ C. At Site A, the average<br />

temperatures <strong>in</strong> the same time frame were: (10.8 ± 0.5) ◦ C at 0.1 m, (11.7 ± 0.5) ◦ C at 1.0 m and<br />

(13.1 ± 0.5) ◦ C at 2.5 m. At Site B they were: (11.0 ± 0.5) ◦ C at 0.1 m, (12.2 ± 0.5) ◦ C at 1.0 m<br />

and (12.4 ± 0.5) ◦ C at 2.5 m. Accuracy is based on the sensor’s accuracy.<br />

Dur<strong>in</strong>g the irrigation of Site A the temperature measurements at 1.0 m and 2.5 m show spikes<br />

co<strong>in</strong>cid<strong>in</strong>g with the water <strong>in</strong>trusion 1 (see Figure 6.1). Whether these spikes represent accurate<br />

1 The lack of reaction to the first irrigation is due to the loggers record<strong>in</strong>g at hourly <strong>in</strong>tervals at that time,<br />

50


5 Results 5.3. O2 and CO2 profiles<br />

temperature changes is questionable, as the temperature record shows several spikes unrelated to<br />

active irrigation and rem<strong>in</strong>iscent of previously observed sensor failures due to moisture <strong>in</strong>trusion.<br />

These spikes do <strong>in</strong>dicate though that the water <strong>in</strong>filtrated the soil to a depth of 2.5 m with<strong>in</strong> an<br />

hour, and therefore likely reached the lower depths as well with<strong>in</strong> a short time.<br />

5.3 O2 and CO2 profiles<br />

O2, CO2 and CH4 were measured at Site A and Site B start<strong>in</strong>g <strong>in</strong> February 2011. S<strong>in</strong>ce<br />

the seal<strong>in</strong>g of Site B is questionable (see Section 5.4) only the measured gas concentrations at<br />

Site A are discussed here <strong>in</strong> detail, the entire dataset is <strong>in</strong>cluded <strong>in</strong> Table B.11. Furthermore,<br />

CH4 was never measured <strong>in</strong> any significant amount: the rare and unsystematic occurrences of<br />

maximum read<strong>in</strong>gs of 0.1 Vol% can be ignored with respect to the sensor’s accuracy of 0.5 Vol%.<br />

Dur<strong>in</strong>g February to April 2011, O2 and CO2 were measured only when noble gas samples were<br />

taken to m<strong>in</strong>imize disturbance of the site’s soil. Dur<strong>in</strong>g that time, the depth profiles (see<br />

Figure 5.2, A and B) of both O2 and CO2 showed uniform concentrations, <strong>in</strong>dicat<strong>in</strong>g low CO2<br />

production. CO2 concentrations at all depths fell dur<strong>in</strong>g that time from an <strong>in</strong>itial maximum of<br />

(4.7 ± 0.5) Vol% to (3.4 ± 0.5) Vol% while O2 concentration rema<strong>in</strong>ed constant at all depths<br />

(with<strong>in</strong> the measurement’s accuracy) at (18 ± 1) Vol% (see Figure C.13). While O2 levels<br />

were depleted compared to atmospheric concentration at that time, the deficit was more than<br />

compensated <strong>by</strong> the rise <strong>in</strong> CO2 concentrations.<br />

Irrigation of Site A was started on May 9 th , 2011 as a reaction to the prolonged dry period,<br />

accompanied <strong>by</strong> an <strong>in</strong>crease of the frequency of O2 and CO2 measurements (see Figure 5.3).<br />

Measurements were executed directly before each irrigation and hourly afterwards. The response<br />

of the O2 and CO2 concentrations was immediate at all three depths. The most pronounced<br />

reaction occurred at the shallowest depth of 2 m, while at 6 m only little change was observed<br />

(see Figure 5.2, C, D and E).<br />

The response of the CO2 concentrations is characterized <strong>by</strong> two phases: the <strong>in</strong>itial reaction is<br />

a steep drop <strong>in</strong> CO2 concentrations at all depths, the maximum at 2 m was a drop <strong>by</strong> 1.7 Vol%<br />

while at 6 m the largest drop was 0.2 Vol%. Dur<strong>in</strong>g the follow<strong>in</strong>g hours the CO2 levels recovered<br />

to the preced<strong>in</strong>g levels at 6 m, at 2 and 4 m depth they rose above them. The ga<strong>in</strong> <strong>in</strong> CO2 was<br />

most highest at 2 m, where the CO2 concentration peaked at (6.8 ± 1.0) Vol% about two weeks<br />

after the first irrigation (see Figure 5.2, E). After the last irrigation was done on May 19 th ,<br />

the CO2 concentration at 2 and 4 m started to decrease slowly, but steadily. At 6 m, the CO2<br />

concentrations never peaked but rose slowly and steadily <strong>by</strong> about 1.0 Vol% dur<strong>in</strong>g the first 50<br />

days after the <strong>in</strong>itial irrigation (see Figures 5.3 and 5.2, F).<br />

O2 concentrations at 2 m rose slightly dur<strong>in</strong>g the first few hours after each irrigation, the overly<strong>in</strong>g<br />

trend however was a depletion of O2 dur<strong>in</strong>g the first two weeks, reach<strong>in</strong>g a m<strong>in</strong>imum<br />

concentration of (10.9 ± 1.0) Vol% ten days after the first irrigation (see Figure 5.2, D). Once<br />

irrigation stopped, the levels rose aga<strong>in</strong>. At 4 m the <strong>in</strong>fluence of irrigation was much less pronounced<br />

but visible, lead<strong>in</strong>g to a m<strong>in</strong>imum <strong>in</strong> O2 concentrations at (13.2 ± 1.0) Vol% 15 days<br />

logg<strong>in</strong>g frequency was subsequently <strong>in</strong>creased to 5 m<strong>in</strong>utes to be able to resolve the water <strong>in</strong>cursion.<br />

51


5.3. O2 and CO2 profiles 5 Results<br />

Figure 5.2: O2 and CO2 depth profiles taken at Site A at various stages. Values at 0 m<br />

depth are based on literature data for atmospheric air [Porcelli et al., 2002]. A-B: before<br />

irrigation. C: dur<strong>in</strong>g irrigation, show<strong>in</strong>g the m<strong>in</strong>imal CO2 concentration measured. D: 10<br />

days after irrigation, m<strong>in</strong>imum of the sum of O2 and CO2. E-F: several weeks after irrigation.<br />

52


5 Results 5.3. O2 and CO2 profiles<br />

Figure 5.3: O2 and CO2 concentrations <strong>in</strong> all three depths of Site A, as well as the sum of<br />

O2 and CO2 at 2 m depth, start<strong>in</strong>g <strong>in</strong> May 2011. Artificially <strong>in</strong>troduced precipitation and<br />

ra<strong>in</strong>fall, data is limited to the <strong>in</strong>terval May 9 th – May 18 th .<br />

after the first irrigation. No effect of the irrigation was visible at 6 m, however over the entire<br />

time s<strong>in</strong>ce <strong>in</strong>itial irrigation the concentration slowly fell <strong>by</strong> about 1.5 Vol%.<br />

The sum of O2 and CO2 concentrations was most affected <strong>by</strong> the reaction to the irrigation at<br />

2 m depth, lead<strong>in</strong>g to a m<strong>in</strong>imum of 16.5 Vol%, caus<strong>in</strong>g a deficit compared to atmospheric composition<br />

of 4.4 Vol%. However, such a deficit was never observed dur<strong>in</strong>g conditions un<strong>in</strong>fluenced<br />

<strong>by</strong> the irrigation. In fact the O2+CO2 concentration at 2 and 4 m was usually a bit above<br />

20.9 Vol%, though well with<strong>in</strong> the sum’s accuracy of 1.4 Vol%.<br />

Figure C.13 shows an overview of the entire O2 and CO2 data available at Site A while Figure<br />

C.14 provides a detailed view of the first ten days of irrigation.<br />

The measurements of O2 and CO2 at Site B show little variation over time (Figure C.15), the<br />

CO2 and O2 concentrations at 1 and 3 m depth are nearly identical. They fluctuate slightly<br />

around 2 Vol% dur<strong>in</strong>g the entire measurement period and are approximately 2 Vol% lower than<br />

Site A’s CO2 concentrations at 2 m depth dur<strong>in</strong>g the same time. At 5 m depth, the CO2<br />

concentrations are <strong>in</strong>creased and the O2 concentrations decreased compared to 1 and 3 m.<br />

53


5.4. Borehole seal<strong>in</strong>g 5 Results<br />

5.4 Borehole seal<strong>in</strong>g<br />

Seal<strong>in</strong>g of the sampl<strong>in</strong>g screens from atmospheric contam<strong>in</strong>ation as well as from each other is an<br />

important factor for the reliability of the depth <strong>in</strong>formation of the measured gas concentrations.<br />

While the estimated depth accuracy lies with<strong>in</strong> the range of ±0.2 m, assum<strong>in</strong>g homogeneous<br />

soil structure, the actual depth accuracy is highly dependent on vertical gas permeability. The<br />

weakest l<strong>in</strong>k of the sampl<strong>in</strong>g site setup is the refill<strong>in</strong>g of the boreholes, possibly lead<strong>in</strong>g to large<br />

pore space and channels along the buried sampl<strong>in</strong>g tubes potentially allow<strong>in</strong>g direct air exchange<br />

between the different sampl<strong>in</strong>g screens or even exchange with atmospheric air.<br />

Radon, O2 and CO2 concentrations were used as tracers for atmospheric air contam<strong>in</strong>ation as<br />

well as <strong>in</strong>terconnectivity between the different screen<strong>in</strong>g depths. Comb<strong>in</strong>ation of all three tracers<br />

allows for discrim<strong>in</strong>ation between the two k<strong>in</strong>ds of seal<strong>in</strong>g breaches.<br />

5.4.1 Site A<br />

The small variations <strong>in</strong> 222 Rn measurements taken on Site A (see Figure 5.4) imply a fully<br />

conta<strong>in</strong>ed seal<strong>in</strong>g aga<strong>in</strong>st atmospheric contam<strong>in</strong>ation spann<strong>in</strong>g the entire sampl<strong>in</strong>g period. O2<br />

and CO2 measurements, taken dur<strong>in</strong>g the first 2 m<strong>in</strong>utes as well as dur<strong>in</strong>g the last 10 m<strong>in</strong>utes of<br />

each 30 m<strong>in</strong>ute radon sampl<strong>in</strong>g showed no significant variation dur<strong>in</strong>g each sampl<strong>in</strong>g, support<strong>in</strong>g<br />

the case of <strong>in</strong>tact seal<strong>in</strong>g as well. A first <strong>in</strong>dication of seal<strong>in</strong>g failure was observed only after<br />

the last noble gas sample (A23) was taken: A longer than usual O2 and CO2 measurement on<br />

May 21 st showed an <strong>in</strong>crease of O2 and a decrease of CO2 after around 3 m<strong>in</strong>utes of pump<strong>in</strong>g,<br />

displayed <strong>in</strong> Figure 5.6. A 222 Rn measurement taken the day before did not show any signs of<br />

dilution with 222 Rn-free 2 atmospheric air. However, the analysis of the entire 222 Rn dataset of<br />

Site A led to the discovery of the slight upwards shift <strong>in</strong> 222 Rn concentrations at the 2 m depth<br />

start<strong>in</strong>g <strong>in</strong> January 2011. This <strong>in</strong>crease likely represents an <strong>in</strong>terconnectivity of the 2 m and 4 m<br />

screen<strong>in</strong>g depths. This would not have been visible <strong>in</strong> the O2 and CO2 data before May 2011,<br />

s<strong>in</strong>ce until then the concentrations of O2 and CO2 were of very similar magnitude <strong>in</strong> all three<br />

depths, there<strong>by</strong> mask<strong>in</strong>g mix<strong>in</strong>g effects. The assumption of an <strong>in</strong>terconnectivity is supported <strong>by</strong><br />

an <strong>in</strong>crease <strong>in</strong> the concentration of the short-lived 220 Rn from the first to the second measur<strong>in</strong>g<br />

cycle at each sampl<strong>in</strong>g of the 2 m depth, <strong>in</strong>dicat<strong>in</strong>g the <strong>in</strong>flow of soil air from a lower depth<br />

rather than atmospheric air.<br />

Based on these observations, the samples taken until January 2011 (A01 to A09) conta<strong>in</strong> air<br />

from the respective sampl<strong>in</strong>g depths, while samples taken afterwards until April 2011 (A10 to<br />

A16) at depths 2 m and 4 m likely conta<strong>in</strong> mixed soil air. The samples collected <strong>in</strong> May 2011<br />

(A17 to A23) were taken us<strong>in</strong>g the BM2000 <strong>in</strong>stead of the RAD7, with less than 2 m<strong>in</strong>utes of<br />

pump<strong>in</strong>g time. Based on the measurement shown <strong>in</strong> Figure 5.6, these samples conta<strong>in</strong> unmixed<br />

air from the respective sampl<strong>in</strong>g depths.<br />

2 The atmospheric volume fraction of radon is about 6 × 10 −20 [Porcelli et al., 2002].<br />

54


5 Results 5.4. Borehole seal<strong>in</strong>g<br />

�� �� ������������ �� � � � �� ���� ��� �� �� � �� ��� � � �� � � � � ��� �� � � � � � �<br />

�� �� ������������������������� � �� �� �� �� ��� � � � � �<br />

�� �� ������������ �� � � � �� ���� ��� �� �� � �� ��� � � �� � � � � � � �� �� � �<br />

� �<br />

�� �� ��� �� ������������������� � �� �� �� �� ��� � � � �<br />

Figure 5.4: Radon concentrations dur<strong>in</strong>g sampl<strong>in</strong>g at Site A. The 2 m values show a slight<br />

upwards shift of about 10 % start<strong>in</strong>g as of January 2011.<br />

Figure 5.5: Radon concentrations dur<strong>in</strong>g sampl<strong>in</strong>g at Site B. Higher than usual 222 Rn<br />

values for atmospheric air are due to immediate previous use of the RAD7 with ground air<br />

on a different sampl<strong>in</strong>g site and are unlikely to be actual atmospheric fluctuations.<br />

55


5.4. Borehole seal<strong>in</strong>g 5 Results<br />

Figure 5.6: Cont<strong>in</strong>uous O2 and CO2 measurements at Site A, show<strong>in</strong>g an <strong>in</strong>crease of O2 and<br />

a decrease of CO2 start<strong>in</strong>g at around 200 s of pump<strong>in</strong>g, <strong>in</strong>dicat<strong>in</strong>g an <strong>in</strong>flow of air more closely<br />

resembl<strong>in</strong>g atmospheric composition. This air <strong>in</strong>flow is not necessarily atmospheric though.<br />

Based on the associated 222 Rn and 220 Rn data it is more likely that an <strong>in</strong>terconnection<br />

between the screens at 2 and 4 m exists. S<strong>in</strong>ce the 4 m soil air conta<strong>in</strong>ed more O2 and<br />

less CO2 than the 2 m soil air at the time, the mix<strong>in</strong>g only appears to be atmospherically<br />

<strong>in</strong>fluenced. Accuracy for both O2 and CO2 is 1 %, error bars have been omitted for better<br />

visibility.<br />

Figure 5.7: Cont<strong>in</strong>uous O2 and CO2 measurements at Site B, show<strong>in</strong>g an <strong>in</strong>crease of O2<br />

and a decrease of CO2 start<strong>in</strong>g at around 30 s of pump<strong>in</strong>g, <strong>in</strong>dicat<strong>in</strong>g an <strong>in</strong>flow of air more<br />

closely resembl<strong>in</strong>g atmospheric composition. Accuracies are 1 % for O2 and 0.5 % for CO2,<br />

error bars have been omitted for better visibility.<br />

56


5 Results 5.5. Noble gases<br />

5.4.2 Site B<br />

The 222 Rn data <strong>in</strong> Figure 5.5 clearly <strong>in</strong>dicates an atmospheric leak lead<strong>in</strong>g to a dilution of the<br />

ground air with mostly 222 Rn-free atmospheric air on several, if not all, sampl<strong>in</strong>gs. O2 and CO2<br />

measurements shown <strong>in</strong> Figure 5.7 support the atmospheric leak scenario, show<strong>in</strong>g an <strong>in</strong>trusion<br />

of differently composed air with<strong>in</strong> 30 seconds of pump<strong>in</strong>g at all depths except 5 m. A contrast<strong>in</strong>g<br />

plot for Site A show<strong>in</strong>g no atmospheric air <strong>in</strong>flow can be found <strong>in</strong> Figure C.6. Additionally,<br />

CO2 concentrations measured at 1 and 3 m at Site B are significantly lower than at 2 m at Site<br />

A while O2 concentrations are higher.<br />

While the reseal<strong>in</strong>g of the borehole at Site B <strong>in</strong>itially seemed to have been successful 3 , the<br />

relatively dry period of September to October 2010 probably caused the deterioration of the<br />

seal<strong>in</strong>g <strong>by</strong> <strong>in</strong>duc<strong>in</strong>g the formation of contraction cracks [Scheffer and Schachtschabel, 2010] <strong>in</strong><br />

the dry<strong>in</strong>g clay soil. The recovery of the 222 Rn values dur<strong>in</strong>g w<strong>in</strong>ter was likely caused <strong>by</strong> a<br />

partial and temporal reseal<strong>in</strong>g of the upper soil layer due to the <strong>in</strong>crease <strong>in</strong> precipitation.<br />

This leads to the conclusion that most, if not all, samples taken at Site B do not conta<strong>in</strong> soil<br />

air from the respective sampl<strong>in</strong>g depths, but rather a mixture of soil and atmospheric air and<br />

therefore cannot be used as a record to answer this <strong>thesis</strong>’ ma<strong>in</strong> objective.<br />

5.5 Noble gases<br />

The accuracy achieved <strong>by</strong> the mass spectrometer was best for the isotopes measured with the<br />

faraday cup and significantly worse for those measured with the multiplier. The best relative<br />

accuracy was achieved for 20 Ne (see Table B.7). Based on these accuracy patterns, the five<br />

isotopes 4 He, 20 Ne, 22 Ne, 36 Ar and 40 Ar were chosen to be presented <strong>in</strong> detail as their average<br />

measurement error was below 1.5 %. The f<strong>in</strong>al accuracy of the presented data is <strong>in</strong>fluenced <strong>by</strong><br />

the <strong>in</strong>troduction of measured gas amount of each sample. The rema<strong>in</strong><strong>in</strong>g three isotopes 3 He,<br />

84 Kr and 132 Xe mirrored the general trends found <strong>in</strong> this study, but suffered from low accuracies<br />

and more pronounced scatter<strong>in</strong>g. Plots and data for all measured isotopes can be found <strong>in</strong> the<br />

Appendices B and C.<br />

5.5.1 Effects of storage<br />

Immediate measurement of the sampled air <strong>in</strong> the mass spectrometer was usually not possible<br />

as the spectrometer was only available dur<strong>in</strong>g certa<strong>in</strong> time frames. Because the copper tubes<br />

conta<strong>in</strong> air with high O2 concentrations and humidity, the possibility of chemical reactions<br />

occurr<strong>in</strong>g dur<strong>in</strong>g storage affect<strong>in</strong>g the noble gases was tested. This was done three times <strong>by</strong><br />

measur<strong>in</strong>g all three samples of a s<strong>in</strong>gle depth, one only a few hours after collect<strong>in</strong>g the samples.<br />

The rema<strong>in</strong><strong>in</strong>g two backups were measured after stor<strong>in</strong>g them for different periods, the one for<br />

3 Assum<strong>in</strong>g the difference between absolute 222 Rn concentrations of Site A and Site B were caused <strong>by</strong> differ<strong>in</strong>g<br />

source strengths. Otherwise it might even be the case that only the 222 Rn concentrations measured <strong>in</strong> February<br />

2011 at Site B represent a sealed borehole.<br />

57


5.5. Noble gases 5 Results<br />

8 days, the other for more than 6 months. All three samples of B06 at 1 m depth (I – III) and<br />

of B06 at 3 m depth (IV – VI) were successfully measured, of sample A23 at 2 m depth one (II)<br />

of the three samples proved to be leaky and was lost.<br />

The results show no change, scatter<strong>in</strong>g with<strong>in</strong> the accuracy of the measurements (see Figure<br />

C.16). This <strong>in</strong>dicates that noble gas measurements were unaffected <strong>by</strong> potential chemical reactions<br />

with<strong>in</strong> the sample tubes dur<strong>in</strong>g prolonged storage.<br />

5.5.2 Atmospheric air samples<br />

The atmospheric air samples display a -2 % shift relative to literature isotope mix<strong>in</strong>g ratios of<br />

atmospheric air at STP [Porcelli et al., 2002], as shown <strong>in</strong> Figure 5.8. Only the heavy noble<br />

gas isotope 84 Kr does not display this behavior, as shown <strong>in</strong> Figure C.11. An isotope mass<br />

dependency of the effect is not visible however, as 132 Xe does display the shift<strong>in</strong>g as well.<br />

Possible mechanisms lead<strong>in</strong>g to such a shift are connected to the noble gas concentrations of<br />

the calibration gas 4 and the gas amount of each <strong>in</strong>dividual sample, calculated from various<br />

parameters: <strong>in</strong>let pressure p<strong>in</strong>let, <strong>in</strong>let volume V<strong>in</strong>let, relative humidity φ<strong>in</strong>let, sample tube mass m<br />

and laboratory temperature TLab. While the latter three parameters would require unreasonably<br />

high changes (> 30 %) to account for a 2 % change <strong>in</strong> gas amount n or have no effect at all <strong>in</strong> an<br />

order of magnitude of percent, the gas amount is quite sensitive on changes of V<strong>in</strong>let and p<strong>in</strong>let<br />

and will be discussed below.<br />

Errors <strong>in</strong> the calculation of noble gas amounts <strong>in</strong> the calibration gas are fairly improbable, but<br />

possible. Measurements of atmospheric air equilibrated water samples us<strong>in</strong>g the same calibration<br />

gas were <strong>in</strong>conspicuous however [T. Kaudse, personal note] and led to the conclusion that the<br />

calibration gas concentrations were accurate.<br />

Inaccuracy of the calculation of the <strong>in</strong>let volume is another parameter that could account for the<br />

-2 % shift, however the volume would have to be 0.54 ml smaller than measured. The accuracy<br />

of the volumetric measurement is 0.15 ml or 0.6 %, based on the accuracies of the known volumes<br />

used. While it is unlikely that an <strong>in</strong>accurately determ<strong>in</strong>ed <strong>in</strong>let volume is responsible for the<br />

entire shift <strong>in</strong> the noble gas concentrations, part of the effect may be attributed to it.<br />

To account for the entire -2 % shift, the <strong>in</strong>let pressure would have to have been overestimated <strong>by</strong><br />

2 %. The Keller pressure gauge regularly displayed a peak read<strong>in</strong>g right after open<strong>in</strong>g a sample<br />

that decreased slightly dur<strong>in</strong>g the next few seconds <strong>by</strong> 0.1 to 0.3 mbar, from an average total of<br />

about 60 mbar. Read<strong>in</strong>gs were taken consistently dur<strong>in</strong>g the first 15 to 20 seconds after open<strong>in</strong>g<br />

the sample, except for samples that barely cracked and consequently required up to several<br />

m<strong>in</strong>utes to vent <strong>in</strong>to the <strong>in</strong>let volume. The slow vent<strong>in</strong>g of these samples (A20-IV, A21-VIII<br />

and A23-III) caused the Keller gauge to approach the f<strong>in</strong>al read<strong>in</strong>g slowly from below, rather<br />

than overshoot<strong>in</strong>g it like the <strong>in</strong>stantly vent<strong>in</strong>g samples did. A comparison of the slowly vented<br />

samples with <strong>in</strong>stantaneously vented samples allows for a qualitative estimation of the effect:<br />

Sample A23-III is a backup sample of the previously measured A23-I (which vented <strong>in</strong>stantaneously<br />

on crack<strong>in</strong>g) from the same date and depth. While A23-III is slightly larger <strong>in</strong> sample<br />

4 The noble gas content of the calibration gas was calculated from measurements of temperature, pressure and<br />

humidity at the time that the calibration gas was sampled from atmospheric air.<br />

58


5 Results 5.5. Noble gases<br />

3<br />

He [% atm. air]<br />

4<br />

He [% atm. air]<br />

40<br />

Ar [% atm. air]<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

110<br />

105<br />

100<br />

95<br />

90<br />

105<br />

100<br />

95<br />

90<br />

105<br />

100<br />

95<br />

90<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

Date<br />

20<br />

Ne [% atm. air]<br />

22<br />

Ne [% atm. air]<br />

36<br />

Ar [% atm. air]<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

110<br />

105<br />

100<br />

95<br />

90<br />

105<br />

100<br />

95<br />

90<br />

105<br />

100<br />

95<br />

90<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

Figure 5.8: Noble gas concentrations of atmospheric air samples relative to literature values<br />

from Porcelli et al. [2002] for 3 He, 4 He, 20 Ne, 22 Ne, 36 Ar and 40 Ar. All isotopes show an<br />

underestimation of roughly 2 %.<br />

tube volume than A23-I, the measured <strong>in</strong>let pressure is actually 0.6 mbar lower, <strong>in</strong>dicat<strong>in</strong>g a<br />

pressure overestimation on sample A23-I. A comparison of the noble gas concentrations calculated<br />

for these two samples shows that A23-I displays a shift relative to A23-III <strong>by</strong> roughly -2 %<br />

on most noble gas isotopes. Samples A20-IV and A21-VIII show elevated noble gas concentrations<br />

compared with samples from the same depth and roughly the same time as well, however<br />

compar<strong>in</strong>g samples from different sampl<strong>in</strong>g dates can only deliver a vague <strong>in</strong>dication at best. To<br />

quantify the observed behavior of slowly decreas<strong>in</strong>g pressure read<strong>in</strong>gs the Keller gauge and the<br />

<strong>in</strong>let volume were re<strong>in</strong>stalled <strong>in</strong> the mass spectrometer preparation l<strong>in</strong>e and eight atmospheric<br />

air samples were opened. This experiment was unable to replicate the previously observed effect:<br />

the read<strong>in</strong>gs for those <strong>in</strong>stantaneously vented samples did not decrease. However this led to the<br />

observation that the Keller gauge is highly sensitive to its spatial orientation, show<strong>in</strong>g a 1 mbar<br />

difference when rotated <strong>by</strong> 180 ◦ , possibly add<strong>in</strong>g to the observed shift <strong>in</strong> atmospheric noble gas<br />

values.<br />

It was deemed most likely that the read<strong>in</strong>g of the <strong>in</strong>let pressure was executed prematurely, but<br />

consistently, lead<strong>in</strong>g to a systematical overestimation of sample gas amount and there<strong>by</strong> caus<strong>in</strong>g<br />

the offset displayed <strong>by</strong> the atmospheric air samples. Therefore all measured samples’ <strong>in</strong>dividual<br />

noble gas isotope values discussed below (except those that vented slowly) were standardized to<br />

the mean fitted values of the measured atmospheric air.<br />

59<br />

Date


5.5. Noble gases 5 Results<br />

5.5.3 Soil atmosphere samples from Site A<br />

A total of 23 noble gas depth profiles were sampled at Site A and measured, the entire dataset<br />

can be found <strong>in</strong> Table B.9. A few read<strong>in</strong>gs (marked red <strong>in</strong> Table B.9) were omitted from<br />

analysis as they scatter unreasonably strong. Except for the 36 Ar read<strong>in</strong>g of sample A01-I, all<br />

these outliers were connected to technical problems that occurred dur<strong>in</strong>g the measurement <strong>in</strong> the<br />

mass spectrometer. The 36 Ar read<strong>in</strong>g of sample A01-I is the only isotope of that sample show<strong>in</strong>g<br />

such elevated read<strong>in</strong>gs and is therefore regarded as an artifact of an undiscovered problem dur<strong>in</strong>g<br />

that s<strong>in</strong>gle measurement.<br />

The sampl<strong>in</strong>g frequency was <strong>in</strong>creased dur<strong>in</strong>g the irrigation period. Samples were taken right<br />

before irrigation (A21), with<strong>in</strong> one hour after irrigation (A22) and days after irrigation events.<br />

The noble gas sample that is accompanied <strong>by</strong> the lowest sum of O2 and CO2 (16.7 Vol%) is A19<br />

2 m depth.<br />

Figure 5.9 shows the measured concentrations of the four isotopes 20 Ne, 22 Ne, 36 Ar and 40 Ar<br />

standardized to literature data on atmospheric composition, plotted over time. The plots for the<br />

rema<strong>in</strong><strong>in</strong>g isotopes can be found <strong>in</strong> Figure C.17. It is quite clear that most of the noble gas data<br />

is located above or at 100 % relative to atmospheric air composition. A significant drop below<br />

100 %, consistent <strong>in</strong> all isotopes, was never observed. Only the heavy noble gas isotopes 84 Kr<br />

and 132 Xe showed such reduced noble gas concentrations, though suffer<strong>in</strong>g from low accuracy<br />

and strong scatter<strong>in</strong>g their reliability is questionable.<br />

The data can be separated <strong>in</strong>to two different time frames based on the differ<strong>in</strong>g visible trends.<br />

The first <strong>in</strong>terval is August 2010 – April 2011, when all noble gases show a decl<strong>in</strong>e from <strong>in</strong>itial<br />

values well above 100 % at the beg<strong>in</strong>n<strong>in</strong>g to values scatter<strong>in</strong>g closely around 100 % with<strong>in</strong> the<br />

data’s accuracy. This trend is visible at all three depths, however it is most pronounced at<br />

6 m, where the <strong>in</strong>itial values are between 104 – 106 %, depend<strong>in</strong>g on the isotope. The shallower<br />

depths both start at 102 – 105 % and reach 100 % earlier than at 6 m. Between December 2010<br />

and April 2011, the noble gas concentrations scattered around 100 % at all depths.<br />

The second <strong>in</strong>terval is May – June 2011, mark<strong>in</strong>g the artificial irrigation of Site A. The reaction<br />

of the noble gases at a depth of 2 m is immediate and strong, concentrations jump from 100 %<br />

to 104 – 107 % (depend<strong>in</strong>g on the noble gas isotope) at the last measured noble gas sample<br />

(A23), taken eleven days after the first irrigation (see Figure 6.1). At depths 4 and 6 m noble<br />

gas concentrations show no reaction at all, rema<strong>in</strong><strong>in</strong>g at around 100 %.<br />

The noble gas isotope ratios 3 He/ 4 He, 4 He/ 20 Ne, 20 Ne/ 22 Ne and 40 Ar/ 36 Ar were calculated,<br />

however a specific evaluation of the mass spectrometer data focused on these ratios was not<br />

done. The ratios displayed no irregularities or significant deviations from literature values (see<br />

Table B.9). The change of O2 and CO2 concentrations had no effect on these ratios, as shown<br />

exemplarily <strong>in</strong> Figures C.20 and C.21 for 4 He/ 20 Ne and 20 Ne/ 22 Ne.<br />

60


5 Results 5.5. Noble gases<br />

20<br />

Ne (2m) [% atm. air]<br />

20<br />

Ne (4m) [% atm. air]<br />

20<br />

Ne (6m) [% atm. air]<br />

36<br />

Ar (2m) [% atm. air]<br />

36<br />

Ar (4m) [% atm. air]<br />

36<br />

Ar (6m) [% atm. air]<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

Date<br />

22<br />

Ne (2m) [% atm. air]<br />

22<br />

Ne (4m) [% atm. air]<br />

22<br />

Ne (6m) [% atm. air]<br />

40<br />

Ar (2m) [% atm. air]<br />

40<br />

Ar (4m) [% atm. air]<br />

40<br />

Ar (6m) [% atm. air]<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

Figure 5.9: Concentrations of 20 Ne, 22 Ne, 36 Ar and 40 Ar <strong>in</strong> soil atmosphere samples from<br />

the different depths of Site A over time, relative to atmospheric literature values from<br />

Porcelli et al. [2002].<br />

61<br />

Date


5.5. Noble gases 5 Results<br />

5.5.4 Soil atmosphere samples from Site B<br />

Noble gas measurements at Site B show a similar, but less dist<strong>in</strong>ct trend of a decl<strong>in</strong>e of concentrations<br />

from elevated values to 100 % <strong>in</strong> w<strong>in</strong>ter. The decl<strong>in</strong>e is similar at all sampled depths<br />

(1, 3 and 5 m), <strong>in</strong> contrast to observations at Site A. S<strong>in</strong>ce it is very likely that the sampled<br />

soil atmosphere at Site B was contam<strong>in</strong>ated to an unknown extent with atmospheric air, the<br />

quality of the data is questionable and the dataset has been omitted from discussion. Plots can<br />

be found <strong>in</strong> Figures C.18 and C.19, the complete dataset is listed <strong>in</strong> Table B.10.<br />

62


Chapter 6<br />

Discussion<br />

The O2 and CO2 measurements executed at Site A show the <strong>in</strong>fluence of several different factors<br />

on soil respiration, all conform<strong>in</strong>g with theoretical predictions and the results of other studies<br />

as presented <strong>in</strong> Chapter 2, Section 2.3.<br />

Dur<strong>in</strong>g late w<strong>in</strong>ter and early spr<strong>in</strong>g 2011, the O2 and CO2 profiles were uniform from 2 to 6 m,<br />

show<strong>in</strong>g a moderate depletion of O2, the sum of O2 and CO2 never fell below 20 Vol% though.<br />

The dom<strong>in</strong>at<strong>in</strong>g factor on soil respiration dur<strong>in</strong>g early spr<strong>in</strong>g was temperature, lead<strong>in</strong>g to very<br />

low CO2 production rates (see Chapter 2, Figure 2.10). The slight decl<strong>in</strong>e <strong>in</strong> CO2 and <strong>in</strong>crease<br />

<strong>in</strong> O2 concentrations <strong>in</strong> the shallower depths, as visible towards May 2011, is probably caused<br />

<strong>by</strong> decreas<strong>in</strong>g water content due to the dry period <strong>in</strong> spr<strong>in</strong>g. This would <strong>in</strong>crease the diffusivity<br />

of the soil, lead<strong>in</strong>g to stronger diffusive exchange with the atmosphere.<br />

While soil temperatures started to rise dur<strong>in</strong>g March 2011, O2 and CO2 measurements showed<br />

no <strong>in</strong>crease of soil respiration. In light of the ris<strong>in</strong>g temperatures and the low amounts of precipitation<br />

dur<strong>in</strong>g March and April (see Figure C.10), soil moisture content likely became the<br />

dom<strong>in</strong>ant limit<strong>in</strong>g factor on biological activity and there<strong>by</strong> on soil respiration. The <strong>in</strong>troduction<br />

of water <strong>in</strong>to the system caused a dramatical change, as shown <strong>in</strong> Figure 5.3. The immediate<br />

reaction to this event was a steep drop <strong>in</strong> CO2 concentrations, caused <strong>by</strong> parts of the CO2 dissolv<strong>in</strong>g.<br />

The abundance of water then activated microbial activity lead<strong>in</strong>g to quickly <strong>in</strong>creas<strong>in</strong>g<br />

CO2 concentrations, as previously observed <strong>in</strong> other studies [Liu et al., 2002; Tang et al., 2005].<br />

The measured maximum of soil respiration was located at 2 m depth, actual activity was likely<br />

even higher at shallower depths [Fierer et al., 2003]. The smaller <strong>in</strong>itial drop <strong>in</strong> CO2 concentrations<br />

<strong>in</strong> response to the water at depths 4 and 6 m was presumably caused <strong>by</strong> smaller amounts of<br />

water reach<strong>in</strong>g those depths. Furthermore, due to the spatial distribution of microbial biomass,<br />

the subsequent changes <strong>in</strong> O2 and CO2 concentrations at 4 m are likely <strong>in</strong>fluenced <strong>in</strong> part <strong>by</strong><br />

diffusive transport from shallower depths. The slow but constant rise of CO2 concentration at<br />

6 m from <strong>in</strong>itial 3.4 Vol% to 4.0 Vol% was probably caused entirely <strong>by</strong> diffusive transfer rather<br />

than <strong>by</strong> <strong>in</strong>-situ production<br />

The drop <strong>in</strong> O2 concentration at 2 m was quite significant and was amplified <strong>by</strong> the high water<br />

content of the upper soil layers slow<strong>in</strong>g down diffusive O2 replenishment from the atmosphere.<br />

63


6 Discussion<br />

This rather extreme amount of water <strong>in</strong>trud<strong>in</strong>g with<strong>in</strong> a short time period created a situation<br />

where oxygen depletion led to a gas deficit of up to 4.4 Vol% at the shallow depth of 2 m.<br />

Cont<strong>in</strong>ued measurements dur<strong>in</strong>g the follow<strong>in</strong>g two months <strong>in</strong>dicate that this is <strong>in</strong> fact unusual:<br />

O2 and CO2 read<strong>in</strong>gs at both 2 and 4 m depth returned towards their <strong>in</strong>itial values despite the<br />

onset of regular natural precipitation (see Figure 5.3). Ra<strong>in</strong>fall had only limited and short-lived<br />

effect on the concentrations at 2 m. The sum of O2 and CO2 rema<strong>in</strong>ed below 20.9 Vol% only<br />

at 6 m depth dur<strong>in</strong>g the two months after irrigation, trend<strong>in</strong>g slightly towards further decrease.<br />

The difference at the time this <strong>thesis</strong> was written was below 1 Vol% however.<br />

O2 and CO2 depth profiles as measured <strong>by</strong> Schneider [2010] <strong>in</strong> August 2009 show that the sum<br />

of O2 and CO2 descends well below 20.9 Vol% under certa<strong>in</strong> natural conditions. The maximum<br />

of O2 depletion and consequent gas deficit was located at greater depths, a m<strong>in</strong>imum of the sum<br />

of O2 and CO2 was measured <strong>by</strong> Schneider [2010] at 3.5 and 4.5 m depth with 15 Vol%. While<br />

such O2 and CO2 profiles were not encountered dur<strong>in</strong>g February to June 2011, it is very likely<br />

that, depend<strong>in</strong>g on variations of weather, a similar situation will be encountered <strong>in</strong> late summer<br />

or autumn 2011.<br />

Despite be<strong>in</strong>g a slightly unusual <strong>in</strong>trusion <strong>in</strong>to the system, the irrigation did provide valuable<br />

<strong>in</strong>formation on the behavior of the noble gas concentrations. As Figure 6.1 shows, noble gas concentrations<br />

did <strong>in</strong>deed rise significantly after the sum of O2 and CO2 dropped below 20.9 Vol%.<br />

To test if the <strong>in</strong>trusion of water itself played a factor <strong>in</strong> this, samples A21 and A22 were taken<br />

with<strong>in</strong> an hour before and after an irrigation event. As visible <strong>in</strong> Figure 6.1, the difference<br />

between the two samples is well with<strong>in</strong> their accuracy and with<strong>in</strong> the normal scatter<strong>in</strong>g of the<br />

mass spectrometer measurement when compared with the double measurement of sample A23.<br />

Therefore a direct <strong>in</strong>fluence of the water <strong>in</strong>trusion is improbable.<br />

The correlation between noble gas concentrations and O2+CO2 deficit is shown <strong>in</strong> Figure 6.2 (<br />

the entire dataset is plotted <strong>in</strong> Figures C.22 and C.23), where the relative change of noble gas<br />

concentrations compared to atmospheric air concentrations is plotted aga<strong>in</strong>st the sum of O2 and<br />

CO2. Also <strong>in</strong>cluded is the theoretical prediction derived from equation 2.19. The accuracies<br />

of the O2 and CO2 measurements are relatively low. It is apparent nonetheless that noble gas<br />

concentrations roughly follow the predictions of equation 2.19 when an O2+CO2 deficit occurs.<br />

This holds also true when O2+CO2 <strong>in</strong>creases slightly above 20.9 Vol%, which was often the case<br />

at depths 2 and 4 m outside the irrigation period.<br />

Furthermore, l<strong>in</strong>ear error weighted fitt<strong>in</strong>g of the dataset of Site A’s 2 m depth samples (see<br />

Figure 6.2) <strong>in</strong>dicates that there might be a mass discrim<strong>in</strong>ation <strong>in</strong>volved: the lighter noble gas<br />

isotopes appear to be less affected <strong>by</strong> changes <strong>in</strong> O2+CO2 concentrations as expected, while the<br />

heavier isotopes conform more closely to the theoretical expectation. Gravitational separation as<br />

described <strong>by</strong> Stute and Schlosser [1993] can not be responsible for this observation at this study’s<br />

shallow depths. A plausible explanation of this observation might be the high diffusion coefficient<br />

of helium <strong>in</strong> air [Cussler, 2005]. The rise <strong>in</strong> noble gas partial pressures creates a concentration<br />

gradient between enriched soil air and atmospheric air. Diffusive transport reduces this gradient,<br />

the rate of change is dependent on the diffusion coefficient as described <strong>by</strong> Fick’s second law<br />

(Equation 2.24). Therefore, the observed low 4 He enrichment could be caused <strong>by</strong> much more<br />

rapid reduction of the concentration gradient of 4 He than of those of the other noble gases.<br />

64


6 Discussion<br />

� � � � ��� � ���� �� � � ��� �� �<br />

� � � � � � ��� � ���� � �� �<br />

� � � � �� �� � � �� � ��� � �<br />

� � � � � � � � � � � ��� � �<br />

Figure 6.1: 40 Ar data from a depth of 2 m (top), sum of O2 and CO2 concentrations at 2 m<br />

(center), soil temperature at 1.0 and 2.5 m and precipitation per day before break, per hour<br />

after break (bottom), all from Site A over the period of January to March 2011, and <strong>in</strong><br />

May 2011. Scales on the date axis differ before and after the break.<br />

65


)!<br />

)!<br />

*+,-./-01234*-56-20301.7*38*9-82/1:-9*17*;./


6 Discussion<br />

F 20 Ne<br />

F 84 Kr<br />

1.04<br />

1.02<br />

1.00<br />

0.98<br />

0.96<br />

1.10<br />

1.05<br />

1.00<br />

0.95<br />

Data from 2 m depth, A01 - A10<br />

Data from 6 m depth, A01 - A10<br />

L<strong>in</strong>ear Fit of 2 m data<br />

L<strong>in</strong>ear Fit of 6 m data<br />

0.96 0.98 1.00 1.02 1.04<br />

F 4 He<br />

Data from 2 m depth, A01 -- A10<br />

Data from 6 m depth, A01 -- A10<br />

L<strong>in</strong>ear Fit of 2 m data<br />

L<strong>in</strong>ear Fit of 6 m data<br />

0.90<br />

0.90 0.95 1.00 1.05 1.10<br />

F 4 He<br />

F 20 Ne<br />

1.04<br />

1.02<br />

1.00<br />

0.98<br />

Data from 2 m depth, A11 - A23<br />

Data from 6 m depth, A11 - A23<br />

L<strong>in</strong>ear Fit of 2 m data<br />

L<strong>in</strong>ear Fit of 6 m data<br />

(a) 0.96<br />

(b)<br />

F 84 Kr<br />

1.10<br />

1.05<br />

1.00<br />

0.95<br />

0.96 0.98 1.00 1.02 1.04<br />

F 4 He<br />

Data from 2 m depth, A11 - A23<br />

Data from 6 m depth, A11 - A23<br />

L<strong>in</strong>ear Fit of 2 m data<br />

L<strong>in</strong>ear Fit of 6 m data<br />

(c) (d)<br />

0.90<br />

0.90 0.95 1.00 1.05 1.10<br />

Figure 6.3: F-Values (see Section A.3 for def<strong>in</strong>ition) of 20 Ne (a, b) and 84 Kr (c, d) versus<br />

F-Values of 4 He. Data <strong>in</strong> the left row (a, c) is from August 2010 – January 2011 (lack<strong>in</strong>g<br />

O2 and CO2 read<strong>in</strong>gs). Data <strong>in</strong> the right row (b, c) is from February – May 2011.<br />

Unfortunately the dataset of noble gas measurements with correspond<strong>in</strong>g O2+CO2 measurements<br />

is quite small. While noble gas concentration changes were <strong>in</strong> fact measured at 6 m<br />

depth <strong>in</strong> late summer 2010, the lack of accompany<strong>in</strong>g O2 and CO2 data renders these noble gas<br />

records useless for this specific analysis. Therefore a different method was employed to test if<br />

the relatively lower enrichment of 4 He can also be found <strong>in</strong> the samples A01 – A10 (that lack<br />

the O2+CO2 data): The F-Values (see Section A.3 for def<strong>in</strong>ition) of the isotopes 3 He, 20 Ne,<br />

22 Ne, 36 Ar, 84 Kr and 132 Xe were plotted versus the F-Value of 4 He (see Figure C.24). Figure<br />

6.3 displays a selective analysis where the data was separated <strong>by</strong> depth and <strong>by</strong> measur<strong>in</strong>g date.<br />

Figure 6.3 (b) shows that 4 He <strong>in</strong> samples from 2 m was slightly depleted relative to the standard<br />

( 40 Ar) while 20 Ne was not, as expected based on Figure 6.2. S<strong>in</strong>ce the highest noble gas<br />

enrichment dur<strong>in</strong>g August 2010 – January 2011 was found at 6 m <strong>in</strong>stead of 2 ˙m, Figure 6.3 (a)<br />

should show a reversed behavior. This is apparently not the case, suggest<strong>in</strong>g that 4 He might<br />

have been <strong>in</strong>fluenced differently at 6 m. However, the accuracy of the data appears to be too<br />

low to draw conclusions, as Figures 6.3 (c) and (d) show. Further improvement of the accuracy<br />

as well as more extensive data is called for to test whether these trends can be confirmed and<br />

whether they appear at all depths.<br />

67<br />

F 4 He


6 Discussion<br />

Table 6.1: Synthetic data of noble gas concentrations <strong>in</strong> water at a temperature of 12 ◦ C.<br />

A is the fraction of air entrapped <strong>in</strong> the water volume, POD is the change <strong>in</strong> noble gas<br />

partial pressures as def<strong>in</strong>ed <strong>in</strong> equation 2.19, Ci are the noble gas concentrations. Ti are<br />

the temperatures calculated <strong>by</strong> the respective model used to fit the synthetic data.<br />

Input<br />

ID A POD CHe CNe CAr CKr CXe<br />

EQ0 0.000 0.00 4.61×10 −8 1.98×10 −7 3.69×10 −4 8.60×10 −8 1.23×10 −8<br />

UA0 0.002 1.00 5.65×10 −8 2.34×10 −7 3.88×10 −4 8.83×10 −8 1.25×10 −8<br />

OD3 0.002 1.03 5.79×10 −8 2.40×10 −7 3.99×10 −4 9.09×10 −8 1.28×10 −8<br />

OD6 0.002 1.06 5.93×10 −8 2.46×10 −7 4.10×10 −4 9.35×10 −8 1.32×10 −8<br />

Output<br />

ID TCE ∆TCE TOD ∆TOD<br />

UA0 12.0 0.4 12.0 2.1<br />

OD3 11.2 0.4 12.0 2.1<br />

OD6 10.5 0.4 12.0 2.1<br />

The O2+CO2 deficits observed <strong>by</strong> this study are of the same order of magnitude as those<br />

recorded <strong>by</strong> Schneider [2010], yet the f<strong>in</strong>d<strong>in</strong>gs on noble gas enrichment significantly differ. A<br />

reason for this discrepancy could not be identified as both study’s setups, sampl<strong>in</strong>g procedures<br />

and measurement methods appear reasonably reliable.<br />

The high noble gas concentrations at 6 m depth <strong>in</strong> summer 2010, reach<strong>in</strong>g up to (106.0 ± 1.7) %<br />

relative to atmospheric air (and slightly lower at 4 m depth, see Figure 5.9), prove that the<br />

OD model’s basic assumption of noble gas enrichment with<strong>in</strong> the soil atmosphere, proposed <strong>by</strong><br />

Hall et al. [2005], is <strong>in</strong>deed encountered with<strong>in</strong> natural soil environments. Based on theoretical<br />

descriptions of soil CO2 depth profiles and the one-time depth profile from August 2009 <strong>by</strong><br />

Schneider [2010], it is reasonable to conclude that this measured <strong>in</strong>crease <strong>in</strong> noble gases is<br />

caused <strong>by</strong> the O2+CO2 deficit occurr<strong>in</strong>g dur<strong>in</strong>g summer and autumn. S<strong>in</strong>ce this study could<br />

not complete a full annual dataset at the time this <strong>thesis</strong> was written, giv<strong>in</strong>g an annual average<br />

of noble gas enrichment is not possible at this time. In light of the observed fluctuations of the<br />

soil atmosphere composition and the seasonal patterns of precipitation lead<strong>in</strong>g to ground water<br />

recharge, it is questionable whether such an average should be used at all.<br />

Estimat<strong>in</strong>g the maximum <strong>in</strong>fluence of noble gas enrichment on noble gas temperatures is however<br />

reasonable and possible <strong>by</strong> us<strong>in</strong>g the maximum measured noble gas concentrations of one of the<br />

most accurately measured isotopes, 40 Ar. Its concentration reached, both <strong>in</strong> summer 2009 at<br />

6 m as well as dur<strong>in</strong>g irrigation at 2 m, 106 % relative to atmospheric air concentrations. A small<br />

set of synthetic data of noble gas concentrations <strong>in</strong> water was created, based on equilibrium<br />

concentrations Ci,eq <strong>in</strong> water at 12 ◦ C and atmospheric pressure (EQ0). Three different variations<br />

of this set were calculated: the first assumes atmospheric noble gas concentrations (UA0), the<br />

second assumes an <strong>in</strong>crease <strong>by</strong> 3 % (OD3) and the third an <strong>in</strong>crease <strong>by</strong> 6 % (OD6). The <strong>in</strong>itial<br />

datasets and the result<strong>in</strong>g temperatures are detailed <strong>in</strong> Table 6.1. These datasets were fitted<br />

68


6 Discussion<br />

us<strong>in</strong>g a software solution developed <strong>by</strong> von Oehsen [2008], both the CE model and the OD model<br />

were employed.<br />

The OD model does reconstruct the <strong>in</strong>itial temperature of 12 ◦ C well for all datasets but suffers<br />

from low accuracy. While the CE model successfully reproduces the temperature of dataset UA0<br />

at a much better accuracy compared to the OD model, it underestimates the temperatures of<br />

datasets OD3 and OD6 <strong>by</strong> 0.8 ◦ C and 1.5 ◦ C respectively.<br />

The observed elevated noble gas concentrations at this study’s sampl<strong>in</strong>g site were therefore<br />

able to <strong>in</strong>fluence the temperatures reproduced <strong>by</strong> the CE model significantly. However, they<br />

were not large enough to account entirely for deviations of noble gas temperatures from mean<br />

annual air temperatures as observed <strong>by</strong> Ma et al. [2004] and Hall et al. [2005] <strong>in</strong> studies of<br />

recent ground waters that prompted the proposition of the OD model to better describe the<br />

excess air component of dissolved noble gases. Yet it is not unlikely that stronger O2+CO2<br />

depletion occurs naturally at <strong>in</strong>creased depths closer to the capillary fr<strong>in</strong>ge. Additionally, the<br />

magnitude of <strong>in</strong>fluence on noble gas temperatures is also dependent on local annual temperature<br />

and precipitation patterns, determ<strong>in</strong><strong>in</strong>g whether ground water recharge occurs at the same time<br />

as the maximum of oxygen depletion: If both temperature and precipitation variations were<br />

<strong>in</strong> phase (or the temperature basically constant at a level ideal for microbial activity), soil<br />

respiration would presumably be at a maximum (disregard<strong>in</strong>g other parameters) parallel to<br />

ground water recharge, lead<strong>in</strong>g to a strong <strong>in</strong>crease <strong>in</strong> noble gas concentrations. The CE model<br />

would then underestimate the noble gas temperatures. The other extreme would be a situation<br />

where temperature and precipitation were out of phase, lead<strong>in</strong>g to m<strong>in</strong>imal soil respiration dur<strong>in</strong>g<br />

the time of ground water recharge, limit<strong>in</strong>g the <strong>in</strong>fluence of oxygen depletion on ground water<br />

noble gas concentrations. In that case the CE model would likely deliver accurate results while<br />

the OD model, if based on maximum or mean O2+CO2 deficits, would overestimate noble gas<br />

temperatures. This consideration is a rather schematic estimation, as it ignores that parameters<br />

like <strong>in</strong>filtration speed, depths of ground water tables, soil structure and diffusive transport<br />

variations may <strong>in</strong>troduce more complex relations between soil temperature, soil atmosphere<br />

composition and ground water recharge.<br />

Further complicat<strong>in</strong>g the situation is the fact that the OD model handles O2+CO2 deficits <strong>in</strong> a<br />

s<strong>in</strong>gle parameter POD to account for the additional fraction Ci,od <strong>in</strong>troduced <strong>by</strong> oxygen depletion<br />

<strong>in</strong>to the total measured concentration Ci,m. This parameter is optimized for an entire dataset. It<br />

is unlikely, though, that the local and global climate parameters that <strong>in</strong>fluenced soil respiration<br />

would rema<strong>in</strong> constant over time scales relevant to paleoclimatology, and therefore the use of the<br />

same, s<strong>in</strong>gle parameter for every sample collected from an aquifer appears unreasonable to properly<br />

describe the <strong>in</strong>fluence of oxygen depletion. Noble gas paleotemperatures have been used to<br />

measure the temperature differences between the last glacial maximum and today. These differences<br />

were found to be with<strong>in</strong> the range of 4 – 5 ◦ C [Stute et al., 1995; Aeschbach-Hertig et al.,<br />

2000; Wieser, 2011]. Based on the Q10 factor provided <strong>by</strong> W<strong>in</strong>kler et al. [1996], assum<strong>in</strong>g soil<br />

respiration rates to be 25 % smaller dur<strong>in</strong>g the glacial maximum just due to the temperature<br />

change appears reasonable. Us<strong>in</strong>g the same parameter POD to characterize the effect of oxygen<br />

depletion for the entire dataset unlikely reflects the physical conditions.<br />

69


Chapter 7<br />

Summary<br />

The results of this study’s measurements confirm that O2 and CO2 concentrations with<strong>in</strong> the<br />

soil atmosphere fluctuate strongly. The measured changes <strong>in</strong> noble gas concentrations should<br />

not be neglected when calculat<strong>in</strong>g paleotemperatures from dissolved noble gas concentrations <strong>in</strong><br />

ground water. These changes are clearly correlated to the deviation of O2+CO2 concentration<br />

from the atmospheric value. The comb<strong>in</strong>ation of noble gas, O2 and CO2 read<strong>in</strong>gs implies that<br />

the change <strong>in</strong> noble gas concentrations might be dependent on isotope mass, for reasons hitherto<br />

unknown. It should be noted that the measurements of both the lightest isotope 3 He as well as<br />

the two heaviest isotopes 84 Kr and 132 Xe suffer from low accuracy and significant scatter.<br />

However, the observed <strong>in</strong>creases <strong>in</strong> noble gas concentrations at this study’s sampl<strong>in</strong>g site were<br />

not large enough to entirely account for deviations of noble gas temperatures and mean annual<br />

air temperatures as observed <strong>by</strong> Ma et al. [2004] and Hall et al. [2005] <strong>in</strong> studies of recent ground<br />

waters that prompted the proposition of the oxygen depletion model to better describe the excess<br />

air component of dissolved noble gases. It is questionable whether it is reasonable to transfer and<br />

compare these results directly to any ground water recharge area, as the processes controll<strong>in</strong>g O2<br />

consumption and CO2 production as well as removal are manifold and complicated. Variations of<br />

diverse parameters on local as well as temporal scales have strong <strong>in</strong>fluences on the magnitude<br />

of soil respiration and there<strong>by</strong> on the noble gas concentrations of soil atmospheres. Due to<br />

these mostly meteorological <strong>in</strong>fluences, soil respiration probably does not <strong>in</strong>fluence the absolute<br />

calculated noble gas temperatures <strong>by</strong> <strong>in</strong>troduc<strong>in</strong>g a constant offset but rather <strong>by</strong> <strong>in</strong>dividually<br />

<strong>in</strong>fluenc<strong>in</strong>g every s<strong>in</strong>gle data po<strong>in</strong>t depend<strong>in</strong>g on the local conditions at the recharge area.<br />

This study could therefore confirm that the basic assumption of the OD model as proposed <strong>by</strong><br />

Hall et al. [2005] is valid, as noble gas enrichment <strong>in</strong>deed occurs <strong>in</strong> natural soil atmospheres and<br />

is able to <strong>in</strong>fluence ground water noble gas concentrations significantly. In light of the measured<br />

fluctuations of soil atmosphere compositions and the <strong>in</strong>fluence of several different parameters on<br />

microbial activity <strong>in</strong> the soil, it is however likely that the OD model <strong>in</strong> its current implementation,<br />

despite deliver<strong>in</strong>g statistically good results, does not capture the physical reality of most ground<br />

water recharge sites very well.<br />

71


Chapter 8<br />

Outlook<br />

Though it was successfully shown that noble gas concentrations are <strong>in</strong>deed <strong>in</strong>fluenced <strong>by</strong> fluctuations<br />

<strong>in</strong> soil atmosphere composition, some questions rema<strong>in</strong> unanswered and new ones were<br />

raised. One <strong>in</strong>terest<strong>in</strong>g aspect is why there appears to be a mass discrim<strong>in</strong>ation when it comes<br />

to the reaction of noble gas concentrations on O2+CO2 changes. The presently available data is<br />

unfortunately quite scarce and of low accuracy, especially at the light and heavy end of the spectrum<br />

of measured noble gas isotopes. Improv<strong>in</strong>g accuracy as well as an <strong>in</strong>crease of the dataset<br />

size seems to be called for to confirm this observation. Additionally, the goal of a full annual<br />

dataset could not be achieved due to time constra<strong>in</strong>s, and while the irrigation did provide valuable<br />

<strong>in</strong>sight, it did disturb the natural annual cycle. Therefore a cont<strong>in</strong>uation and improvement<br />

of this <strong>in</strong>vestigation should be attempted.<br />

Dur<strong>in</strong>g the work on this study several flaws <strong>in</strong> the construction of the sampl<strong>in</strong>g sites became<br />

apparent, leav<strong>in</strong>g only one of three sites <strong>in</strong>tact at the time this <strong>thesis</strong> was written. Based on the<br />

accumulated experience, improvements to the sampl<strong>in</strong>g sites as well as the sampl<strong>in</strong>g protocol<br />

are possible. Future sampl<strong>in</strong>g sites should consist of a group of boreholes of different depths<br />

<strong>in</strong> close proximity, each equipped with only one sampl<strong>in</strong>g tube. While significantly <strong>in</strong>creas<strong>in</strong>g<br />

<strong>in</strong>itial setup time and effort, this would reduce and compartmentalize the risk of seal<strong>in</strong>g failure.<br />

Increas<strong>in</strong>g sampl<strong>in</strong>g depth should be attempted as well, ideally until reach<strong>in</strong>g the ground water<br />

table. Many microbiological processes occur at much deeper depths than those explored <strong>in</strong><br />

this study. These might have different <strong>in</strong>fluences on the total soil atmosphere composition.<br />

Especially the soil atmosphere just above the capillary fr<strong>in</strong>ge would be most <strong>in</strong>terest<strong>in</strong>g to<br />

analyze as it is closest to the aquifer where the last air contact of the <strong>in</strong>filtrat<strong>in</strong>g water occurs.<br />

An automatization of O2 and CO2 measurement would improve the capability to record the<br />

fluctuations the soil atmosphere with <strong>in</strong>creased resolution and might allow resolv<strong>in</strong>g diurnal<br />

fluctuations.<br />

Measurement of the noble gas samples should be further improved as well, especially for the<br />

heavier isotopes 84 Kr and 132 Xe. The determ<strong>in</strong>ation of the sample gas amount <strong>by</strong> measur<strong>in</strong>g<br />

the <strong>in</strong>let pressure needs improvement as well, ma<strong>in</strong>ly <strong>by</strong> <strong>in</strong>creas<strong>in</strong>g the accuracy of both the<br />

pressure the volume measurement.<br />

73


8 Outlook<br />

Consider<strong>in</strong>g the manifold of parameters <strong>in</strong>fluenc<strong>in</strong>g the microbial activity, an expansion of the<br />

sampl<strong>in</strong>g sites to different soil environments and climates might help to understand how sensitive<br />

the noble gas concentrations are to changes <strong>in</strong> these parameters. This might allow for a better<br />

parameterization of the effect of oxygen depletion on noble gas concentrations <strong>in</strong> ground water.<br />

74


Appendix A<br />

Calculations<br />

A.1 Gas sample size<br />

A.1.1 Vapor pressure of water<br />

Equation from Se<strong>in</strong>feld and Pandis [2006].<br />

pH2O[mbar] = φ[%]<br />

100 · p0 H2O[mbar] (A.1)<br />

p 0 H2O[mbar] = 1013.25 · exp [13.3185a − 1.97a 2 − 0.6445a 3 − 0.1299a 4 ] (A.2)<br />

a = 1 − 373.15<br />

TLab[K]<br />

(A.3)<br />

• φ : relative humidity of the sample gas<br />

• p0 H2O : saturated vapor pressure of water<br />

A.1.2 Calculation of sample size us<strong>in</strong>g the <strong>in</strong>let pressure<br />

�<br />

(p<strong>in</strong>let[mbar] − pH2O[mbar]) · VC[ml] −<br />

n(p)[mol] =<br />

mCu[g]<br />

ρCu[ g<br />

cc ]<br />

�<br />

(TLab[ ◦C] + 273.15) · R � �<br />

J · 10000<br />

• n(p) : amount of gas <strong>in</strong>side the sample tube<br />

molK<br />

• p<strong>in</strong>let : pressure <strong>in</strong>side the <strong>in</strong>let volume after crack<strong>in</strong>g the sample tube<br />

• pH2O : water vapor pressure at laboratory temperature<br />

• VC : <strong>in</strong>let volume<br />

75<br />

(A.4)


A.1. Gas sample size A Calculations<br />

• mCu : mass of the sample tube<br />

• ρCu : density of copper<br />

• TLab : laboratory temperature<br />

• R : universal gas constant<br />

Gaussian error propagation<br />

(∆pH2O) 2 =<br />

(∆n(P )) 2 =<br />

�<br />

∆a =<br />

373.15<br />

(TLab + 273.15)<br />

2 · ∆TLab<br />

�<br />

∆φ<br />

100 · 1013.25 · exp [13.3185a − 1.97a2 − 0.6445a 3 − 0.1299a 4 ]<br />

�<br />

φ<br />

+<br />

100 · 1013.25 · � 13.3185 − 2 · 1.97a − 3 · 0.6445a 2 − 4 · 0.1299a 3�<br />

· exp � 13.3185a − 1.97a 2 − 0.6445a 3 − 0.1299a 4� · ∆a � 2<br />

VC − mCu<br />

ρCu<br />

· ∆pKeller<br />

(TLab + 273.15) · R · 10000<br />

�<br />

�2 p<strong>in</strong>let − pH2O<br />

+<br />

· ∆VC<br />

(TLab + 273.15) · R · 10000<br />

⎛<br />

+ ⎝ (p<strong>in</strong>let<br />

�<br />

− pH2O) · VC − mCu<br />

�<br />

ρCu<br />

(TLab + 273.15) 2 · R · 10000<br />

� 2<br />

· ∆TLab<br />

+<br />

�<br />

+<br />

⎞<br />

⎠<br />

�<br />

� 2<br />

VC − mCu<br />

�2 ρCu<br />

· ∆pH2O<br />

(TLab + 273.15) · R · 10000<br />

p<strong>in</strong>let − pH2O<br />

· ∆mCu<br />

(TLab + 273.15) · R · ρCu · 10000<br />

A.1.3 Calculation of sample size us<strong>in</strong>g the sample tube length<br />

n(L)[mol] = (psample[mbar] − pH2O[mbar]) · V (L)[ml]<br />

(T [ ◦C] + 273.15) · R � �<br />

J · 10000<br />

• n(L) : amount of gas <strong>in</strong>side the sample tube<br />

• psample : sample pressure <strong>in</strong>side the copper tube<br />

• pH2O : water vapor pressure dur<strong>in</strong>g sampl<strong>in</strong>g<br />

2<br />

molK<br />

• V (L) : <strong>in</strong>ner volume of the sample tube, see below for calculation<br />

• T : temperature of the soil air dur<strong>in</strong>g sampl<strong>in</strong>g<br />

• R : universal gas constant<br />

76<br />

(A.5)<br />

� 2


A Calculations A.2. Mass spectrometer <strong>in</strong>let volume<br />

Calculation of the sample tube volume<br />

The <strong>in</strong>ner volume of the copper sample tubes is calculated us<strong>in</strong>g an empirical formula <strong>by</strong> Wieser<br />

[2006], who characterized the relation of length to volume of copper tubes of the same k<strong>in</strong>d,<br />

squeezed shut us<strong>in</strong>g the same method as applied here. The <strong>in</strong>ner volume is calculated from the<br />

length of the copper tube, measured after the sample was extracted:<br />

• L : length of the sample tube<br />

• A = (−0.12284 ± 0.00336) cm 3 : empirical parameter<br />

• B = (0.13583 ± 6.35 × 10 −4 ) cm 2 : empirical parameter<br />

Gaussian error propagation<br />

(∆n(L)) 2 =<br />

V (L)[cm 3 ] = B · L[cm] + A (A.6)<br />

�<br />

V (L)<br />

· ∆psample<br />

(T + 273.15) · R · 10000<br />

�<br />

psample − pH2O<br />

+<br />

· ∆V (L)<br />

(T + 273.15) · R · 10000<br />

A.2 Mass spectrometer <strong>in</strong>let volume<br />

� 2<br />

� 2<br />

�<br />

V (L)<br />

+<br />

· ∆pH2O<br />

(T + 273.15) · R · 10000<br />

�<br />

psample − pH2O<br />

+<br />

(T + 273.15) 2 · ∆T<br />

· R · 10000<br />

The volume of the mass spectrometer <strong>in</strong>let was volumetrically measured, us<strong>in</strong>g the follow<strong>in</strong>g set<br />

of previously measured [T. Marx, personal note] volumes:<br />

V1 = (38, 876 ± 0, 470) ml<br />

Vkl = (77, 338 ± 0, 040) ml<br />

Volumes V1 and Vkl were connected and flooded with atmospheric air at pressures between 1 and<br />

11 mbar. The <strong>in</strong>let volume V<strong>in</strong>let and the unknown volume V2 connect<strong>in</strong>g V<strong>in</strong>let and (V1+Vkl) were<br />

evacuated. All volumes were at the same temperature when the gas was expanded successively<br />

<strong>in</strong>to the evacuated volumes. The pressures p1, p2 and p3 correspond to the volumes (V1 + Vkl),<br />

(V1 + Vkl + V2) and (V1 + Vkl + V2 + V<strong>in</strong>let) respectively. The <strong>in</strong>let volume is then calculated us<strong>in</strong>g<br />

the follow<strong>in</strong>g equation:<br />

p1(V1 + Vkl) = p2 [(V1 + Vkl) + V2]<br />

� �<br />

p1<br />

⇒ V2 = (V1 + Vkl) − 1<br />

p1(V1 + Vkl) = p3 [(V1 + Vkl) + V2 + VC]<br />

77<br />

p2<br />

� 2<br />

� 2


A.3. Fractionation-Values A Calculations<br />

For the measured pressures see Table B.6.<br />

Gaussian error propagation<br />

(∆V2) 2 =<br />

(∆V<strong>in</strong>let) 2 =<br />

�� p1<br />

p2<br />

A.3 Fractionation-Values<br />

⇒ V<strong>in</strong>let = p1<br />

(V1 + Vkl) − (V1 + Vkl) − V2<br />

p3<br />

�<br />

p1<br />

= (V1 + Vkl) − 1 −<br />

p3<br />

p1<br />

�<br />

+ 1<br />

p2<br />

� �<br />

p1<br />

= (V1 + Vkl)<br />

� �2 ��<br />

p1<br />

− 1 ∆V1 +<br />

p2<br />

p3<br />

− p1<br />

p2<br />

� �2 − 1 ∆Vkl<br />

�� � �2 ��<br />

(V1 + Vkl) +<br />

∆p1 + −<br />

p2<br />

(V1 + Vkl)p1<br />

p2 2<br />

� �2 ∆p2<br />

��<br />

p1<br />

−<br />

p3<br />

p1<br />

� �2 ��<br />

p1<br />

∆V1 + −<br />

p2<br />

p3<br />

p1<br />

� �2 ∆Vkl<br />

p2<br />

��<br />

(V1 + Vkl) +<br />

−<br />

p3<br />

(V1<br />

� �2 ��<br />

+ Vkl) ∆p1 + −<br />

p2<br />

(V1 + Vkl)p1<br />

p2 2<br />

� �2 ∆p2<br />

��<br />

+ − (V1 + Vkl)p1<br />

p3 2<br />

� �2 ∆p3<br />

The Fractionation-Values are calculated based on Kendrick et al. [2006], but use 40 Ar <strong>in</strong>stead<br />

of 36 Ar as the reference isotope, s<strong>in</strong>ce <strong>in</strong> this study’s environment, <strong>in</strong>fluences <strong>by</strong> radiogenic 40 Ar<br />

are negligible.<br />

F X =<br />

�<br />

X<br />

�<br />

�<br />

40Ar sample<br />

� X<br />

�<br />

40Ar air<br />

(A.7)<br />

where X stands for the noble gas isotopes. The F-Value def<strong>in</strong>e a mix<strong>in</strong>g l<strong>in</strong>e between atmospheric<br />

composition and measured composition, if a sample noble gas isotope has an F-Value of 1 its<br />

volume mix<strong>in</strong>g ratio equals that of atmospheric air.<br />

78


Appendix B<br />

Data<br />

Table B.1: Parameters for fitt<strong>in</strong>g noble gas solubilities from Benson and Krause [1976].<br />

Noble gas a0 a1 [ ◦ K] a2 [ ◦ K 2 ]<br />

He -5.0746 -4127.8 627250<br />

Ne -4.2988 -4871.1 793580<br />

Ar -4.2123 -5239.6 995240<br />

Kr -3.6326 -5664.0 1122400<br />

Xe -2.0917 -6693.5 1341700<br />

Table B.2: Results of measur<strong>in</strong>g a test gas comprised of 15.0 % O2, 2.0 % CO2, 2.2 % CH4 and<br />

100 ppm CO us<strong>in</strong>g the BM2000. The data <strong>in</strong>dicates that the device slightly overestimates<br />

the O2 levels, but rema<strong>in</strong>s well with<strong>in</strong> the given accuracy of 1 %.<br />

CO2 [Vol%] CH4 [Vol%] O2 [Vol%] CO [ppm]<br />

1.8 2.1 15.3 107<br />

1.8 2.1 15.4 107<br />

1.9 2.1 15.5 108<br />

1.9 2.0 15.6 107<br />

1.8 2.1 15.4 106<br />

1.8 2.1 15.3 106<br />

1.8 2.1 15.3 106<br />

1.9 2.1 15.2 106<br />

1.9 2.1 15.2 105<br />

1.9 2.1 15.3 105<br />

1.9 2.1 15.1 104<br />

1.9 2.1 15.1 104<br />

79


Table B.3: Noble gas volume mix<strong>in</strong>g ratios and isotope composition <strong>in</strong> dry atmospheric air,<br />

compiled <strong>by</strong> Porcelli et al. [2002].<br />

Noble gas Volume fraction Isotope Relative isotopic<br />

<strong>in</strong> atmospheric air abundance<br />

He (5.24 ± 0.05)×10 −6 3 He 0.000140<br />

4 He 100<br />

Ne (1.82 ± 0.04)×10 −5 20 Ne 90.5<br />

21 Ne 0.268<br />

22 Ne 9.23<br />

Ar (9.34 ± 0.01)×10 −3 36 Ar 0.3364<br />

38 Ar 0.0632<br />

40 Ar 99.60<br />

Kr (1.14 ± 0.01)×10 −6 78 Kr 0.3469<br />

80 Kr 2.2571<br />

82 Kr 11.523<br />

83 Kr 11.477<br />

84 Kr 57.00<br />

86 Kr 17.398<br />

Xe (8.7 ± 0.1)×10 −8 124 Xe 0.0951<br />

126 Xe 0.0887<br />

128 Xe 1.919<br />

129 Xe 26.44<br />

130 Xe 4.070<br />

131 Xe 21.22<br />

132 Xe 26.89<br />

134 Xe 10.430<br />

136 Xe 8.857<br />

80<br />

B Data


B Data<br />

Table B.4: Calculated values of relative humidity with<strong>in</strong> mass spectrometer <strong>in</strong>let for small<br />

sample tube size and sampl<strong>in</strong>g temperatures rang<strong>in</strong>g from 0 to 25 ◦ CC, based on an average<br />

sample volume of (0.745 ± 0.028) cm 3 , a temperature uncerta<strong>in</strong>ty of ∆T = 5 ◦ C and 100 %<br />

relative humidity of soil air.<br />

T [ ◦ C] φ<strong>in</strong>let [%] ∆φ<strong>in</strong>let [%]<br />

0 0.6 0.3<br />

1 0.7 0.3<br />

2 0.7 0.3<br />

3 0.8 0.4<br />

4 0.8 0.4<br />

5 0.9 0.4<br />

6 0.9 0.4<br />

7 1.0 0.5<br />

8 1.0 0.5<br />

9 1.1 0.5<br />

10 1.2 0.5<br />

11 1.3 0.6<br />

12 1.4 0.6<br />

13 1.5 0.6<br />

14 1.5 0.7<br />

15 1.6 0.7<br />

16 1.7 0.8<br />

17 1.9 0.8<br />

18 2.0 0.9<br />

19 2.1 0.9<br />

20 2.2 1.0<br />

21 2.3 1.0<br />

22 2.5 1.1<br />

23 2.6 1.1<br />

24 2.8 1.2<br />

25 2.9 1.2<br />

81


Table B.5: Calculated values of relative humidity with<strong>in</strong> mass spectrometer <strong>in</strong>let for large<br />

sample tubes and sampl<strong>in</strong>g temperatures rang<strong>in</strong>g from 0 to 25 ◦ C, based on an average<br />

sample volume of (1.579 ± 0.029) cm 3 , a temperature uncerta<strong>in</strong>ty of ∆T = 5 ◦ C and 100 %<br />

relative humidity of soil air.<br />

T [ ◦ C] φ<strong>in</strong>let [%] ∆φ<strong>in</strong>let [%]<br />

0 1.3 0.4<br />

1 1.4 0.4<br />

2 1.5 0.4<br />

3 1.6 0.4<br />

4 1.7 0.5<br />

5 1.8 0.5<br />

6 2.0 0.5<br />

7 2.1 0.6<br />

8 2.2 0.6<br />

9 2.4 0.6<br />

10 2.5 0.7<br />

11 2.7 0.7<br />

12 2.9 0.8<br />

13 3.1 0.8<br />

14 3.3 0.9<br />

15 3.5 0.9<br />

16 3.7 1.0<br />

17 3.9 1.0<br />

18 4.2 1.1<br />

19 4.4 1.1<br />

20 4.7 1.2<br />

21 5.0 1.3<br />

22 5.2 1.4<br />

23 5.6 1.4<br />

24 5.9 1.5<br />

25 6.2 1.6<br />

82<br />

B Data


B Data<br />

Table B.6: Results of the volumetric measurements done to calculate the <strong>in</strong>let volume V<strong>in</strong>let<br />

for the setup dur<strong>in</strong>g the measur<strong>in</strong>g runs 1 – 3 (M22 – M33) and run 4 (M42 – M56), which<br />

featured a slightly different setup. Measurements M1 to M21 and M34 to M41 were omitted<br />

due to systematic errors.<br />

ID p1 [mbar] p2 [mbar] p3 [mbar]<br />

M22 2.206 1.738 1.472<br />

M23 5.191 4.088 3.462<br />

M24 9.094 7.161 6.062<br />

M25 9.464 7.448 6.303<br />

M26 9.864 7.760 6.565<br />

M27 6.981 5.490 4.642<br />

M28 4.935 3.880 3.284<br />

M29 3.492 2.747 2.324<br />

M30 3.470 2.728 2.308<br />

M31 2.452 1.927 1.630<br />

M32 4.759 3.740 3.161<br />

M33 7.838 6.162 5.210<br />

M42 8.912 7.025 5.936<br />

M43 3.935 3.101 2.621<br />

M44 10.770 8.487 7.166<br />

M45 5.051 3.977 3.359<br />

M46 8.848 6.966 5.883<br />

M47 9.503 7.484 6.320<br />

M48 6.716 5.288 4.462<br />

M49 4.757 3.747 3.162<br />

M50 3.366 2.651 2.239<br />

M51 6.002 4.722 3.986<br />

M52 10.680 8.409 7.098<br />

M53 4.694 3.696 3.120<br />

M54 7.407 5.829 4.921<br />

M55 10.580 8.336 7.034<br />

M56 8.557 6.733 5.682<br />

Table B.7: M<strong>in</strong>imum, maximum and average relative errors ∆ of the noble gas isotope gas<br />

amount measurements from the mass spectrometer before recalculation to concentrations.<br />

The low accuracy of the 132 Xe measurements was heavily <strong>in</strong>fluenced <strong>by</strong> 132 Xe calibration<br />

measurements splitt<strong>in</strong>g <strong>in</strong> two dist<strong>in</strong>ct branches for reasons unknown.<br />

3 He [%] 4 He [%] 20 Ne [%] 22 Ne [%] 36 Ar [%] 40 Ar [%] 84 Kr [%] 132 Xe [%]<br />

∆m<strong>in</strong> 1.71 0.44 0.24 0.34 1.39 1.08 1.86 2.84<br />

∆avg 1.89 0.46 0.29 0.42 1.48 1.09 2.13 3.26<br />

∆max 2.77 0.60 0.43 0.78 1.82 1.36 2.62 4.68<br />

83


Table B.8: Precipitation record of the months February, March and April <strong>in</strong> the past 20<br />

years, recorded <strong>by</strong> a Deutscher Wetterdienst weather station (ID 10729) near Mannheim.<br />

The total precipitation dur<strong>in</strong>g these months <strong>in</strong> 2011, 53.0 mm is well below the 20 year<br />

average of 133.2 mm.<br />

Precipitation [mm]<br />

Year February March April Total<br />

1991 18.3 35.2 27.4 80.9<br />

1992 36.3 43.7 38.5 118.5<br />

1993 9.7 23.0 23.8 56.5<br />

1994 30.9 56.9 63.2 151.0<br />

1995 37.4 53.6 98.5 189.5<br />

1996 52.0 31.2 13.8 97.0<br />

1997 49.7 28.6 34.0 112.3<br />

1998 24.4 33.8 106.8 165.0<br />

1999 61.7 73.6 42.8 178.1<br />

2000 59.3 66.3 41.9 167.5<br />

2001 57.4 125.7 56.0 239.1<br />

2002 105.8 41.6 50.6 198.0<br />

2003 10.3 27.7 17.6 55.6<br />

2004 28.9 22.8 20.0 71.7<br />

2005 59.0 28.8 71.2 159.0<br />

2006 22.2 64.6 43.7 130.5<br />

2007 72.0 70.1 0.7 142.8<br />

2008 44.4 59.1 72.0 175.5<br />

2009 61.2 54.7 44.6 160.5<br />

2010 35.8 35.7 23.3 94.8<br />

2011 23.7 16.3 13.0 53.0<br />

� 42.9 47.3 43.0 133.2<br />

84<br />

B Data


B Data<br />

Table B.9: Complete dataset (Offset corrected) of noble gases measured at Site A. Values<br />

marked red were excluded from analysis. Values (except ratios) are given <strong>in</strong> % atmospheric<br />

air, mean<strong>in</strong>g they are standardized to atmospheric air noble gas concentrations taken from<br />

Porcelli et al. [2002].<br />

!" #$%&'()*+"$,- .-/0.-1 2.-/0.-1 3-4503-44 23-4503-44 6715067/8 26715067/8 .-103-45 2.-103-45 .-/ 2.-/ .-1 2.-1 3-45 23-45 3-44 23-44 67/8 267/8 6715 26715 97:1 297:1 ;-


Table B.10: Complete dataset of noble gases measured at Site B and Site C. Values (except<br />

ratios) are given <strong>in</strong> % atmospheric air, mean<strong>in</strong>g they are standardized to atmospheric air<br />

noble gas concentrations taken from Porcelli et al. [2002].<br />

!" #$%&'()*+"$,- .-/0.-1 2.-/0.-1 3-4503-44 23-4503-44 6715067/8 26715067/8 .-103-45 2.-103-45 .-/ 2.-/ .-1 2.-1 3-45 23-45 3-44 23-44 67/8 267/8 6715 26715 97:1 297:1 ;-


B Data<br />

Table B.11: Complete dataset of O2 and CO2 concentrations measured at Site A. All values<br />

are given <strong>in</strong> Vol%. NG ID identifies the noble gas sample ID correspond<strong>in</strong>g to those O2 and<br />

CO2 read<strong>in</strong>gs.<br />

Date NG ID<br />

2m 4m 6m 2m 4m 6m<br />

2m 4m 6m<br />

O2 ΔO2 O2 ΔO2 O2 ΔO2 CO2 ΔCO2 CO2 ΔCO2 CO2 ΔCO2 O2+CO2 ΔO2+CO2 O2+CO2 ΔO2+CO2 O2+CO2 ΔO2+CO2<br />

15.02.11 A11 17.9 1.0 17.8 1.0 17.3 1.0 4.6 0.5 4.7 0.5 4.7 0.5 22.5 1.1 22.5 1.1 22.0 1.1<br />

26.02.11 A12 19.1 1.0 19.0 1.0 18.6 1.0 4.4 0.5 4.5 0.5 4.5 0.5 23.5 1.1 23.5 1.1 23.1 1.1<br />

28.03.11 A13 18.1 1.0 18.3 1.0 18.6 1.0 3.7 0.5 3.7 0.5 3.6 0.5 21.8 1.1 22.0 1.1 22.2 1.1<br />

30.03.11 17.0 1.0 17.3 1.0 17.3 1.0 3.3 0.5 3.7 0.5 3.6 0.5 20.3 1.1 21.0 1.1 20.9 1.1<br />

07.04.11 A14 17.7 1.0 17.5 1.0 18.0 1.0 3.6 0.5 3.5 0.5 3.4 0.5 21.3 1.1 21.0 1.1 21.4 1.1<br />

12.04.11 A15 18.7 1.0 18.2 1.0 18.7 1.0 4.0 0.5 4.0 0.5 3.8 0.5 22.7 1.1 22.2 1.1 22.5 1.1<br />

22.04.11 A16 17.0 1.0 17.9 1.0 18.5 1.0 3.6 0.5 3.5 0.5 3.4 0.5 20.6 1.1 21.4 1.1 21.9 1.1<br />

28.04.11 17.6 1.0 16.2 1.0 17.6 1.0 3.8 0.5 3.9 0.5 3.6 0.5 21.4 1.1 20.1 1.1 21.2 1.1<br />

04.05.11 18.5 1.0 17.9 1.0 17.9 1.0 3.3 0.5 3.5 0.5 3.4 0.5 21.8 1.1 21.4 1.1 21.3 1.1<br />

05.05.11 18.1 1.0 17.1 1.0 16.8 1.0 3.2 0.5 3.4 0.5 3.4 0.5 21.3 1.1 20.5 1.1 20.2 1.1<br />

09.05.11 18.8 1.0 18.2 1.0 17.7 1.0 2.8 0.5 3.0 0.5 3.3 0.5 21.6 1.1 21.2 1.1 21.0 1.1<br />

09.05.11 18.4 1.0 17.8 1.0 17.4 1.0 2.4 0.5 3.1 0.5 3.3 0.5 20.8 1.1 20.9 1.1 20.7 1.1<br />

09.05.11 17.6 1.0 17.4 1.0 17.1 1.0 3.0 0.5 2.9 0.5 3.1 0.5 20.6 1.1 20.3 1.1 20.2 1.1<br />

09.05.11 17.9 1.0 17.5 1.0 17.3 1.0 3.2 0.5 3.1 0.5 3.2 0.5 21.1 1.1 20.6 1.1 20.5 1.1<br />

09.05.11 18.1 1.0 18.1 1.0 18.1 1.0 3.6 0.5 3.1 0.5 3.2 0.5 21.7 1.1 21.2 1.1 21.3 1.1<br />

10.05.11 17.0 1.0 17.2 1.0 17.6 1.0 3.7 0.5 3.6 0.5 3.3 0.5 20.7 1.1 20.8 1.1 20.9 1.1<br />

10.05.11 17.0 1.0 17.1 1.0 17.4 1.0 2.0 0.5 3.2 0.5 3.3 0.5 19.0 1.1 20.3 1.1 20.7 1.1<br />

10.05.11 A17 17.1 1.0 17.2 1.0 17.5 1.0 2.4 0.5 3.3 0.5 3.2 0.5 19.5 1.1 20.5 1.1 20.7 1.1<br />

10.05.11 16.9 1.0 17.2 1.0 17.5 1.0 2.6 0.5 3.2 0.5 3.1 0.5 19.5 1.1 20.4 1.1 20.6 1.1<br />

11.05.11 15.3 1.0 16.6 1.0 17.4 1.0 3.9 0.5 3.7 0.5 3.3 0.5 19.2 1.1 20.3 1.1 20.7 1.1<br />

11.05.11 A18 15.0 1.0 16.7 1.0 17.4 1.0 2.5 0.5 3.4 0.5 3.3 0.5 17.5 1.1 20.1 1.1 20.7 1.1<br />

11.05.11 15.3 1.0 17.1 1.0 17.6 1.0 2.8 0.5 3.4 0.5 3.3 0.5 18.1 1.1 20.5 1.1 20.9 1.1<br />

11.05.11 15.7 1.0 17.2 1.0 17.7 1.0 3.0 0.5 3.3 0.5 3.2 0.5 18.7 1.1 20.5 1.1 20.9 1.1<br />

12.05.11 14.4 1.0 16.7 1.0 18.1 1.0 4.2 0.5 3.8 0.5 3.4 0.5 18.6 1.1 20.5 1.1 21.5 1.1<br />

12.05.11 14.4 1.0 17.3 1.0 18.0 1.0 4.2 0.5 3.7 0.5 3.4 0.5 18.6 1.1 21.0 1.1 21.4 1.1<br />

12.05.11 13.8 1.0 16.9 1.0 17.7 1.0 4.2 0.5 3.7 0.5 3.4 0.5 18.0 1.1 20.6 1.1 21.1 1.1<br />

13.05.11 13.4 1.0 16.1 1.0 17.5 1.0 4.5 0.5 3.9 0.5 3.4 0.5 17.9 1.1 20.0 1.1 20.9 1.1<br />

13.05.11 A19 13.4 1.0 16.9 1.0 18.3 1.0 3.3 0.5 3.8 0.5 3.4 0.5 16.7 1.1 20.7 1.1 21.7 1.1<br />

13.05.11 13.5 1.0 17.0 1.0 17.8 1.0 3.5 0.5 3.7 0.5 3.4 0.5 17.0 1.1 20.7 1.1 21.2 1.1<br />

13.05.11 13.5 1.0 16.8 1.0 17.4 1.0 3.6 0.5 3.7 0.5 3.4 0.5 17.1 1.1 20.5 1.1 20.8 1.1<br />

13.05.11 13.9 1.0 17.1 1.0 17.7 1.0 3.8 0.5 3.7 0.5 3.3 0.5 17.7 1.1 20.8 1.1 21.0 1.1<br />

13.05.11 14.0 1.0 17.3 1.0 18.0 1.0 4.0 0.5 3.7 0.5 3.4 0.5 18.0 1.1 21.0 1.1 21.4 1.1<br />

14.05.11 13.1 1.0 16.9 1.0 18.1 1.0 4.6 0.5 3.8 0.5 3.4 0.5 17.7 1.1 20.7 1.1 21.5 1.1<br />

14.05.11 A20 12.6 1.0 16.1 1.0 16.8 1.0 4.6 0.5 3.7 0.5 3.3 0.5 17.2 1.1 19.8 1.1 20.1 1.1<br />

14.05.11 13.0 1.0 17.2 1.0 18.0 1.0 4.6 0.5 3.8 0.5 3.4 0.5 17.6 1.1 21.0 1.1 21.4 1.1<br />

14.05.11 12.9 1.0 17.1 1.0 17.8 1.0 4.7 0.5 3.8 0.5 3.4 0.5 17.6 1.1 20.9 1.1 21.2 1.1<br />

14.05.11 12.5 1.0 16.1 1.0 18.1 1.0 4.8 0.5 4.0 0.5 3.4 0.5 17.3 1.1 20.1 1.1 21.5 1.1<br />

15.05.11 12.0 1.0 15.6 1.0 17.5 1.0 5.0 1.0 4.1 0.5 3.4 0.5 17.0 1.4 19.7 1.1 20.9 1.1<br />

15.05.11 12.3 1.0 17.0 1.0 18.2 1.0 5.0 1.0 4.0 0.5 3.4 0.5 17.3 1.4 21.0 1.1 21.6 1.1<br />

15.05.11 12.2 1.0 17.1 1.0 18.0 1.0 5.0 1.0 3.9 0.5 3.4 0.5 17.2 1.4 21.0 1.1 21.4 1.1<br />

15.05.11 11.9 1.0 16.3 1.0 17.6 1.0 5.0 1.0 4.0 0.5 3.4 0.5 16.9 1.4 20.3 1.1 21.0 1.1<br />

15.05.11 12.1 1.0 15.9 1.0 18.2 1.0 5.1 1.0 4.2 0.5 3.4 0.5 17.2 1.4 20.1 1.1 21.6 1.1<br />

15.05.11 12.2 1.0 16.1 1.0 18.1 1.0 5.1 1.0 4.2 0.5 3.4 0.5 17.3 1.4 20.3 1.1 21.5 1.1<br />

16.05.11 12.0 1.0 16.0 1.0 18.3 1.0 5.2 1.0 4.3 0.5 3.4 0.5 17.2 1.4 20.3 1.1 21.7 1.1<br />

16.05.11 11.9 1.0 16.2 1.0 17.9 1.0 5.2 1.0 4.1 0.5 3.4 0.5 17.1 1.4 20.3 1.1 21.3 1.1<br />

16.05.11 12.0 1.0 17.1 1.0 18.0 1.0 5.2 1.0 3.9 0.5 3.4 0.5 17.2 1.4 21.0 1.1 21.4 1.1<br />

16.05.11 12.7 1.0 17.2 1.0 17.8 1.0 5.1 1.0 3.8 0.5 3.4 0.5 17.8 1.4 21.0 1.1 21.2 1.1<br />

16.05.11 12.7 1.0 17.2 1.0 17.9 1.0 5.1 1.0 3.8 0.5 3.4 0.5 17.8 1.4 21.0 1.1 21.3 1.1<br />

17.05.11 11.8 1.0 15.5 1.0 17.7 1.0 5.4 1.0 4.3 0.5 3.4 0.5 17.2 1.4 19.8 1.1 21.1 1.1<br />

17.05.11 11.9 1.0 16.2 1.0 18.0 1.0 5.4 1.0 4.2 0.5 3.5 0.5 17.3 1.4 20.4 1.1 21.5 1.1<br />

17.05.11 11.5 1.0 16.1 1.0 17.5 1.0 5.4 1.0 4.1 0.5 3.4 0.5 16.9 1.4 20.2 1.1 20.9 1.1<br />

17.05.11 11.7 1.0 16.9 1.0 17.9 1.0 5.5 1.0 3.9 0.5 3.4 0.5 17.2 1.4 20.8 1.1 21.3 1.1<br />

18.05.11 11.9 1.0 15.7 1.0 18.0 1.0 5.6 1.0 4.4 0.5 3.5 0.5 17.5 1.4 20.1 1.1 21.5 1.1<br />

18.05.11 11.7 1.0 15.0 1.0 17.3 1.0 5.7 1.0 4.5 0.5 3.5 0.5 17.4 1.4 19.5 1.1 20.8 1.1<br />

18.05.11 A21 11.8 1.0 15.3 1.0 17.0 1.0 5.6 1.0 4.4 0.5 3.4 0.5 17.4 1.4 19.7 1.1 20.4 1.1<br />

18.05.11 A22 12.0 1.0 16.5 1.0 17.7 1.0 5.3 1.0 4.3 0.5 3.5 0.5 17.3 1.4 20.8 1.1 21.2 1.1<br />

18.05.11 12.2 1.0 16.7 1.0 17.5 1.0 5.3 1.0 4.1 0.5 3.5 0.5 17.5 1.4 20.8 1.1 21.0 1.1<br />

18.05.11 12.2 1.0 16.5 1.0 17.2 1.0 5.2 1.0 3.9 0.5 3.4 0.5 17.4 1.4 20.4 1.1 20.6 1.1<br />

19.05.11 10.9 1.0 15.8 1.0 17.5 1.0 5.7 1.0 4.2 0.5 3.5 0.5 16.6 1.4 20.0 1.1 21.0 1.1<br />

19.05.11 10.9 1.0 16.0 1.0 17.1 1.0 5.7 1.0 4.1 0.5 3.4 0.5 16.6 1.4 20.1 1.1 20.5 1.1<br />

19.05.11 11.0 1.0 15.9 1.0 17.5 1.0 5.8 1.0 4.2 0.5 3.5 0.5 16.8 1.4 20.1 1.1 21.0 1.1<br />

20.05.11 11.1 1.0 15.1 1.0 17.7 1.0 6.0 1.0 4.7 0.5 3.5 0.5 17.1 1.4 19.8 1.1 21.2 1.1<br />

20.05.11 11.0 1.0 14.9 1.0 17.3 1.0 6.0 1.0 4.6 0.5 3.5 0.5 17.0 1.4 19.5 1.1 20.8 1.1<br />

20.05.11 A23 11.1 1.0 16.0 1.0 17.8 1.0 5.9 1.0 4.3 0.5 3.5 0.5 17.0 1.4 20.3 1.1 21.3 1.1<br />

21.05.11 15.9 1.0 17.0 1.0 17.8 1.0 4.3 0.5 3.8 0.5 3.4 0.5 20.2 1.1 20.8 1.1 21.2 1.1<br />

22.05.11 11.3 1.0 16.2 1.0 17.2 1.0 5.5 1.0 3.8 0.5 3.3 0.5 16.8 1.4 20.0 1.1 20.5 1.1<br />

23.05.11 11.1 1.0 15.5 1.0 17.1 1.0 6.0 1.0 4.3 0.5 3.4 0.5 17.1 1.4 19.8 1.1 20.5 1.1<br />

24.05.11 11.0 1.0 14.3 1.0 17.4 1.0 6.6 1.0 5.1 1.0 3.5 0.5 17.6 1.4 19.4 1.4 20.9 1.1<br />

25.05.11 11.1 1.0 14.0 1.0 17.1 1.0 6.7 1.0 5.3 1.0 3.5 0.5 17.8 1.4 19.3 1.4 20.6 1.1<br />

26.05.11 11.4 1.0 13.2 1.0 16.4 1.0 6.8 1.0 5.6 1.0 3.4 0.5 18.2 1.4 18.8 1.4 19.8 1.1<br />

26.05.11 11.9 1.0 14.4 1.0 17.5 1.0 6.8 1.0 5.3 1.0 3.5 0.5 18.7 1.4 19.7 1.4 21.0 1.1<br />

27.05.11 12.2 1.0 14.0 1.0 16.5 1.0 6.7 1.0 5.3 1.0 3.5 0.5 18.9 1.4 19.3 1.4 20.0 1.1<br />

28.05.11 12.8 1.0 14.5 1.0 17.0 1.0 6.7 1.0 5.3 1.0 3.5 0.5 19.5 1.4 19.8 1.4 20.5 1.1<br />

29.05.11 13.1 1.0 15.3 1.0 17.4 1.0 6.6 1.0 5.0 1.0 3.5 0.5 19.7 1.4 20.3 1.4 20.9 1.1<br />

30.05.11 13.4 1.0 15.2 1.0 17.4 1.0 6.6 1.0 5.2 1.0 3.6 0.5 20.0 1.4 20.4 1.4 21.0 1.1<br />

31.05.11 14.0 1.0 15.5 1.0 17.3 1.0 6.4 1.0 4.9 0.5 3.6 0.5 20.4 1.4 20.4 1.1 20.9 1.1<br />

01.06.11 14.3 1.0 15.4 1.0 17.1 1.0 6.3 1.0 5.0 1.0 3.6 0.5 20.6 1.4 20.4 1.4 20.7 1.1<br />

02.06.11 14.8 1.0 15.0 1.0 16.9 1.0 6.2 1.0 5.5 1.0 3.6 0.5 21.0 1.4 20.5 1.4 20.5 1.1<br />

31.05.11 14.5 1.0 14.5 1.0 15.9 1.0 6.1 1.0 5.3 1.0 3.6 0.5 20.6 1.4 19.8 1.4 19.5 1.1<br />

31.05.11 15.1 1.0 15.3 1.0 17.0 1.0 6.2 1.0 5.3 1.0 3.6 0.5 21.3 1.4 20.6 1.4 20.6 1.1<br />

01.06.11 15.6 1.0 15.7 1.0 17.3 1.0 5.7 1.0 5.2 1.0 3.7 0.5 21.3 1.4 20.9 1.4 21.0 1.1<br />

02.06.11 15.3 1.0 15.4 1.0 16.8 1.0 5.6 1.0 5.1 1.0 3.7 0.5 20.9 1.4 20.5 1.4 20.5 1.1<br />

03.06.11 15.5 1.0 15.0 1.0 16.1 1.0 5.7 1.0 5.4 1.0 3.6 0.5 21.2 1.4 20.4 1.4 19.7 1.1<br />

04.06.11 15.6 1.0 15.5 1.0 16.6 1.0 6.0 1.0 5.3 1.0 3.6 0.5 21.6 1.4 20.8 1.4 20.2 1.1<br />

06.06.11 16.5 1.0 16.0 1.0 16.9 1.0 5.3 1.0 5.5 1.0 3.7 0.5 21.8 1.4 21.5 1.4 20.6 1.1<br />

07.06.11 16.2 1.0 16.1 1.0 16.8 1.0 5.6 1.0 5.2 1.0 3.7 0.5 21.8 1.4 21.3 1.4 20.5 1.1<br />

08.06.11 17.1 1.0 16.7 1.0 17.1 1.0 5.0 1.0 5.1 1.0 3.7 0.5 22.1 1.4 21.8 1.4 20.8 1.1<br />

09.06.11 16.4 1.0 16.0 1.0 16.6 1.0 5.2 1.0 4.9 0.5 3.7 0.5 21.6 1.4 20.9 1.1 20.3 1.1<br />

10.06.11 16.7 1.0 16.2 1.0 16.7 1.0 5.1 1.0 4.9 0.5 3.7 0.5 21.8 1.4 21.1 1.1 20.4 1.1<br />

11.06.11 17.0 1.0 16.7 1.0 16.9 1.0 4.9 0.5 4.7 0.5 3.8 0.5 21.9 1.1 21.4 1.1 20.7 1.1<br />

13.06.11 17.2 1.0 16.8 1.0 16.8 1.0 4.6 0.5 4.5 0.5 3.7 0.5 21.8 1.1 21.3 1.1 20.5 1.1<br />

14.06.11 17.9 1.0 17.5 1.0 16.8 1.0 4.0 0.5 4.2 0.5 3.7 0.5 21.9 1.1 21.7 1.1 20.5 1.1<br />

14.06.11 17.1 1.0 16.5 1.0 16.6 1.0 4.4 0.5 4.2 0.5 3.7 0.5 21.5 1.1 20.7 1.1 20.3 1.1<br />

14.06.11 17.3 1.0 16.9 1.0 16.7 1.0 4.5 0.5 4.4 0.5 3.7 0.5 21.8 1.1 21.3 1.1 20.4 1.1<br />

15.06.11 17.6 1.0 17.0 1.0 16.5 1.0 4.0 0.5 4.1 0.5 3.7 0.5 21.6 1.1 21.1 1.1 20.2 1.1<br />

15.06.11 17.0 1.0 16.5 1.0 16.6 1.0 4.8 0.5 4.5 0.5 3.7 0.5 21.8 1.1 21.0 1.1 20.3 1.1<br />

16.06.11 17.5 1.0 16.6 1.0 16.6 1.0 4.2 0.5 4.3 0.5 3.8 0.5 21.7 1.1 20.9 1.1 20.4 1.1<br />

16.06.11 17.8 1.0 17.4 1.0 16.8 1.0 3.9 0.5 4.1 0.5 3.7 0.5 21.7 1.1 21.5 1.1 20.5 1.1<br />

17.06.11 17.4 1.0 16.8 1.0 16.6 1.0 4.1 0.5 4.2 0.5 3.8 0.5 21.5 1.1 21.0 1.1 20.4 1.1<br />

18.06.11 17.5 1.0 17.1 1.0 16.7 1.0 4.1 0.5 4.2 0.5 3.9 0.5 21.6 1.1 21.3 1.1 20.6 1.1<br />

20.06.11 17.3 1.0 17.2 1.0 16.6 1.0 4.4 0.5 4.2 0.5 3.9 0.5 21.7 1.1 21.4 1.1 20.5 1.1<br />

21.06.11 17.6 1.0 17.2 1.0 16.3 1.0 3.9 0.5 4.0 0.5 3.9 0.5 21.5 1.1 21.2 1.1 20.2 1.1<br />

22.06.11 17.4 1.0 17.3 1.0 16.2 1.0 3.7 0.5 3.7 0.5 4.0 0.5 21.1 1.1 21.0 1.1 20.2 1.1<br />

23.06.11 16.7 1.0 16.4 1.0 16.0 1.0 4.6 0.5 4.2 0.5 4.0 0.5 21.3 1.1 20.6 1.1 20.0 1.1<br />

24.06.11 A24 17.5 1.0 17.3 1.0 16.6 1.0 3.9 0.5 4.0 0.5 4.0 0.5 21.4 1.1 21.3 1.1 20.6 1.1<br />

24.06.11 17.4 1.0 16.6 1.0 16.3 1.0 4.3 0.5 4.2 0.5 4.0 0.5 21.7 1.1 20.8 1.1 20.3 1.1<br />

27.06.11 17.9 1.0 17.4 1.0 16.1 1.0 3.5 0.5 3.8 0.5 4.0 0.5 21.4 1.1 21.2 1.1 20.1 1.1<br />

27.06.11 17.8 1.0 17.3 1.0 16.0 1.0 3.5 0.5 3.8 0.5 4.0 0.5 21.3 1.1 21.1 1.1 20.0 1.1<br />

28.06.11 18.0 1.0 17.7 1.0 16.0 1.0 3.2 0.5 3.4 0.5 4.0 0.5 21.2 1.1 21.1 1.1 20.0 1.1<br />

28.06.11 17.6 1.0 17.2 1.0 15.8 1.0 3.4 0.5 3.5 0.5 3.9 0.5 21.0 1.1 20.7 1.1 19.7 1.1<br />

87


Appendix C<br />

Additional plots and figures<br />

Figure C.1: Detailed view of the perforated tip of a sampl<strong>in</strong>g tube through which the soil<br />

atmosphere was sampled. Install<strong>in</strong>g f<strong>in</strong>er filters to shield the tube from <strong>in</strong>filtration of sludge<br />

was deemed unnecessary based on laboratory experiments with fully saturated soil where<br />

sufficient flow rates were achieved. This was confirmed at Site C’s 2 m depth tube, where<br />

the pump rates through the tube rema<strong>in</strong>ed viable even after part of the tube was filled with<br />

soil after pump<strong>in</strong>g sludge/water upwards and had dried up <strong>in</strong>side.<br />

89


Figure C.2: Tak<strong>in</strong>g samples at Site A <strong>in</strong> December 2010.<br />

Figure C.3: Tak<strong>in</strong>g samples at Site B <strong>in</strong> April 2011.<br />

90<br />

C Additional plots and figures


C Additional plots and figures<br />

Figure C.4: Copper sample tubes, ∼0.7 cc variety (top) and ∼1.5 cc variety (bottom).<br />

Figure C.5: Access<strong>in</strong>g the LogTag data loggers at Site A.<br />

91


C Additional plots and figures<br />

Figure C.6: Cont<strong>in</strong>uous O2 and CO2 measurements at Site A, show<strong>in</strong>g O2 and CO2 each<br />

reach<strong>in</strong>g a constant value soon after <strong>in</strong>itiation of pump<strong>in</strong>g. Accuracies are 1 % for O2 and<br />

0.5 % for CO2, error bars have been omitted for better visibility.<br />

�� � �� � � � � � � � � � ��<br />

�� � � � �� � �� � �� ���� � ����� � �� � �� �� �� � ��� � ��<br />

Figure C.7: All atmospheric O2 measurements taken at Site A and Site B with the BM2000<br />

over time, the error weighted mean fit gives (20.89 ± 0.04) Vol% O2.<br />

92


C Additional plots and figures<br />

� � � � � � � � � � � ��� � �<br />

Figure C.8: Temperature measurements taken at Site A at 0.1 m, 1.0 m and 2.5 m depth<br />

and s<strong>in</strong>usoidal fits. Depth placement accuracy is ±0.1 m, temperature accuracy is ±0.5 ◦ C.<br />

The atmospheric daily mean temperatures were provided <strong>by</strong> meteomedia, measured <strong>by</strong> a<br />

weather station close to Site A. The offset visible at 1.0 m depth <strong>in</strong> August 2010 is caused<br />

<strong>by</strong> a moisture <strong>in</strong>trusion <strong>in</strong>to the sensor <strong>in</strong> August 2010 after refill<strong>in</strong>g the borehole with<br />

sludge. Comparison with the data from Site B leads to the conclusion that the sensor was<br />

affected <strong>by</strong> this until early October 2010. The smaller offset at 2.5 m depth <strong>in</strong> December 2010<br />

and January 2011 co<strong>in</strong>cide with heavy precipitation (Figure C.10). The low atmospheric<br />

temperatures at the time <strong>in</strong>dicate that this also is an offset caused <strong>by</strong> moisture <strong>in</strong>trusion<br />

rather than an actual temperature change <strong>by</strong> <strong>in</strong>filtrat<strong>in</strong>g ra<strong>in</strong> water of different temperature.<br />

Such an event is visible <strong>in</strong> May 2011, when cold tap water was used for irrigation, lead<strong>in</strong>g to<br />

spikes <strong>in</strong> the temperature curves of 1.0 m and 2.5 m. The sensor at 0.1 m depth was shielded<br />

aga<strong>in</strong>st these sudden and short-lived temperature changes <strong>by</strong> be<strong>in</strong>g <strong>in</strong>sulated <strong>by</strong> air <strong>in</strong>side<br />

the <strong>in</strong>strumentation box.<br />

93


C Additional plots and figures<br />

Figure C.9: Model data of daily top soil moisture of Europe for May 29 th , 2011, provided<br />

<strong>by</strong> the European Commission Jo<strong>in</strong>t Research Centre Institute for Environment and Susta<strong>in</strong>ability<br />

(http://edo.jrc.ec.europa.eu/php/<strong>in</strong>dex.php?action=viewid=20). Based<br />

on measured meteorological data from JRC-MARS (http://www.marsop.<strong>in</strong>fo/marsop3)<br />

94


C Additional plots and figures<br />

Figure C.10: Daily sum of precipitation and daily mean relative humidity, provided <strong>by</strong><br />

meteomedia’s weather station at Site A.<br />

� � �� �� �� ��<br />

� � �� �� �� ��<br />

Figure C.11: Noble gas concentrations of atmospheric air samples relative to literature values<br />

from Porcelli et al. [2002], for 84 Kr and 132 Xe.<br />

95


C Additional plots and figures<br />

Figure C.12: Depth profile of O2 and CO2 at a one-time borehole on August 11th, 2009<br />

with<strong>in</strong> a radius of 10 m of Site A, based on data <strong>by</strong> Schneider [2010].<br />

96


C Additional plots and figures<br />

O 2 [Vol%]<br />

CO 2 [Vol%]<br />

O 2 + CO 2 [Vol%]<br />

20<br />

19<br />

18<br />

17<br />

16<br />

15<br />

14<br />

13<br />

12<br />

11<br />

10<br />

7<br />

6<br />

5<br />

4<br />

3<br />

2<br />

24<br />

23<br />

22<br />

21<br />

20<br />

19<br />

18<br />

17<br />

16<br />

15<br />

2m<br />

4m<br />

6m<br />

2m<br />

4m<br />

6m<br />

Feb Mar Apr May Jun<br />

2m<br />

4m<br />

6m<br />

Month, 2011<br />

Feb Mar Apr May Jun<br />

Month, 2011<br />

Feb Mar Apr May Jun<br />

Month, 2011<br />

Figure C.13: Concentrations of O2, CO2 and O2+CO2 at Site A dur<strong>in</strong>g February to June<br />

2011. Accuracies are 1 % for O2 and 0.5 % for CO2 below 5 % absolute and 1 % above 5 %,<br />

error bars have been omitted for better visibility.<br />

97


O 2 [Vol%]<br />

CO 2 [Vol%]<br />

O 2 + CO 2 [Vol%]<br />

20<br />

19<br />

18<br />

17<br />

16<br />

15<br />

14<br />

13<br />

12<br />

11<br />

10<br />

7<br />

6<br />

5<br />

4<br />

3<br />

2<br />

23<br />

22<br />

21<br />

20<br />

19<br />

18<br />

17<br />

16<br />

2m<br />

4m<br />

6m<br />

2m<br />

4m<br />

6m<br />

15<br />

9 May 11 May 13 May 15 May 17 May 19 May<br />

Date, 2011<br />

C Additional plots and figures<br />

Figure C.14: Concentrations of O2, CO2 and O2+CO2 at Site A dur<strong>in</strong>g the irrigation<br />

period. Accuracies are 1 % for O2 and 0.5 % for CO2 below 5 % absolute and 1 % above 5 %,<br />

error bars have been omitted for better visibility.<br />

98<br />

2m<br />

4m<br />

6m


C Additional plots and figures<br />

4<br />

He [% atm. air]<br />

4<br />

He [% atm. air]<br />

4<br />

He [% atm. air]<br />

Figure C.15: Concentrations of O2 and CO2 at Site B. Accuracies are 1 % for O2 and 0.5 %<br />

for CO2 error bars have been omitted for better visibility.<br />

Oct Nov Apr May<br />

110<br />

105<br />

100<br />

95<br />

90<br />

105<br />

100<br />

95<br />

90<br />

105<br />

100<br />

95<br />

B06 at 1 m depth<br />

B06 at 3 m depth<br />

A23 at 2 m depth<br />

90<br />

Oct Nov Apr May<br />

Date of measurement <strong>in</strong> mass spectrometer<br />

20<br />

Ne [% atm. air]<br />

20<br />

Ne [% atm. air]<br />

20<br />

Ne [% atm. air]<br />

Oct Nov Apr May<br />

110<br />

105<br />

100<br />

95<br />

90<br />

105<br />

100<br />

95<br />

90<br />

105<br />

100<br />

95<br />

B06 at 1 m depth<br />

B06 at 3 m depth<br />

A23 at 2 m depth<br />

90<br />

Oct Nov Apr May<br />

Date of measurement <strong>in</strong> mass spectrometer<br />

Figure C.16: Concentrations of 4 He and 20 Ne of samples that were measured multiple times,<br />

relative to atmospheric literature values from Porcelli et al. [2002], <strong>in</strong> relation to their date<br />

of measur<strong>in</strong>g.<br />

99


3<br />

He (2m) [% atm. air]<br />

3<br />

He (4m) [% atm. air]<br />

3<br />

He (6m) [% atm. air]<br />

84<br />

Kr (2m) [% atm. air]<br />

84<br />

Kr (4m) [% atm. air]<br />

84<br />

Kr (6m) [% atm. air]<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

Date<br />

4<br />

He (2m) [% atm. air]<br />

4<br />

He (4m) [% atm. air]<br />

4<br />

He (6m) [% atm. air]<br />

132<br />

Xe (2m) [% atm. air]<br />

132<br />

Xe (4m) [% atm. air]<br />

132<br />

Xe (6m) [% atm. air]<br />

110<br />

105<br />

100<br />

95<br />

90<br />

C Additional plots and figures<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

Aug Sep Oct Nov Dec Jan Feb Mar Apr May<br />

Figure C.17: Concentrations of 3 He, 4 He, 84 Kr and 132 Xe <strong>in</strong> soil atmosphere samples from<br />

Site A over time, relative to atmospheric literature values from Porcelli et al. [2002].<br />

100<br />

Date


C Additional plots and figures<br />

20<br />

Ne (1m) [% atm. air]<br />

20<br />

Ne (3m) [% atm. air]<br />

20<br />

Ne (5m) [% atm. air]<br />

36<br />

Ar (1m) [% atm. air]<br />

36<br />

Ar (3m) [% atm. air]<br />

36<br />

Ar (5m) [% atm. air]<br />

Aug Sep Oct Nov Dec Jan<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

Aug Sep Oct Nov Dec Jan<br />

Date<br />

22<br />

Ne (1m) [% atm. air]<br />

22<br />

Ne (3m) [% atm. air]<br />

22<br />

Ne (5m) [% atm. air]<br />

40<br />

Ar (1m) [% atm. air]<br />

40<br />

Ar (3m) [% atm. air]<br />

40<br />

Ar (5m) [% atm. air]<br />

Aug Sep Oct Nov Dec Jan<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

Aug Sep Oct Nov Dec Jan<br />

Figure C.18: Concentrations of 20 Ne, 22 Ne, 36 Ar and 40 Ar <strong>in</strong> soil atmosphere samples from<br />

Site B over time, relative to atmospheric literature values from Porcelli et al. [2002].<br />

101<br />

Date


3<br />

He (1m) [% atm. air]<br />

3<br />

He (3m) [% atm. air]<br />

3<br />

He (5m) [% atm. air]<br />

84<br />

Kr (1m) [% atm. air]<br />

84<br />

Kr (3m) [% atm. air]<br />

84<br />

Kr (5m) [% atm. air]<br />

Aug Sep Oct Nov Dec Jan<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

Aug Sep Oct Nov Dec Jan<br />

Date<br />

4<br />

He (1m) [% atm. air]<br />

4<br />

He (3m) [% atm. air]<br />

4<br />

He (5m) [% atm. air]<br />

132<br />

Xe (1m) [% atm. air]<br />

132<br />

Xe (3m) [% atm. air]<br />

132<br />

Xe (5m) [% atm. air]<br />

110<br />

105<br />

100<br />

95<br />

90<br />

C Additional plots and figures<br />

Aug Sep Oct Nov Dec Jan<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

115<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

110<br />

105<br />

100<br />

95<br />

90<br />

Aug Sep Oct Nov Dec Jan<br />

Figure C.19: Concentrations of 3 He, 4 He, 84 Kr and 132 Xe <strong>in</strong> soil atmosphere samples from<br />

Site B over time, relative to atmospheric literature values from Porcelli et al. [2002].<br />

102<br />

Date


C Additional plots and figures<br />

Figure C.20: 20 Ne/ 22 Ne ratios of Site A 2 m samples from before (A11 – A16) and dur<strong>in</strong>g<br />

(A17 - A23) the irrigation period. The data is absolute measured gas amounts.<br />

Figure C.21: 4 He/ 20 Ne ratios of Site A 2 m samples from before (A11 – A16) and dur<strong>in</strong>g<br />

(A17 - A23) the irrigation period. The data is absolute measured gas amounts.<br />

103


10<br />

10<br />

Theoretical expectation as described <strong>in</strong> Formula 2.19<br />

4<br />

He from samples A11 - A23, all depths<br />

Theoretical expectation as described <strong>in</strong> Formula 2.19<br />

3<br />

He from samples A11 - A23, all depths<br />

8<br />

8<br />

6<br />

6<br />

4<br />

4<br />

2<br />

He compared to atm. air [%]<br />

2<br />

He compared to atm. air [%]<br />

0<br />

0<br />

-2<br />

Relative change of 4<br />

-2<br />

Relative change of 3<br />

-4<br />

-4<br />

-6<br />

-6<br />

16 18 20 22 24<br />

16 18 20 22 24<br />

10<br />

10<br />

Theoretical expectation as described <strong>in</strong> Formula 2.19<br />

22<br />

Ne from samples A11 - A23, all depths<br />

Theoretical expectation as described <strong>in</strong> Formula 2.19<br />

20<br />

Ne from samples A11 - A23, all depths<br />

8<br />

8<br />

6<br />

6<br />

4<br />

4<br />

C Additional plots and figures<br />

2<br />

Ne compared to atm. air [%]<br />

2<br />

Ne compared to atm. air [%]<br />

0<br />

0<br />

-2<br />

Relative change of 22<br />

-2<br />

Relative change of 20<br />

Figure C.22: The measured relative deviation of 3 He, 4 He, 20 Ne and 22 Ne concentrations<br />

from atmospheric values of samples with exist<strong>in</strong>g O2 and CO2 data (A11 – A23, all depths),<br />

plotted aga<strong>in</strong>st the sample’s sum of O2 and CO2 concentration. The red l<strong>in</strong>e represents the<br />

theoretically expected behavior as described <strong>in</strong> Section 2.1.3.<br />

104<br />

-4<br />

-4<br />

-6<br />

-6<br />

16 18 20 22 24<br />

16 18 20 22 24<br />

O 2 + CO 2 [Vol%] O 2 + CO 2 [Vol%]


C Additional plots and figures<br />

10<br />

10<br />

Theoretical expectation as described <strong>in</strong> Formula 2.19<br />

40<br />

Ar from samples A11 - A23, all depths<br />

Theoretical expectation as described <strong>in</strong> Formula 2.19<br />

39<br />

Ar from samples A11 - A23, all depths<br />

8<br />

8<br />

6<br />

6<br />

4<br />

4<br />

2<br />

Ar compared to atm. air [%]<br />

2<br />

Ar compared to atm. air [%]<br />

0<br />

0<br />

-2<br />

Relative change of 40<br />

-2<br />

Relative change of 39<br />

-4<br />

-4<br />

-6<br />

-6<br />

16 18 20 22 24<br />

16 18 20 22 24<br />

10<br />

10<br />

Theoretical expectation as described <strong>in</strong> Formula 2.19<br />

132<br />

Xe from samples A11 - A23, all depths<br />

Theoretical expectation as described <strong>in</strong> Formula 2.19<br />

84<br />

Kr from samples A11 - A23, all depths<br />

8<br />

8<br />

6<br />

6<br />

4<br />

4<br />

2<br />

Xe compared to atm. air [%]<br />

2<br />

Kr compared to atm. air [%]<br />

0<br />

0<br />

-2<br />

Relative change of 132<br />

-2<br />

Relative change of 84<br />

Figure C.23: The measured relative deviation of 36 Ar, 40 Ar, 84 Kr and 132 Xe concentrations<br />

from atmospheric values of samples with exist<strong>in</strong>g O2 and CO2 data (A11 – A23, all depths),<br />

plotted aga<strong>in</strong>st the sample’s sum of O2 and CO2 concentration. The red l<strong>in</strong>e represents the<br />

theoretically expected behavior as described <strong>in</strong> Section 2.1.3.<br />

105<br />

-4<br />

-4<br />

-6<br />

-6<br />

16 18 20 22 24<br />

16 18 20 22 24<br />

O 2 + CO 2 [Vol%]<br />

O 2 + CO 2 [Vol%]


F 3 He<br />

F 22 Ne<br />

F 84 Kr<br />

1.10<br />

1.05<br />

1.00<br />

0.95<br />

0.90<br />

0.90 0.95 1.00 1.05 1.10<br />

1.10<br />

1.05<br />

1.00<br />

0.95<br />

F 4 He<br />

0.90<br />

0.90 0.95 1.00 1.05 1.10<br />

1.10<br />

1.05<br />

1.00<br />

0.95<br />

F 4 He<br />

0.90<br />

0.90 0.95 1.00 1.05 1.10<br />

F 4 He<br />

F 20 Ne<br />

F 36 Ar<br />

F 132 Xe<br />

1.10<br />

1.05<br />

1.00<br />

0.95<br />

C Additional plots and figures<br />

0.90<br />

0.90 0.95 1.00 1.05 1.10<br />

1.10<br />

1.05<br />

1.00<br />

0.95<br />

F 4 He<br />

0.90<br />

0.90 0.95 1.00 1.05 1.10<br />

1.10<br />

1.05<br />

1.00<br />

0.95<br />

F 4 He<br />

0.90<br />

0.90 0.95 1.00 1.05 1.10<br />

Figure C.24: F-Values (see Section A.3 for def<strong>in</strong>ition) of 3 He, 20 Ne, 22 Ne, 36 Ar, 84 Kr and<br />

132 Xe versus F-Values of 4 He. The data is from all noble gas samples of Site A and all<br />

depths. Correlations are only visible for 20 Ne and 22 Ne but not systematically as the magnitude<br />

of noble gas enrichment has no discernible <strong>in</strong>fluence on the position of the datapo<strong>in</strong>t.<br />

106<br />

F 4 He


Appendix D<br />

Datasheets<br />

107


Besondere Merkmale des Biogasmonitors BM2000<br />

analytische Systeme und<br />

Componenten GmbH<br />

� tragbar, batteriebetrieben, langzeitstabil, zuverlässige Messwerte,<br />

� misst 4 Gase: Methan, Kohlendioxid, Sauerstoff und Schwefelwasserstoff,<br />

� Infrarot-Technik - ke<strong>in</strong>e gegenseitige Bee<strong>in</strong>flussung von CO 2 und CH4,<br />

� Abwesenheit von O 2 bee<strong>in</strong>flusst Messwerte nicht,<br />

� Datenspeicher für 250 Wertesätze mit Anzeigefunktion am Messgerät und<br />

Datenübertragung auf PC mit speziellem Programm,<br />

� Luftdruckmessung (900 bis 1100 mbar), Druckmessung <strong>in</strong> der Messgasleitung,<br />

� starke Pumpe bis 400 mbar Unterdruck (rel).<br />

Technische Daten des Biogasmonitors BM2000<br />

Messpr<strong>in</strong>zip: Infrarotabsorption für CH 4 und CO 2 bei selektiven Wellenlängen,<br />

elektrochemische Zelle für O 2 (langlebig, ca. 5 Jahre Lebensdauer),<br />

Messbereiche: 0 - 100 Vol.-% für CH 4 , 0 - 100 Vol.-% für CO 2 ,<br />

0 - 25 Vol.-% für O 2 , 0-10 000 ppm für H 2 S (Nachweisgrenze ca. 100 ppm).<br />

(elektrochemische Sensoren zeigen Querempf<strong>in</strong>dlichkeiten,<br />

beim Schwefelwasserstoff (H2S) Sensor z.B. durch Wasserstoff (H2))<br />

Fördervolumen der <strong>in</strong>ternen Probenahmepumpe: ca. 0,5 l/m<strong>in</strong> bei freiem Fluss,<br />

Pumpe stoppt bei ca. 400 mbar Unterdruck (rel).<br />

Genauigkeit: Messkomponente: CH 4 CO 2 O 2<br />

Messbereich: 0-100 0-100 0-25 (Vol.-%)<br />

0 - 5 �0,5 �0,5 �1,0 (Vol.-%)<br />

5 - 15 �1,0 �1,0 �1,0 (Vol.-%)<br />

>15 �3,0 �3,0 �1,0 (Vol.-%)<br />

Genauigkeit: Messparameter: Druck<br />

Messbereich: 900-1100 (mbar)<br />

900-1100 � 5 (mbar)<br />

Temperatur: 0°C bis 40°C,<br />

Feuchte: 0 bis 95 % rH, nicht kondensierend,<br />

Batteriebetrieb: bei normalem Betrieb ca. 8 Stunden (im Auslieferungszustand),<br />

Ladezeit: ca. 2 Stunden bei vollständiger Entladung der Batterien,<br />

Schutzart: Gehäuse IP 65,<br />

Display: LCD Anzeige, beleuchtbar, 40x16 Zeichen,<br />

Maße: ca. 19 cm x 25 cm x 6 cm,<br />

Gewicht: ca. 2,0 kg.<br />

Optionelles Zubehör für den tragbaren Biogasmonitor BM2000<br />

� Temperatursonde: Messbereich 0°C - 100°C<br />

(im Bereich 0°C - 40°C Ex-Zulassung gültig ),<br />

� externe Messe<strong>in</strong>richtungen z.B. für kle<strong>in</strong>e Konzentrationen an H2S,<br />

mit ATEX-Zulassung,<br />

� Sensor zur Messung von Gesamtgasfluss, Anemometer,<br />

(nicht für den Ex-Bereich),<br />

� RS-Kabel und Software zur Datenübertragung (nicht EX-zugelassen).<br />

Weitere Produkte<br />

� zusätzliche Sonden zur Messung des Wasserstandes (Lichtlote), des<br />

Sickerwasserstandes oder der Mächtigkeit von Ölphasen auf Wasser.<br />

Technische Änderungen vorbehalten 08/10 D 1I<br />

ansyco analytische Systeme und Componenten GmbH · Ostr<strong>in</strong>g 4 · 76131 Karlsruhe<br />

Telefon 0721 / 626560 · Telefax 0721 / 621332 · E-Mail <strong>in</strong>fo@ansyco.de · http://www.ansyco.de<br />

Figure D.1: Datasheet Geotech BM2000 Biogas Monitor.<br />

108<br />

D Datasheets


D Datasheets<br />

LogTag TREX-8 Technical Specifications<br />

Part Order Code TREX-8<br />

Remote Temperature Sensor Measurement range -40 ~ +99°C (-40°F ~ +210°F)<br />

Recorder operat<strong>in</strong>g temperature range -40 ~ +85°C (-40°F ~ +185°F)<br />

Rated Temperature read<strong>in</strong>g accuracy *<br />

* Ex-factory values with standard 1.5m LogTag supplied probe<br />

With recorder case sitt<strong>in</strong>g <strong>in</strong> environmental temperature<br />

between 0°C ~ 50°C the rated accuracy is :-<br />

±0.5°C for -10°C~ +40°C<br />

better than ±0.7°C for -10°C~ -30°C & +40°C~+60°C<br />

better than ±0.8°C for -30°C~ -40°C & +60°C~+80°C<br />

better than ±1.0°C for +80°C~+99°C<br />

For generalized environment conditions see Chart below<br />

Rated absolute accuracy<br />

±3.0ºC<br />

±2.5ºC<br />

±2.0ºC<br />

±1.5ºC<br />

±1.0ºC<br />

±.5ºC<br />

LogTag TREX Temperature Recorder<br />

Rated remote sensor temperature read<strong>in</strong>g accuracy<br />

-40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90 100<br />

Temperature (ºC)<br />

recorder @ +25°C recorder @ - 20°C recorder @ -40°C<br />

Actual performance is normally much better than the rated<br />

values.<br />

Accuracy figures can be improved <strong>by</strong> recalibration.<br />

Rated Native Resolution<br />

Rated Temperature read<strong>in</strong>g resolution #<br />

less than 0.1°C for -40°C ~ +40°C,<br />

less than 0.2°C for +40°C ~+80°C<br />

less than 0.6°C for +80°C ~+99°C (see below)<br />

1.0ºC<br />

0.9ºC<br />

0.8ºC<br />

0.7ºC<br />

0.6ºC<br />

0.5ºC<br />

0.4ºC<br />

0.3ºC<br />

0.2ºC<br />

0.1ºC<br />

LogTag TREX Temperature Recorder<br />

Rated native temperature read<strong>in</strong>g resolution<br />

-40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90 100<br />

Temperature (ºC)<br />

# LogTag Analyser currently displays to one decimal place of °C<br />

or °F. The native resolution is what is stored <strong>in</strong> the LogTag.<br />

Capacity 8000 temperature read<strong>in</strong>gs (16K<strong>by</strong>tes memory)<br />

Sampl<strong>in</strong>g frequency adjustable, 30 sec to several hours<br />

Download Time<br />

Typically with full memory <strong>in</strong> less than 5 seconds<br />

depend<strong>in</strong>g on computer or readout device used.<br />

Environmental IP61<br />

Power source 3V Lithium - replaceable <strong>by</strong> qualified technician<br />

Battery life<br />

2~3 years typical use – longer (up to 5-10 years)<br />

if recorder is hibernated between uses.<br />

Standard Remote Sensor Cable Lengths<br />

Standard: 1.5m (4’11”)<br />

Extended: 3 m (9’10”) (recommended maximum)<br />

Remote Sensor Cable Type PTFE (FDA food contact rated) coaxial<br />

Size 86mm(H)x54.5mm(W)x8.6mm(T)<br />

Weight 35grams without ST100<br />

Case Material Polycarbonate<br />

Figure D.2: Datasheet LogTag TREX-8 data logger and sensor. TRIX-8 is identical but<br />

with the sensor build <strong>in</strong>to the cas<strong>in</strong>g.<br />

DOC REV G 28/11/07 – ©Copyright 2007, LogTag Recorders Ltd. release version Page 2 of 4<br />

109


HIGHLY PRECISE DIGITAL MANOMETER<br />

PRECISION: 0,01 %FS *<br />

LEX 1 is a micro-processor controlled, accurate and versatile digital pressure measur<strong>in</strong>g<br />

<strong>in</strong>strument with <strong>in</strong>tegrated Max.-/M<strong>in</strong>.-function for calibration and test<strong>in</strong>g purposes.<br />

The pressure is measured twice per second and displayed. The top display <strong>in</strong>dicates the<br />

actual pressure, the bottom display shows the Max.- or M<strong>in</strong>.-pressure s<strong>in</strong>ce the last RESET.<br />

LEX 1 has two operat<strong>in</strong>g keys. The left key is to turn the <strong>in</strong>strument on, to select the functions<br />

and the pressure units. The right key executes the selected function resp. unit or serves to<br />

display the Max.- and M<strong>in</strong>.-value.<br />

The <strong>in</strong>strument has the follow<strong>in</strong>g functions:<br />

RESET: With the RESET-function, the Max.- and M<strong>in</strong>.-value is set to the actual pressure<br />

value.<br />

ZERO: The ZERO-function allows to set any value as a new Zero reference. Barometric<br />

pressure variations can thus be compensated.<br />

The factory sett<strong>in</strong>g of the Zero for the ranges ≤ 30 bar is at 0 bar absolute. For sealed<br />

gauge pressure measurements, activate “ZERO SEt” at ambient pressure. Instruments<br />

with ranges > 30 bar are calibrated <strong>in</strong> a sealed gauge mode with ambiant<br />

pressure as a Zero reference.<br />

CONT: The <strong>in</strong>strument turns off 15 M<strong>in</strong>. after the last key function. Activat<strong>in</strong>g CONT (Cont<strong>in</strong>uous)<br />

deactivates this automatic turn-off.<br />

UNITS: All standard <strong>in</strong>struments are calibrated <strong>in</strong> bar. The pressure can be <strong>in</strong>dicated <strong>in</strong> 13<br />

different units.<br />

Optional Accessories:<br />

Carry<strong>in</strong>g bag, protective rubber cover<strong>in</strong>g<br />

SPECIFICATIONS<br />

Pressure Ranges, Resolution, Overpressure: Range Resolution Overpressure<br />

-1…2 bar 0,1 mbar 3 bar<br />

-1…20 bar 1 mbar 30 bar<br />

0…200 bar 10 mbar 300 bar<br />

0…400 bar 50 mbar 600 bar<br />

0…1000 bar 100 mbar 1100 bar<br />

Number of Digits 5 Digit<br />

Accuracy (10…30 °C) * 0,05 %FS (<strong>in</strong>clud<strong>in</strong>g l<strong>in</strong>earity, repeatability and hysteresis)<br />

Precision * 0,05 %FS<br />

Precision optional (≥ 20 bar) * 0,025 %FS / 0,01 %FS<br />

Storage- / Operat<strong>in</strong>g Temperature -10…60 °C / 0…50 °C<br />

Compensated Temperature Range 0…50 °C<br />

Supply 3 V battery, type CR 2430<br />

Battery Life 2000 hours cont<strong>in</strong>uous operation<br />

Pressure Connection G1/4”<br />

Interface RS485; rear-sided mat<strong>in</strong>g plug “Fischer”<br />

compatible with PC-converter cable<br />

K103-A (RS232) and K104-A (USB)<br />

Protection IP65<br />

Diameter x Height x Depth 76 x 118 x 42 mm<br />

Weight 210 g<br />

KELLER AG für Druckmesstechnik St. Gallerstrasse 119 CH-8404 W<strong>in</strong>terthur Tel. 052 - 235 25 25 Fax 052 - 235 25 00<br />

KELLER Gesellschaft für Druckmesstechnik mbH Schwarzwaldstrasse 17 D-79798 Jestetten Tel. 07745 - 9214 - 0 Fax 07745 - 9214 - 50<br />

Companies approved to ISO 9001 / EN 29001<br />

Subject to alterations<br />

Figure D.3: Datasheet Keller Lex 1 pressure gauge.<br />

110<br />

Display 5 Digit LEX 1<br />

D Datasheets<br />

LEX 1<br />

* Accuracy and Precision<br />

“Accuracy” is an absolute term, “Precision” a relative<br />

term. Dead weight testers are primary standards for<br />

pressure, where the pressure is def<strong>in</strong>ed <strong>by</strong> the primary<br />

values of mass, length and time. Highest class primary<br />

standards <strong>in</strong> national laboratories <strong>in</strong>dicate the uncerta<strong>in</strong>ty<br />

of their pressure references with 70 to 90 ppM or close to<br />

0,01%.<br />

Commercial dead weight testers as used <strong>in</strong> our facilities to<br />

calibrate the transmitters and manometers <strong>in</strong>dicate an<br />

uncerta<strong>in</strong>ty or accuracy of 0,025 %. Below these levels,<br />

KELLER use the expression “Precision” as the ability of a<br />

pressure transmitter or manometer to be at each pressure<br />

po<strong>in</strong>t with<strong>in</strong> 0.01 %FS relative to these commercial standards.<br />

The manometer’s full-scale output can be set up to match<br />

any standard of your choice <strong>by</strong> correct<strong>in</strong>g the ga<strong>in</strong> with a<br />

calibration software.<br />

www.keller-druck.com<br />

12/04


Bibliography<br />

[Aeschbach-Hertig 1994] Aeschbach-Hertig, W.: Helium und Tritium als Tracer für<br />

physikalische Prozesse <strong>in</strong> Seen, ETH Zürich, Dissertation ETH Nr. 10714, 1994<br />

[Aeschbach-Hertig et al. 2002] Aeschbach-Hertig, W. ; Beyerle, U. ; Holocher, J. ;<br />

Peeters, F. ; Kipfer, R.: Excess air <strong>in</strong> groundwater as a potential <strong>in</strong>dicator of past<br />

environmental changes, 2002, p. 174–183<br />

[Aeschbach-Hertig et al. 2008] Aeschbach-Hertig, W. ; El-Gamal, H. ; Wieser, M. ;<br />

Palcsu, L.: Model<strong>in</strong>g excess air and degass<strong>in</strong>g <strong>in</strong> groundwater <strong>by</strong> equilibrium partition<strong>in</strong>g<br />

with a gas phase. In: Water Resour. Res. 44 (2008), p. W08449, doi:10.1029/2007WR006454<br />

[Aeschbach-Hertig et al. 1999a] Aeschbach-Hertig, W. ; Hofer, M. ; Kipfer, R. ; Imboden,<br />

D. ; Wieler, R.: Accumulation of mantle gases <strong>in</strong> a permanently stratified volcanic<br />

Lake (Lac Pav<strong>in</strong>, France). In: Geochim. Cosmochim. Acta 63 (1999), Nr. 19/20, p. 3357–3372<br />

[Aeschbach-Hertig et al. 1999b] Aeschbach-Hertig, W. ; Peeters, F. ; Beyerle, U. ;<br />

Kipfer, R.: Interpretation of dissolved atmospheric noble gases <strong>in</strong> natural waters. In: Water<br />

Resour. Res. 35 (1999), Nr. 9, p. 2779–2792<br />

[Aeschbach-Hertig et al. 2000] Aeschbach-Hertig, W. ; Peeters, F. ; Beyerle, U. ;<br />

Kipfer, R.: Palaeotemperature reconstruction from noble gases <strong>in</strong> ground water tak<strong>in</strong>g <strong>in</strong>to<br />

account equilibration with entrapped air. In: Nature 405 (2000), p. 1040–1044<br />

[Amundson and Davidson 1990] Amundson, R. G. ; Davidson, E. A.: Carbon dioxide and<br />

nitrogenous gases <strong>in</strong> the soil atmosphere. In: Journal of Geochemical Exploration 38 (1990),<br />

Nr. 1-2, p. 13–41<br />

[Ballent<strong>in</strong>e and Hall 1999] Ballent<strong>in</strong>e, C. ; Hall, C.: Determ<strong>in</strong><strong>in</strong>g paleotemperature and<br />

other variables <strong>by</strong> us<strong>in</strong>g an error-weighted, nonl<strong>in</strong>ear <strong>in</strong>version of noble gas concentrations <strong>in</strong><br />

water. In: Geochim. Cosmochim. Acta 63 (1999), Nr. 16, p. 2315–2336<br />

[Ballent<strong>in</strong>e and Burnard 2002] Ballent<strong>in</strong>e, C. J. ; Burnard, P. G.: Production, Release<br />

and transport of Noble Gases <strong>in</strong> the Cont<strong>in</strong>ental Crust. In: Porcelli, D. (Ed.) ; Ballent<strong>in</strong>e,<br />

C. (Ed.) ; Wieler, R. (Ed.): Noble gases <strong>in</strong> geochemistry and cosmochemistry Bd. 47.<br />

Wash<strong>in</strong>gton, DC : M<strong>in</strong>eralogical Society of America, Geochemical Society, 2002, p. 481–538<br />

[Baumgartner 1996] Baumgartner, A. (Ed.): Allgeme<strong>in</strong>e Hydrologie - quantitative Hydrologie.<br />

2. Aufl. Berl<strong>in</strong> ; Stuttgart [u.a.] : Borntraeger, 1996<br />

111


BIBLIOGRAPHY BIBLIOGRAPHY<br />

[Benson and Krause 1976] Benson, B. B. ; Krause, D.: Empirical laws for dilute aqueous<br />

solutions of nonpolar gases. In: J. Chem. Phys. 64 (1976), Nr. 2, p. 689–709<br />

[Benson and Krause 1980] Benson, B. B. ; Krause, D.: Isotopic fractionation of helium<br />

dur<strong>in</strong>g solution: A probe for the liquid state. In: J. Solution Chem. 9 (1980), Nr. 12,<br />

p. 895–909<br />

[Beyerle et al. 2000] Beyerle, U. ; Aeschbach-Hertig, W. ; Imboden, D. ; Baur, H. ;<br />

Graf, T. ; Kipfer, R.: A mass spectrometric system for the analysis of noble gases and<br />

tritium from water samples. In: Environ. Sci. Technol. 34 (2000), Nr. 10, p. 2042–2050<br />

[Beyerle et al. 2003] Beyerle, U. ; Rueedi, J. ; Leuenberger, M. ; Aeschbach-Hertig,<br />

W. ; Peeters, F. ; Kipfer, R. ; Dodo, A.: Evidence of periods of wetter and cooler climate<br />

<strong>in</strong> the Sahel between 6 and 40 kyr BP derived from groundwater. In: Geophys. Res. Lett. 30<br />

(2003), Nr. 4, p. 1173, doi:10.1029/2002GL016310<br />

[Bourg and Sposito 2008] Bourg, I. ; Sposito, G.: Isotopic fractionation of noble gases <strong>by</strong><br />

diffusion <strong>in</strong> liquid water: Molecular dynamics simulations and hydrologic applications. In:<br />

Geochim. Cosmochim. Acta 72 (2008), p. 2237–2247<br />

[Boynton and Compton 1944] Boynton, B. ; Compton, O. C.: Normal seasonal changes of<br />

oxygen and carbon dioxide percentages <strong>in</strong> gas from the larger pores of three orchard soils. In:<br />

Soil Science 57 (1944), p. 108–117<br />

[Brook et al. 1983] Brook, G. A. ; Folkloff, M. E. ; Box, E. O.: A world model of soil<br />

carbon dioxide. In: Earth Surface Processes and Landforms 8 (1983), Nr. 1, p. 79–88<br />

[Buyanovsky and Wagner 1983] Buyanovsky, G. A. ; Wagner, G. H.: Annual Cycles of<br />

Carbon Dioxide Level <strong>in</strong> Soil Air. In: Soil Science Society of America Journal 47 (1983),<br />

Nr. 6, p. 1139–1145<br />

[Castro et al. 2007] Castro, M. ; Hall, C. ; Patriarche, D. ; Goblet, P. ; Ellis, B.: A<br />

new noble gas paleoclimate record <strong>in</strong> Texas — Basic assumptions revisited. In: Earth Planet.<br />

Sci. Lett. 257 (2007), p. 170–187<br />

[Cussler 2005] Cussler, Edward L.: Diffusion. 2. ed., 7th pr<strong>in</strong>t. Cambridge : Cambridge<br />

University Press, 2005. – 580 p<br />

[Dowdell and Smith 1974] Dowdell, R. J. ; Smith, K. A.: Field studies of the soil atmosphere<br />

II. Occurrence of nitrous oxide. In: European Journal of Soil Science 25 (1974), Nr. 2, p. 231–<br />

238<br />

[Fang and Moncrieff 1999] Fang, C. ; Moncrieff, J.: A model for soil CO2 production and<br />

transport 1: Model development. In: Agricultural and Forest Meteorology 95 (1999), Nr. 4,<br />

p. 225–236<br />

[Fierer et al. 2003] Fierer, N. ; Schimel, J. P. ; Holden, P. A.: Variations <strong>in</strong> microbial<br />

community composition through two soil depth profiles. In: Soil Biology and Biochemistry<br />

35 (2003), Nr. 1, p. 167–176<br />

112


BIBLIOGRAPHY BIBLIOGRAPHY<br />

[Friedrich 2007] Friedrich, R.: Grundwassercharakterisierung mit Umwelttracern: Erkundung<br />

des Grundwassers der Odenwald-Region sowie Implementierung e<strong>in</strong>es neuen Edelgas-<br />

Massenspektrometersystems, Universität Heidelberg, Dissertation, 2007<br />

[Hall et al. 2005] Hall, C. ; Castro, M. ; Lohmann, K. ; Ma, L.: Noble gases and<br />

stable isotopes <strong>in</strong> a shallow aquifer <strong>in</strong> southern Michigan: Implications for noble gas paleotemperature<br />

reconstructions for cool climates. In: Geophys. Res. Lett. 32 (2005), p. L18404,<br />

doi:10.1029/2005GL023582<br />

[Hamada and Tanaka 2001] Hamada, Y. ; Tanaka, T.: Dynamics of carbon dioxide <strong>in</strong> soil<br />

profiles based on long-term field observation. In: Hydrological Processes 15 (2001), Nr. 10,<br />

p. 1829–1845. – ISSN 1099-1085<br />

[Heaton and Vogel 1981] Heaton, T. H. E. ; Vogel, J. C.: “Excess air” <strong>in</strong> groundwater. In:<br />

J. Hydrol. 50 (1981), p. 201–216<br />

[Hillel 1980] Hillel, D.: Fundamentals of soil physics. New York [u.a.] : Academic Pr., 1980<br />

[Holocher et al. 2002] Holocher, J. ; Peeters, F. ; Aeschbach-Hertig, W. ; Hofer,<br />

M. ; Brennwald, M. ; K<strong>in</strong>zelbach, W. ; Kipfer, R.: Experimental <strong>in</strong>vestigations on the<br />

formation of excess air <strong>in</strong> quasi-saturated porous media. In: Geochim. Cosmochim. Acta 66<br />

(2002), Nr. 23, p. 4103–4117<br />

[Hölt<strong>in</strong>g and Coldewey 2009] Hölt<strong>in</strong>g, B. ; Coldewey, W. G.: Hydrogeologie. 7., neu bearb.<br />

und erw. Aufl. Heidelberg : Spektrum Akad. Verl., 2009. – XXVIII, 383 S. p<br />

[Howard and Howard 1993] Howard, D. M. ; Howard, P. J. A.: Relationships between CO2<br />

evolution, moisture content and temperature for a range of soil types. In: Soil Biology and<br />

Biochemistry 25 (1993), Nr. 11, p. 1537 – 1546. – ISSN 0038-0717<br />

[Jähne et al. 1987] Jähne, B. ; He<strong>in</strong>z, G. ; Dietrich, W.: Measurement of the diffusion<br />

coefficients of spar<strong>in</strong>gly soluble gases <strong>in</strong> water. In: J. Geophys. Res. 92 (1987), Nr. C10,<br />

p. 10767–10776<br />

[Kendrick et al. 2006] Kendrick, M. ; Duncan, R. ; Phillips, D.: Noble gas and halogen<br />

constra<strong>in</strong>ts on m<strong>in</strong>eraliz<strong>in</strong>g fluids of metamorphic versus surficial orig<strong>in</strong>: Mt Isa, Australia.<br />

In: Chem. Geol. 235 (2006), p. 325–351<br />

[Kipfer et al. 2002] Kipfer, R. ; Aeschbach-Hertig, W. ; Peeters, F. ; Stute, M.:<br />

Noble gases <strong>in</strong> lakes and ground waters. In: Porcelli, D. (Ed.) ; Ballent<strong>in</strong>e, C. (Ed.) ;<br />

Wieler, R. (Ed.): Noble gases <strong>in</strong> geochemistry and cosmochemistry Bd. 47. Wash<strong>in</strong>gton,<br />

DC : M<strong>in</strong>eralogical Society of America, Geochemical Society, 2002, p. 615–700<br />

[Klump et al. 2008] Klump, S. ; Cirpka, O. ; Surbeck, H. ; Kipfer, R.: Experimental<br />

and numerical studies on excess-air formation <strong>in</strong> quasi-saturated porous media. In: Water<br />

Resour. Res. 44 (2008), p. W05402, doi:10.1029/2007WR006280<br />

[Klump et al. 2007] Klump, S. ; Tomonaga, Y. ; Kienzler, P. ; K<strong>in</strong>zelbach, W. ; Baumann,<br />

T. ; Imboden, D. ; Kipfer, R.: Field experiments yield new <strong>in</strong>sights <strong>in</strong>to gas exchange<br />

and excess air formation <strong>in</strong> natural porous media. In: Geochim. Cosmochim. Acta 71 (2007),<br />

p. 1385–1397<br />

113


BIBLIOGRAPHY BIBLIOGRAPHY<br />

[Liu et al. 2002] Liu, X. ; Wan, S. ; Su, B. ; Hui, D. ; Luo, Y.: Response of soil CO 2 efflux<br />

to water manipulation <strong>in</strong> a tallgrass prairie ecosystem. In: Plant and Soil 240 (2002), Nr. 2,<br />

p. 213–223<br />

[Ma et al. 2004] Ma, L. ; Castro, M. ; Hall, C.: A late Pleistocene–Holocene noble gas<br />

paleotemperature record <strong>in</strong> southern Michigan. In: Geophys. Res. Lett. 31 (2004), p. L23204,<br />

doi:10.1029/2004GL021766<br />

[Magnusson 1992] Magnusson, T.: Studies of the soil atmosphere and related physical site<br />

characteristics <strong>in</strong> m<strong>in</strong>eral forest soils. In: European Journal of Soil Science 43 (1992), Nr. 4,<br />

p. 767–790<br />

[Magnusson 1994] Magnusson, T.: Studies of the soil atmosphere and related physical<br />

characteristics <strong>in</strong> peat forest soils. In: Forest Ecology and Management 67 (1994), Nr. 1-3,<br />

p. 203–224<br />

[Mol<strong>in</strong>s and Mayer 2007] Mol<strong>in</strong>s, S. ; Mayer, K. U.: Coupl<strong>in</strong>g between geochemical reactions<br />

and multicomponent gas and solute transport <strong>in</strong> unsaturated media: A reactive transport<br />

model<strong>in</strong>g study. In: Water Resources Research 43 (2007), Nr. 5, p. W05435<br />

[Nieder and Benbi 2008] Nieder, R. ; Benbi, D. K.: Carbon and nitrogen <strong>in</strong> the terrestrial<br />

environment. [Dordrecht] : Spr<strong>in</strong>ger, 2008. – XI, 430 S. p. – ISBN 1-402-08432-3 ; 978-1-<br />

4020-8432-4 ; 978-1-4020-8433-1<br />

[von Oehsen 2008] Oehsen, A. von: Parameter Estimation and Model Validation for Models of<br />

dissolved Nobel Gas Concentrations <strong>in</strong> Groundwater, Ruprecht-Karls-UniversitÃt Heidelberg,<br />

<strong>Diploma</strong>rbeit, 2008<br />

[Orchard and Cook 1983] Orchard, V. A. ; Cook, F. J.: Relationship between soil respiration<br />

and soil moisture. In: Soil Biology and Biochemistry 15 (1983), Nr. 4, p. 447–453<br />

[Peeters et al. 2003] Peeters, F. ; Beyerle, U. ; Aeschbach-Hertig, W. ; Holocher,<br />

J. ; Brennwald, M. ; Kipfer, R.: Improv<strong>in</strong>g noble gas based paleoclimate reconstruction<br />

and groundwater dat<strong>in</strong>g us<strong>in</strong>g 20Ne/22Ne ratios. In: Geochim. Cosmochim. Acta 67 (2003),<br />

Nr. 4, p. 587–600<br />

[Porcelli et al. 2002] Porcelli, D. (Ed.) ; Ballent<strong>in</strong>e, C. (Ed.) ; Wieler, R. (Ed.): Rev.<br />

M<strong>in</strong>eral. Geochem.. Bd. 47: Noble gases <strong>in</strong> geochemistry and cosmochemistry. Wash<strong>in</strong>gton,<br />

DC : M<strong>in</strong>eralogical Society of America, Geochemical Society, 2002<br />

[Prunty and Bell 2005] Prunty, L. ; Bell, J.: Soil Temperature Change over Time dur<strong>in</strong>g<br />

Infiltration. In: Soil Science Society of America Journal 69 (2005), Jan, Nr. 3, p. 766<br />

[Ratkowsky et al. 1982] Ratkowsky, D. A. ; Olley, J. ; McMeek<strong>in</strong>, T. A. ; Ball, A.:<br />

Relationship between temperature and growth rate of bacterial cultures. In: Journal of<br />

Bacteriology 149 (1982), Nr. 1, p. 1<br />

[Richter and Großgebauer 1978] Richter, J. ; Großgebauer, A.: Untersuchungen zum Bodenlufthaushalt<br />

<strong>in</strong> e<strong>in</strong>em Bodenbearbeitungsversuch. 2. Gasdiffusionskoeffizienten als Strukturmaße<br />

für Böden. In: Zeitschrift für Pflanzenernährung und Bodenkunde 141 (1978),<br />

p. 181–202<br />

114


BIBLIOGRAPHY BIBLIOGRAPHY<br />

[Riveros-Iregui et al. 2007] Riveros-Iregui, D. A. ; Emanuel, R. E. ; Muth, D. J. ; Mcglynn,<br />

B. L. ; Epste<strong>in</strong>, H. E. ; Welsch, D. L. ; Pacific, V. J. ; Wraith, J. M.: Diurnal<br />

hysteresis between soil CO2 and soil temperature is controlled <strong>by</strong> soil water content. In:<br />

Geophys. Res. Lett. 34 (2007), Sep, Nr. 17, p. L17404<br />

[Riveros-Iregui et al. 2011] Riveros-Iregui, D. A. ; Mcglynn, B. L. ; Marshall, L. A. ;<br />

Welsch, D. L. ; Emanuel, R. E. ; Epste<strong>in</strong>, H. E.: A watershed-scale assessment of a process<br />

soil CO 2production and efflux model. In: Water Resour. Res. 47 (2011), May, p. W00J04<br />

[Rowell 1997] Rowell, D. L.: Bodenkunde. Berl<strong>in</strong> ; Heidelberg [u.a.] : Spr<strong>in</strong>ger, 1997. –<br />

XVIII, 614 S. p<br />

[Saito and ˇ Sim˚unek 2009] Saito, H. ; ˇ Sim˚unek, J.: Effects of meteorological models on<br />

the solution of the surface energy balance and soil temperature variations <strong>in</strong> bare soils. In:<br />

Journal of Hydrology 373 (2009), Jan, Nr. 3-4, p. 545–561<br />

[Scheffer and Schachtschabel 2010] Scheffer, F. ; Schachtschabel, P. ; Blume, H. (Ed.) ;<br />

Kandeler, E. (Ed.) ; Stahr, K. (Ed.): Lehrbuch der Bodenkunde. 16. Aufl. Heidelberg :<br />

Spektrum Akademischer Verlag, 2010<br />

[Schimel et al. 1999] Schimel, J. P. ; Gulledge, J. M. ; Cle<strong>in</strong>-Curley, J. S. ; L<strong>in</strong>dstrom,<br />

J. E. ; Braddock, J. F.: Moisture effects on microbial activity and community structure <strong>in</strong><br />

decompos<strong>in</strong>g birch litter <strong>in</strong> the Alaskan taiga. In: Soil Biology and Biochemistry 31 (1999),<br />

Nr. 6, p. 831–838<br />

[Schneider 2010] Schneider, T.: E<strong>in</strong>fluss von Sauerstoffzehrung auf Edelgaspartialdrücke <strong>in</strong><br />

Bodenluft, Universität Heidelberg, <strong>Diploma</strong>rbeit, 2010<br />

[Se<strong>in</strong>feld and Pandis 2006] Se<strong>in</strong>feld, J. H. ; Pandis, S. N.: Atmospheric chemistry and<br />

physics. 2. ed. Hoboken, NJ : Wiley, 2006<br />

[Sturchio et al. 2004] Sturchio, N. C. ; Du, X. ; Purtschert, R. ; Lehmann, B. E. ;<br />

Sultan, M. ; Patterson, L. J. ; Lu, Z. T. ; Müller, P. ; Bigler, T. ; Bailey, K. ;<br />

O’Connor, T. P. ; Young, L. ; Lorenzo, R. ; Becker, R. ; El Alfy, Z. ; El Kalioubly,<br />

B. ; Dawood, Y. ; Abdallah, A. M. A.: One million year old groundwater <strong>in</strong> the Sahara<br />

revealed <strong>by</strong> krypton-81 and chlor<strong>in</strong>e-36. In: Geophys. Res. Lett. 31 (2004), Nr. L05503,<br />

p. doi:10.1029/2003GL019234<br />

[Stute 1989] Stute, M.: Edelgase im Grundwasser - Bestimmung von PalÃotemperaturen und Untersuchung der Dynamik von GrundwasserflieÃsystemen, Universität Heidelberg, Dissertation,<br />

1989<br />

[Stute et al. 1995] Stute, M. ; Forster, M. ; Frischkorn, H. ; Serejo, A. ; Clark, J. ;<br />

Schlosser, P. ; Broecker, W. ; Bonani, G.: Cool<strong>in</strong>g of tropical Brazil (5 ◦ C) dur<strong>in</strong>g the<br />

Last Glacial Maximum. In: Science 269 (1995), p. 379–383<br />

[Stute and Schlosser 1993] Stute, M. ; Schlosser, P.: Pr<strong>in</strong>ciples and applications of the<br />

noble gas paleothermometer. In: Swart, P. K. (Ed.) ; Lohmann, K. C. (Ed.) ; McKenzie, J.<br />

(Ed.) ; Sav<strong>in</strong>, S. (Ed.): Climate Change <strong>in</strong> Cont<strong>in</strong>ental Isotopic Records Bd. 78. Wash<strong>in</strong>gton,<br />

DC : American Geophysical Union, 1993, p. 89–100<br />

115


BIBLIOGRAPHY BIBLIOGRAPHY<br />

[Stute and Sonntag 1992] Stute, M. ; Sonntag, C.: Paleotemperatures derived from noble<br />

gases dissolved <strong>in</strong> groundwater and <strong>in</strong> relation to soil temperature, 1992, p. 111–122<br />

[Suarez and ˇ Sim˚unek 1993] Suarez, D. L. ; ˇ Sim˚unek, J.: Model<strong>in</strong>g of carbon dioxide<br />

transport and production <strong>in</strong> Soil 2. Parameter selection, sensitivity analysis, and comparison<br />

of model predictions to field data. In: Water Resources Research 29 (1993), Nr. 2, p. 499–513<br />

[Sun et al. 2008] Sun, T. ; Hall, C. ; Castro, M. ; Lohmann, K. ; Goblet, P.: Excess air<br />

<strong>in</strong> the noble gas groundwater paleothermometer: A new model based on diffusion <strong>in</strong> the gas<br />

phase. In: Geophys. Res. Lett. 35 (2008), p. L19401, doi:10.1029/2008GL035018<br />

[Sun et al. 2010] Sun, T. ; Hall, C. M. ; Castro, M. C.: Statistical properties of groundwater<br />

noble gas paleoclimate models: Are they robust and unbiased estimators? In: Geochem.<br />

Geophys. Geosyst. 11 (2010), Feb, Nr. 2, p. Q02002<br />

[Tang et al. 2005] Tang, J. ; Misson, L. ; Gershenson, A. ; Cheng, W. ; Goldste<strong>in</strong>,<br />

A. H.: Cont<strong>in</strong>uous measurements of soil respiration with and without roots <strong>in</strong> a ponderosa<br />

p<strong>in</strong>e plantation <strong>in</strong> the Sierra Nevada Mounta<strong>in</strong>s. In: Agricultural and Forest Meteorology 132<br />

(2005), Nr. 3-4, p. 212–227<br />

[Welsch and Hornberger 2004] Welsch, D. L. ; Hornberger, G. M.: Spatial and temporal<br />

simulation of soil CO2 concentrations <strong>in</strong> a small forested catchment <strong>in</strong> Virg<strong>in</strong>ia. In:<br />

Biogeochemistry 71 (2004), Nr. 3, p. 413–434<br />

[Wieser 2006] Wieser, M.: Entwicklung und Anwendung von Diffusionssamplern zur<br />

Beprobung gelöster Edelgase <strong>in</strong> Wasser, Universität Heidelberg, <strong>Diploma</strong>rbeit, 2006<br />

[Wieser 2011] Wieser, M.: Impr<strong>in</strong>ts of climatic and environmental change <strong>in</strong> a regional<br />

aquifer system <strong>in</strong> an arid part of India us<strong>in</strong>g noble gases and other environmental tracers,<br />

Universität Heidelberg, Dissertation, 2011<br />

[W<strong>in</strong>kler et al. 1996] W<strong>in</strong>kler, J. P. ; Cherry, R. S. ; Schles<strong>in</strong>ger, W. H.: The Q10 relationship<br />

of microbial respiration <strong>in</strong> a temperate forest soil. In: Soil Biology and Biochemistry<br />

28 (1996), Nr. 8, p. 1067–1072<br />

[WMO 2008] WMO, World Meteorological Organization: Guide to Meteorological Instruments<br />

and Methods of Observation. Bd. WMO No. 8. Chap. 4, p. 29. Geneva, 2008<br />

[Yamaguchi et al. 1967] Yamaguchi, M. ; Flocker, W. J. ; Howard, F. D.: Soil Atmosphere<br />

as Influenced <strong>by</strong> Temperature and Moisture. In: Soil Sci. Soc. Am. J. 31 (1967), Nr. 2, p. 164–<br />

167<br />

[Yuste et al. 2003] Yuste, J. C. ; Janssens, I. A. ; Carrara, A. ; Meiresonne, L. ;<br />

Ceulemans, R.: Interactive effects of temperature and precipitation on soil respiration <strong>in</strong> a<br />

temperate maritime p<strong>in</strong>e forest. In: Tree Physiology 23 (2003), Nr. 18, p. 1263<br />

116


Acknowledegment<br />

At the end of this <strong>thesis</strong> a few words of thanks are due: to all the people who helped me get<br />

to this po<strong>in</strong>t, both personally as well as professionally. First of all I would like to thank my<br />

parents for support<strong>in</strong>g me through my entire time at the Heidelberg University, through all the<br />

good and all the difficult times.<br />

Of course I would also like to thank my supervisor, Professor Dr. Werner Aeschbach-Hertig, for<br />

giv<strong>in</strong>g me the chance to work on this <strong>thesis</strong> and for lett<strong>in</strong>g me handle it quite <strong>in</strong>dependently<br />

while still provid<strong>in</strong>g valuable <strong>in</strong>put and guidance. For all the <strong>in</strong>put and shared knowledge, as<br />

well as the great help at construct<strong>in</strong>g the sampl<strong>in</strong>g sites I would like to thank my friend and<br />

colleague Tim Schneider. In fact, the entire team was extremely k<strong>in</strong>d and helpful with the many<br />

problems and questions that arose. Primarily, but not exclusively, I would like to thank Mart<strong>in</strong><br />

Wieser, Tillmann Kaudse, Lisa Broeder and for all their help and the extensive proofread<strong>in</strong>g<br />

and generally for the great time.<br />

This study would also not have been possible without the k<strong>in</strong>d assistance of Kurt Elfner, who<br />

provided a small piece of his flower nursery’s land where I was able to set up one of my sampl<strong>in</strong>g<br />

sites. Also very helpful was the company meteomedia, which made their weather data archive<br />

of their nearest weather station available to me.<br />

And of course there were the <strong>in</strong>habitants of office 442: probably hold<strong>in</strong>g both the <strong>in</strong>stitute’s<br />

attendance record on sundays and holidays (which might be a bit ambivalent) and always enterta<strong>in</strong><strong>in</strong>g<br />

visitors (which was a lot of fun as well as <strong>in</strong>sightful). Claudia, Johannes, Joelle and<br />

Fiete, you are the best!<br />

A special and huge thanks goes to my good friend Julia Schaper who helped me a lot with some<br />

great, extensive and always encourag<strong>in</strong>gly funny proofread<strong>in</strong>g and to T<strong>in</strong>a Straße, who somehow<br />

always managed to motivate me and to get my spirit up. Your visits meant a lot to me! I would<br />

not have made it this far without the support of my girlfriend Katja Weiß, who was always there<br />

for me (even when I did not have time for her). Thank you!<br />

117


Erklärung:<br />

Ich versichere, dass ich diese Arbeit selbstständig verfasst und ke<strong>in</strong>e anderen als die angegebenen<br />

Quellen und Hilfsmittel benutzt habe.<br />

Heidelberg, 30.06.2011<br />

.......................................<br />

(Unterschrift)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!