20.12.2013 Views

Karin.Grust - British Hydrological Society

Karin.Grust - British Hydrological Society

Karin.Grust - British Hydrological Society

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.

doi: 10.7558/bhs.2012.ns24<br />

UK trial of the OTT Pluvio 2<br />

<strong>Karin</strong> <strong>Grust</strong> 1* and Dave Stewart 2<br />

1<br />

Scottish Environment Protection Agency; 2 Environment Agency<br />

*Email: <strong>Karin</strong>.<strong>Grust</strong>@SEPA.org.uk<br />

Abstract<br />

The OTT Pluvio 2 is a gravimetric raingauge which works according to the weighing balance<br />

principle. This paper reports on trails of this equipment carried out by The Environment Agency and<br />

SEPA, in cooperation with the Met Office, to assess performance, accuracy and reliability. Detailed<br />

comparisons and analyses for both trials against other co-located raingauges are presented.<br />

Introduction<br />

The OTT Pluvio 2 is a gravimetric raingauge. It works<br />

according to the weighing balance principle. Digital outputs<br />

(impulse/0.1 mm and frequency/status) and a serial interface,<br />

freely configurable as SDI-12 or RS-485, are available for<br />

transferring the data.<br />

The Scottish Environment Protection Agency<br />

(SEPA), in cooperation with the Met Office, has undertaken<br />

a Pluvio 2 trial at Eskdalemuir in the Scottish Borders. The<br />

Environment Agency (EA) has carried out a trial of this<br />

equipment at Tryweryn in North Wales. OTT Hydrometry has<br />

supported both trials, providing support and installation of a<br />

Pluvio 2 with a 200 cm 2 rim area and a 1500 mm m -2 collection<br />

bucket. All Pluvio 2 data used for these investigations are data<br />

collected from a Pluvio 2 without a rim heater.<br />

Purpose of the Pluvio 2 trials<br />

The purpose of the Pluvio 2 trials is:<br />

• to compare performance against existing technology;<br />

• to test the equipment as a potential low maintenance<br />

raingauge in remote areas that will allow reduced frequency<br />

of site visits compared to existing practice for tipping<br />

bucket raingauges (TBRs);<br />

• to investigate the instrument accuracy and reliability to<br />

allow its operation without the need for a comparison gauge<br />

e.g. a monthly check gauge;<br />

• to assess its performance in measurement of solid<br />

precipitation.<br />

Pluvio 2 performance criteria<br />

An observer-read daily raingauge was the reference for the<br />

daily rainfall total. The Pluvio 2 performance is measured by<br />

its deviation to the observer-read daily gauge over a period of<br />

one month and should not exceed 10% (SEPA) and 8% (EA).<br />

These data were quality controlled by the Met Office. The<br />

observer-read data were replaced by the Met Office quality<br />

controlled data for comparison when it became available.<br />

For the SEPA trial, the Pluvio 2 performance is also<br />

measured by its daily and monthly performance comparison<br />

to other automated raingauges which are four different,<br />

commonly used TBRs installed at the test site within close<br />

proximity of the Pluvio 2 .<br />

The EA 8% comparison is further refined as follows:<br />

• Rainfall > 25 mm in period: within 8%<br />

• Rainfall < 25 mm in period: within 2 mm<br />

Note 1: the following condition applies: there has not been<br />

any snow during the check period.<br />

Note 2: The ‘period’ is currently one month at Tryweryn.<br />

The trial will be considered a success if the Pluvio 2<br />

• performs within 8% (EA) or 10% (SEPA) of the observer<br />

read daily raingauge;<br />

• meets or betters the TBR performance criteria over the<br />

period of the trial;<br />

• requires less maintenance than the TBR in order to provide<br />

confidence that it will extend the interval between site visits<br />

required compared to TBRs;<br />

• data can be successfully collected via logger;<br />

• data can be successfully collected from site via existing<br />

telemetry infrastructure;<br />

• data can be successfully saved into the existing data archive<br />

(WISKI 6 and WISKI7) from either loggers or telemetry.<br />

Test sites<br />

Eskdalemuir<br />

The SEPA test site is located at Eskdalemuir Observatory in<br />

the Southern Uplands of Scotland, Dumfries and Galloway<br />

at an altitude of 242 metres (NGR: NT 235 027). The site is<br />

operated by the Met Office and the <strong>British</strong> Geological Survey<br />

(BGS). Eskdalemuir Observatory is a meteorological station<br />

and involved with atmospheric pollution measurements<br />

as well as seismic and magnetic observations. The site is<br />

representative for the climate of inhabited areas of high<br />

ground in the northern United Kingdom.<br />

A Pluvio 2 was installed in a pit of about 50 cm depth.<br />

With a total instrument height of about 100 cm, the collection<br />

rim of the Pluvio 2 was about 50 cm above the surface. For<br />

the evaluation of the rainfall data quality from the Pluvio 2 ,<br />

a daily raingauge (read by a Met Office observer at 9:00<br />

UTC) was installed in close proximity to the Pluvio 2 . This<br />

gauge was used as reference check gauge. Furthermore, in<br />

the same area of the test site four different types of tipping<br />

bucket raingauges were installed (ELE, Casella, Rimco and<br />

SBS 500). The daily reference gauge, as well as the TBRs,<br />

were installed on the ground surface with a rim height of<br />

approximately 30 cm. This is the general installation height<br />

in the SEPA raingauge network. The comparison data in<br />

this paper include those from the daily reference gauge, the<br />

BHS Eleventh National Symposium, Hydrology for a changing world, Dundee 2012. ISBN: 1903741181<br />

© <strong>British</strong> <strong>Hydrological</strong> <strong>Society</strong><br />

1


Pluvio 2 weighing raingauge and the four different types of<br />

TBRs. All TBRs were calibrated at a single simulated rainfall<br />

intensity of about 14 to 21 mm hour -1 .<br />

Data from all automatic raingauges were recorded by<br />

a D7141 Dynamic Logic data logger at 15-minute intervals.<br />

The Pluvio 2 was configured for non real time logging. For the<br />

data analysis in this paper, daily and monthly totals as well<br />

as 15-minute data from a heavy rainfall event in May 2011<br />

have been used. On-site air temperature, wind speed and wind<br />

direction recordings were considered during the data analyses.<br />

The Pluvio 2 trial at Eskdalemuir started in November<br />

2009. At the beginning of the trial the Pluvio 2 was undermeasuring<br />

rainfall by 5 to 10% compared with the daily read<br />

reference gauge. With its collection rim being higher off the<br />

ground than the rim of the daily gauge, it was thought that<br />

wind was responsible for the discrepancies. To reduce the<br />

wind effect on the Pluvio 2 , a wind shield was installed in<br />

November 2010. However, the performance of the Pluvio 2 did<br />

not improve. Significant performance improvement occurred<br />

with the Pluvio2 firmware upgrade (to version 1.27.2) on the<br />

16th May 2011. Therefore, in this paper the data analysis is<br />

based on the data collected from the 17th May 2011 onwards.<br />

Tryweryn<br />

The EA test site is located at Tryweryn Dam in North Wales at<br />

an altitude of 255 metres (NGR: SH 881 399). The automatic<br />

raingauges are operated by EA staff and the daily gauge is<br />

read by a Welsh Water member of staff.<br />

The equipment on site consists of<br />

• Fenced compound<br />

• Instrument house<br />

• Manually read daily storage gauge<br />

• Casella TBR, 2 mm tip, 400 cm 2 rim area, providing time<br />

of tip data, connected to a Dynamic Logic 700 telemetry<br />

outstation<br />

• Casella Tipping Bucket Raingauge, 0.2 mm tip, 400 cm 2<br />

rim area, providing time of tip data, connected to a data<br />

logger. These backup instrument data were not used<br />

for comparison. Both TBRs were calibrated at a single<br />

simulated rainfall intensity of between 10 and 15 mm hour -1<br />

• Pluvio 2 , 200 cm 2 rim area, unheated rim, with rim 1 metre<br />

above ground level. This provides rainfall data at 1 minute<br />

intervals. Firmware version was version 1.27.2 throughout<br />

• DuoSens logger system with solar panel / battery power<br />

supply and enclosure<br />

panel and battery combination is sufficient to provide power<br />

throughout the winter periods.<br />

Data analysis<br />

Monthly data analysis<br />

Monthly rainfall totals from the Pluvio 2 at Eskdalemuir<br />

and Tryweryn in comparison with a daily read reference<br />

gauge and one or more different types of TBRs (Figure 1)<br />

show a good performance. Monthly totals at Eskdalemuir<br />

from December, January and February are influenced by<br />

snowfall, resulting in suspect data from automatic raingauges.<br />

Therefore, precipitation data from these months have<br />

been excluded in the monthly performance evaluation.<br />

Monthly rainfall totals of the Pluvio 2 are predominantly<br />

slightly lower than the totals from the daily reference gauge.<br />

The monthly %-difference between the daily<br />

reference gauge and automatic raingauges at Eskdalemuir<br />

and at Tryweryn (Tables 1 and 2) provide further information<br />

about the performance of the Pluvio 2 . Compared with the<br />

reference gauge, all automatic raingauges at Eskdalemuir,<br />

including the Pluvio 2 fulfil the success criteria of 10% (SEPA)<br />

and 8% (EA) during months without snowfall (Table 1) (see<br />

Trial success criteria). On average, all automatic raingauges<br />

were under-catching monthly rainfall totals. The Pluvio 2 and<br />

three of the four monthly gauges achieved results within<br />

–2% of the observer’s readings. The ELE was under-catching<br />

monthly rainfall totals by 5%. However, examining the<br />

amplitude of variation, the ELE provides the most consistent<br />

monthly measurements with an amplitude of variation of<br />

3.2 mm; this is followed by the Pluvio2 with an amplitude<br />

of variation of 5.7 mm from the observers readings over<br />

seven months, as shown in Table 1. The reason for better<br />

Monthly rainfall totals [mm]<br />

350<br />

300<br />

250<br />

200<br />

150<br />

100<br />

50<br />

0<br />

ELE Casella<br />

Rimco SBS 500<br />

Pluvio2<br />

SEPA daily<br />

Data accuracy<br />

affected by snowfall<br />

In February 2012 it was discovered that a problem with the<br />

telemetry outstation configuration rendered a significant<br />

proportion of the data unusable. Data from the back up<br />

DouSens logger was examined and it too was found to<br />

have problems, possibly due to a dirty contact in the cable<br />

connection. It also suffered from power supply problems.<br />

All data from the summer of 2011 have been lost. Data are<br />

available for the following periods from the DouSens logger<br />

as a pulse output from the Pluvio 2 :<br />

• 6 January 2011 – 16 April 2011<br />

• 2 November 2011 – ongoing. This period includes missing<br />

data for the 7 – 10 January 2012.<br />

Met Office quality controlled daily observer data<br />

have been used up to 16 April 2011. Raw data have been<br />

used for the later period but will be substituted with quality<br />

controlled data when this becomes available.<br />

There was significant data loss from the DuoSens<br />

logger during November and December 2010 and four days<br />

were lost in January 2012. This raises a doubt that the solar<br />

2<br />

Monthly rainfall totals [mm]<br />

350<br />

300<br />

250<br />

200<br />

150<br />

100<br />

50<br />

0<br />

Figure 1<br />

Jun-11<br />

Jan-11<br />

Feb-11<br />

Jul-11<br />

Mar-11<br />

Aug-11<br />

Apr-11<br />

May-11<br />

Sep-11<br />

Jun-11<br />

Oct-11<br />

Jul-11<br />

Nov-11<br />

Dec-11<br />

missing data<br />

Aug-11<br />

Sep-11<br />

Oct-11<br />

Jan-12<br />

Casella<br />

Pluvio2<br />

EA daily<br />

Nov-11<br />

Dec-11<br />

Feb-12<br />

Jan-12<br />

Monthly rainfall totals of the Pluvio2 in comparison with a daily<br />

read reference gauge and one or more different types of TBRs<br />

(ELE, Rimco, Casella and SBS 500) at Eskdalemuir (top) and<br />

Tryweryn (bottom).<br />

Mar-12<br />

Feb-12


Table 1<br />

Monthly difference [%] between SEPA daily reference gauge and automatic raingauges (snow free months only).<br />

Greatest and lowest differences of each raingauge are highlighted.<br />

Eskdalemuir ELE [%] Casella [%] Rimco [%] SBS 500 [%] Pluvio2 [%]<br />

June 2011 -3.7 3.4 4.3 2.0 0.2<br />

July 2011 -5.9 -1.3 -0.5 -1.65 -1.6<br />

August 2012 -3.2 1.6 1.7 0.5 -0.2<br />

September 2011 -3.6 -2.2 -1.9 -1.8 0.2<br />

October 2011 -6.4 -6.2 -3.4 -5.1 -1.0<br />

November 2011 -6.2 -6.1 -4.8 -5.1 -5.5<br />

March 2012 -6.3 -3.4 3.4 -0.2 -5.6<br />

Average difference [%] -5.0 -2.0 0.2 -1.6 -1.9<br />

Amplitude of variation (mm) 3.2 9.6 9.1 7.2 5.7<br />

Table 2<br />

Monthly difference [%] between EA daily reference gauge and automatic raingauge.<br />

Greatest and lowest differences of each raingauge are highlighted.<br />

Casella [%] Pluvio 2 [%] Comments<br />

January 2011 -12.0 -0.1 Data from 6th onwards<br />

February 2011 -12.7 -1.2<br />

March 2011 -11.5 -14.2*<br />

April 2011 4.3 14.5* Data from 1st to 16th<br />

November 2011 -0.2 0.0 Data from 2nd onwards<br />

December 2011 -4.2 -2.6<br />

January 2012 -6.6 -3.4<br />

February 2012 -9.4 -4.3<br />

Average difference [%] -6.5 -1.4<br />

Amplitude of difference variation [mm] 17.0 28.7<br />

* Difference may be due to observer error at the end of March and beginning of April.<br />

performance consistency of the ELE and Pluvio 2 raingauges<br />

could be performance differences at varying rainfall<br />

intensities. Analyses of raingauge performances at different<br />

intensities are required to examine this further.<br />

Daily data analysis<br />

Cumulative daily rainfall totals The daily performance of<br />

the Pluvio 2 at Eskdalemuir and Tryweryn using cumulative<br />

daily rainfall totals in comparison with other raingauges is<br />

shown in Figure 2. Cumulative daily rainfall totals measured<br />

by different types of raingauges at Eskdalemuir cover 301<br />

snow free days from the 17th May 2011 to the 31st March<br />

2012 (Figure 2a). Periods of suspect data during winter due to<br />

snow precipitation have been excluded in this analysis. The<br />

measured daily rainfall totals from Pluvio 2 , Casella, Rimco<br />

and SBS 500 raingauges show similar and consistent results<br />

for the observation period. The cumulative rainfall total of the<br />

SEPA daily reference gauge from the 17th May 2011 to the<br />

31st March 2012 was 1670.7 mm. Automatic raingauges were<br />

under-measuring rainfall totals by about 3 to 6% (Table 3).<br />

The performance of the Pluvio 2 with an under-catch of 3.4%<br />

was very satisfying.<br />

Cumulative daily rainfall totals measured at<br />

Tryweryn cover 235 days from the 6th November 2010<br />

to the 29th February 2012 (Figure 2b). This period was<br />

characterised by an intense rainfall event in February 2011<br />

(data from the 17th April 2011 to 2nd November 2011 are<br />

missing). The TBR significantly under-measured during<br />

this event. Before and after this event all three gauges show<br />

excellent comparison.<br />

Correlation analysis The correlation analysis of daily<br />

rainfall totals of the Pluvio 2 gauges at Eskdalemuir and<br />

Tryweryn with the on-site daily reference gauge shows strong<br />

correlations of 99.61% and 98.69% respectively. A strong<br />

correlation between raingauges installed in close proximity<br />

was expected and the correlation of the Pluvio 2 with the<br />

3<br />

daily reference gauge is better than the correlation of TBRs<br />

with the daily reference gauge. There are two outliers in the<br />

Treweryn data. On both of these occasions the Pluvio 2 did not<br />

record rainfall measured in both the daily raingauge and TBR.<br />

The missing data at the site (see Figure 2b) were caused by<br />

a ‘dirty’ data lead connection. One of the affected days was<br />

close to the end of the first set of data and the other soon after<br />

data restarted. It is possible that the unrecorded data by the<br />

Pluvio 2 on both of these days was a manifestation of the more<br />

serious problems during the period of missing data.<br />

Sub daily data analysis<br />

The significant two-day rainfall event at Tryweryn (Figure 3<br />

and Table 4) had a maximum rainfall intensity of only around<br />

5 mm hour - 1 but was accompanied by gale force winds. It<br />

might be expected that the higher rim height of the Pluvio 2<br />

would have resulted in it under-catching. In fact it shows<br />

excellent comparison with the daily gauge total for the same<br />

period. The TBR shows significant under-catch against both<br />

of the other raingauges. During a later high rainfall intensity<br />

event (10 mm hour -1 ) on 2nd January 2012, without strong<br />

winds, all three raingauges provided very close rainfall totals<br />

of around 45 mm in 24 hours.<br />

During higher intensity rainfall events<br />

(> 5 mm hour -1 ) at Eskdalemuir the cumulative rainfall<br />

recordings from the Pluvio 2 and the TBRs at 15-mins intervals<br />

are very similar, with a trend of the Pluvio 2 recording slightly<br />

higher event totals than the TBRs (Table 4). It is not clear<br />

whether the Pluvio 2 or the TBR readings during this event<br />

are more accurate. The Eskdalemuir heavy rainfall event<br />

on the 23rd May 2011 happened from 03:00 to 10:30 and<br />

occurred therefore within two water days. It appears that the<br />

windy conditions on site with recorded wind speeds of about<br />

24 miles hour -1 during the event have had no negative impact<br />

on the accuracy of the Pluvio 2 data. The two-daily rainfall<br />

totals shown within brackets in Table 4 provide the totals<br />

from all test gauges for comparison. Interestingly, the twodaily<br />

total for the Pluvio 2 is slightly lower than that of most


a)<br />

Cumulative daily rainfall [mm]<br />

1800<br />

1600<br />

1400<br />

1200<br />

1000<br />

800<br />

600<br />

400<br />

ELE<br />

Casella<br />

Rimco<br />

SBS 500<br />

Pluvio2<br />

SEPA daily<br />

200<br />

b)<br />

0<br />

17 May 06 Jul 25 Aug 14 Oct 03 Dec 22 Jan 12 Mar<br />

1600<br />

Cumulative daily rainfall [mm]<br />

1400<br />

1200<br />

1000<br />

800<br />

600<br />

400<br />

Pluvio2<br />

Casella<br />

EA daily<br />

missing data<br />

200<br />

0<br />

06 Nov 26 Dec 14 Feb 05 Apr 25 May 14 Jul 02 Sep 22 Oct 11 Dec 30 Jan<br />

Figure 2<br />

Cumulative daily rainfall totals from different types of raingauges at Eskdalemuir (top) and at Tryweryn (bottom)<br />

Cumulative rainfall totals [mm]<br />

120<br />

Pluvio2<br />

100<br />

Casella<br />

80<br />

60<br />

40<br />

20<br />

0<br />

00:00 12:00 00:00 12:00 00:00 12:00<br />

Figure 3 Cumulative rainfall totals during heavy rainfall events recorded at 15 Minute intervals at Tryweryn on the 5th and 6th February 2011.<br />

4


Table 3<br />

Cumulative rainfall totals from different raingauges at Eskdalemuir from 17th May 2011 to 31st March 2012 and at Tryweryn from the 6th<br />

November 2010 to 29th February 2012 (some missing data) and their differences to a daily reference gauge.<br />

Es k d a l e m u i r<br />

Tr y w e r y n<br />

Raingauge SEPA ELE Casella Rimco SBS 500 Pluvio 2 EA Casella Pluvio 2<br />

reference<br />

reference<br />

Rainfall total [mm] 1670.7 1569.0 1611.4 1622.0 1614.4 1614.3 1372.8 1194.8 1272.4<br />

Difference [mm] - 101.7 59.3 48.7 56.3 56.4 - 103.4 21.6<br />

Difference [%] - 6.1 3.5 2.9 3.4 3.4 - -7.5 -1.6<br />

Table 4<br />

Cumulative rainfall totals of heavy rainfall event Tryweryn on<br />

the 5th and 6th February 2011(48 hour event) and Eskdalemuir<br />

on the 22nd and 23rd May 2011 (7.5 hour event).<br />

Cumulative rainfall Cumulative rainfall<br />

totals [mm]<br />

totals [mm]<br />

05 + 06 Feb 2011 23 May 2011<br />

(Tryweryn, 48 hours) (Eskdalemuir, 7.5hours)<br />

Pluvio 2 109.7 37.1 (50.3*)<br />

Casella 95.2 35.8 (50.6*)<br />

ELE - 35.4 (49.2*)<br />

Rimco - 36.0 (51.2*)<br />

SBS 500 - 36.6 (51.2*)<br />

Two-daily total<br />

(reference gauge) 111.4 53.5<br />

* Two-daily total [mm]<br />

other automatic raingauges.<br />

The impact of the strong winds on the TBR<br />

performance in the February 2011 event at Tryweryn may<br />

be site specific. Results from both trials suggest that further<br />

observations of high intensity rainfall events and events with<br />

high winds would be useful for the further evaluation of the<br />

Pluvio 2 with regard to its recording accuracy.<br />

Comparison of snow data<br />

The performance of raingauges during snowfall events has<br />

been analysed at Eskdalemuir. A snowfall event in December<br />

2011 (Figure 4 and Table 5) is representative for the overall<br />

performance of the Pluvio 2 and TBRs in comparison.<br />

On the 4th December 2011, 15 cm of snow were<br />

ELE<br />

Casella<br />

Rimco SBS 500<br />

Pluvio2 SEPA daily<br />

Recorded daily precipitation totals [mm]<br />

60<br />

50<br />

40<br />

30<br />

20<br />

10<br />

0<br />

Temp<br />

04 Dec 05 Dec 06 Dec 07 Dec 08 Dec 09 Dec 10 Dec<br />

10<br />

5<br />

0<br />

-5<br />

-10<br />

-15<br />

-20<br />

Temperature [°C]<br />

Figure 4<br />

Daily precipitation measured by different raingauges at Eskdalemuir and hourly<br />

temperatures measured at Eskdalemuir from 4th December to 10th December 2011.<br />

Note: SEPA daily measurement on the 6th December is a 48 hour cumulative snow water<br />

equivalent for the 5th and 6th December.<br />

Table 5<br />

Daily precipitation recordings from different raingauges during a period of snowfall, snowmelt and rainfall in December 2011 at<br />

Eskdalemuir<br />

Date ELE Casella Rimco SBS 500 Pluvio2 SEPA daily Snow [cm]**<br />

04 Dec 0.8 0.8 0.8 0.6 12.5 11.2 11.0<br />

05 Dec 2.6 0.6 1.6 0.2 7.0 missing 16.0<br />

06 Dec 12.4 15.0 18.4 0.4 8.0 18.0* 11.0<br />

07 Dec 35.8 43.0 31.8 28.8 26.7 27.5 7.0<br />

08 Dec 10.0 10.0 10.6 10.4 9.8 11.3 -<br />

09 Dec 0.2 0.2 0.2 0.2 1.6 missing -<br />

10 Dec 10.2 10.0 9.6 10.0 9.5 10.7* -<br />

Precipitation total [mm] 72 80 73 50.6 75.1 78.7<br />

* Cumulated total over 2 days<br />

**Automatic snow depth recording at Eskdalemuir at 5:50 UTC (snow depth value displayed within it’s water day i.e. the value measured<br />

on the 5th Dec at 5:50 UTC belongs meteorologically to the 4th Dec).<br />

5


ecorded at Eskdalemuir: this translated into a snow water<br />

equivalent of 11.2 mm measured from the SEPA daily<br />

reference gauge. With 12.5 mm the Pluvio 2 recorded a similar<br />

snow water equivalent as the daily gauge whereas all TBRs<br />

were snow blocked and failed to measure this precipitation<br />

event. With temperatures having dropped below 0 °C over<br />

night, TBRs remained blocked or recorded precipitation too<br />

low. The accuracy of the Pluvio 2 snow water measurement is<br />

unknown as no SEPA daily snow water equivalent is available<br />

for the 5th December. The 48-hour totals of the SEPA daily<br />

gauge (18 mm) and the Pluvio 2 (15 mm) on the 6th December<br />

suggests that the Pluvio 2 is slightly under-catching. The<br />

second half of the 6th December and the 7th December were<br />

characterised by a temperature increase, some precipitation<br />

and snow melt occurring in the TBRs. This resulted in overrecording<br />

of precipitation by the TBRs on the 6th and 7th<br />

December. Further precipitation on the 8th December appears<br />

to be recorded correctly by all raingauges.<br />

Results<br />

Results from Eskdalemuir<br />

The results from the Eskdalemuir Pluvio 2 raingauge trial are<br />

summarised as follows:<br />

• The monthly performance of the Pluvio 2 fulfil the trial<br />

criteria of 10% and better (on average under-catching<br />

rainfall by 1.9% per month)<br />

• Daily rainfall totals measured by the Pluvio 2 were (on<br />

average) under-catching by 3.4%. The TBR performance at<br />

Eskdalemuir was similar to the performance of the Pluvio 2 .<br />

• A strong correlation exists between daily rainfall totals<br />

measured by the Pluvio 2 and the daily reference gauge.<br />

With no outliers, the correlation of the Pluvio 2 is slightly<br />

better than the correlation of the TBRs.<br />

• The most consistent monthly measurements were achieved<br />

by the ELE TBR and the Pluvio 2 .<br />

• The Pluvio 2 is able to measure snow precipitation.<br />

Values appear reliable, but not as accurate as for rainfall<br />

precipitation values. The Pluvio 2 outperforms all TBR<br />

gauges during snowfall events.<br />

Results from Tryweryn<br />

• The Pluvio 2 displays very close comparison against the<br />

daily gauge.<br />

• The TBR under catch of 7.5% may be due it being<br />

calibrated at single rainfall intensity of around 10–15<br />

mm hour -1 rather than across a range of intensities, as<br />

recommended in recent laboratory and field trails of<br />

automatic raingauges in the labortatory (Lanza et al., 2006)<br />

and in the field (Lanza and Vuerlich, 2009).<br />

• A good correlation exists for the Pluvio 2 compared with<br />

the TBR but the trend line gradient again shows the TBR<br />

under-catch.<br />

• The Pluvio 2 has passed the trial success criteria, but with<br />

concern for the two days of very poor performance.<br />

• The Pluvio 2 performed very well in comparison to the daily<br />

gauge during the event at Tryweryn on the 5th and 6th<br />

February. The TBR significantly under-measured this event.<br />

Conclusions<br />

• The Pluvio 2 outperforms TBRs when compared to the<br />

storage daily reference raingauge.<br />

• The Pluvio 2 raingauge compares very well with the TBR<br />

during a high rainfall intensity storm, albeit based upon<br />

examination of a limited number of events only.<br />

• Where a solar panel / battery power is used then this must<br />

be sufficient to last through the winter without power loss.<br />

• The use of a Pluvio 2 can provide an opportunity for time<br />

saving compared to a TBR through reduced site visits and<br />

calibration effort.<br />

• Calibration checks for the Pluvio 2 are easily completed<br />

in the field and take considerably less time to undertake<br />

compared to a TBR; 5 minutes in the field compared to<br />

around 2 hours for a TBR.<br />

• Pluvio 2 data can be successfully collected from sites via<br />

existing telemetry infrastructure.<br />

• Pluvio 2 data can be successfully saved into the existing<br />

data archive (WISKI 6 and WISKI7) from either loggers or<br />

telemetry.<br />

• The windshield installed temporarily over 12 months at<br />

Eskdalemuir did not appear to improve data collection. This<br />

may be site specific.<br />

Recommendations<br />

• Due to the relatively high cost of the Pluvio 2 it should be<br />

deployed where it provides maximum benefits, e.g. remote<br />

locations, logger only sites or where dual TBRs might be<br />

considered.<br />

• The Pluvio 2 rainfall catching orifice is 100 cm high —<br />

considerably higher than a regular TBR or storage gauge.<br />

Although the results from these trials do not indicate any<br />

issues arising from this, consideration should be given to<br />

installing the base on a sunken concrete slab.<br />

• The Pluvio 2 should only be deployed at sites with low<br />

vandalism risk.<br />

• Provided the Pluvio 2 is part of a telemetry system, site visit<br />

frequency in comparison to required TBR maintenance<br />

visits can be significantly reduced, depending on the typical<br />

UK annual rainfall at the site or any other site drivers.<br />

• Where a Pluvio 2 is to be installed with a solar panel and<br />

battery power then if on telemetry it is suggested to set a<br />

low power alarm. The solar panel and battery should be<br />

sized adequately to provide power throughout the winter.<br />

• Further observations of high intensity rainfall events and<br />

events with high winds would be useful for the further<br />

evaluation of the Pluvio 2 with regards to its recording<br />

accuracy.<br />

Acknowledgements<br />

Thanks to Simon Wills (OTT Hydrometry) for the provision<br />

of the Pluvio 2 and installation support at both test sites,<br />

Eskdalemuir and Tryweryn. Many thanks to Ian Dawson<br />

and Billy Jagger (Met Office) for good cooperation and<br />

their thorough daily raingauge readings at Eskdalemuir. We<br />

would like to thank our SEPA colleagues Stuart Smith and<br />

John Scott and EA colleagues Steven Roberts and Christina<br />

McKay for their help with raingauge installations, logger<br />

configurations and maintenance as well as for general<br />

discussions.<br />

• The Pluvio 2 has passed the success criteria at both trials.<br />

• The Pluvio 2 raingauge shows very close relationship with<br />

the storage daily reference raingauge over the period of<br />

both the Tryweryn and SEPA trials.<br />

6<br />

References<br />

Lanza, L., Leroy, M., Alexandropoulos, C., Stagi L. and<br />

Wauben. W. 2006. WMO laboratory intercomparison of


ainfall intensity gauges,.Final report. WMO Geneva.<br />

Lanza, L.G. and Vuerich, E. 2009. The WMO Field<br />

Intercomparison of Rain Intensity Gauges. Atmospheric<br />

Research.<br />

7

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!