09.05.2014 Views

FY2010 - Oak Ridge National Laboratory

FY2010 - Oak Ridge National Laboratory

FY2010 - Oak Ridge National Laboratory

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.

Director’s R&D Fund—<br />

Neutron Sciences<br />

project. Comparative experiments at the two imaging facilities have been performed. Flint #13 sand and<br />

Hanford soil were packed inside an aluminum cylinder with a 2.7 cm OD and initially saturated with<br />

water. Various water potentials (ψ) were applied to the bottom of the column using the hanging water<br />

column, and radiography (2D) images were taken at each equilibrium state during drying and wetting<br />

processes with a 60 s exposure time. Three-dimensional images with a rotational increment of 0.25° were<br />

also acquired at about half way in the drying and wetting stages. The 2D images of the soil column using<br />

NIST BT-2 illustrate water movement into the soil during the wetting process. The average water content<br />

of Flint #13 sand (4 cm height) at each equilibrium state during the drying and wetting cycles was<br />

measured by two methods: (1) volumetric measurements from the hanging water column and (2)<br />

integration of pixel-based water contents using neutron imaging data. The two methods yielded very<br />

similar average water retention curves. Similar measurements were performed at HFIR CG-1, using two<br />

different samples, Hanford and Flint, respectively.<br />

Evaluating water transport limitations in soil–plant systems. Switchgrass seeds were grown in pure silica<br />

sand within aluminum cylinders and subjected to low moisture conditions. Water was applied to the<br />

young seedlings through an injection port at the base of roots. A series of neutron images were taken<br />

every 2 min at HFIR CG-1D (exposure time was 120 s). After 18 min, there was little detectable change<br />

in the water content of the roots, albeit the plants were exposed to low light conditions which likely<br />

limited their uptake and transport of the injected water. Further work has utilized both D 2 O and H 2 O to<br />

enhance image contrast, addition of illumination treatments to enhance water flux, thin rectangular<br />

containers to improve root display, and longer intervals to track water flux.<br />

Maize seeds were germinated in silica sand and watered with D 2 O. After several weeks, 9 mL of water<br />

was injected into the bottom of the container and water distribution was tracked through time. The<br />

analysis reveals areas of increased moisture (large deep root, root tips) and decreased moisture (darker<br />

shallow roots) that establish the utility of this technique to determine water transport limitations in situ,<br />

specifically dynamics of water content surrounding root tissue.<br />

The relative soil–root hydration surrounding a growing root can reveal rhizosphere hydration at 80 m<br />

resolution through time, which will be assessed during periods of drying to determine loss of<br />

conductivity. Our 2D/3D neutron imaging data for water in roots and soil clearly demonstrate the<br />

potential of the technique for tracing the flow of water to roots and quantifying the spatial distribution of<br />

water in partially saturated soil. Our future research will focus on investigating (1) differences in point<br />

water retention curves estimated from conventional hanging water column experiments and those<br />

measured directly by neutron imaging; (2) comparison of forward numerical simulations with direct<br />

neutron-based measurements of water distribution; (3) reduction in hydraulic conductance of rhizosphere<br />

as soil dries under undisturbed conditions; and (4) fine-scale, 3D geometry of air–water interfaces using<br />

neutron tomography.<br />

Assessment of analytical and numerical models for predicting fluid flow. We have extracted point water<br />

retention functions for Flint #13 sand by constructing the drying curve on a pixel-by-pixel basis and then<br />

averaging row-by-row across the cylinder. The range in point functions obtained from different heights in<br />

the column is quite small as expected for a relatively homogenous material like Flint sand. We are now in<br />

the process of modeling the average curves using TRUECELL to analytically predict the point function,<br />

which can be compared with the observed point functions obtained by neutron imaging. Inverse<br />

numerical modeling of the 2D water distributions using HYDRUS 2D is also under way. The degree to<br />

which the observed and predicted functions agree with each other can provide an objective evaluation of<br />

the assumptions inherent within the different models.<br />

61

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

Saved successfully!

Ooh no, something went wrong!