Skip to main content
ARS Home » Research » Publications at this Location » Publication #207176
Title: Representativeness of Point Measure Observations, Upscaling and Assimilation, Quantification and Reduction of Predictive Uncertainty for Sustainable Water Resources Management

Author
item DELANNOY, GABRIELLE - GHENT UNIVERSITY
item PAUWELLS, VALENTUN - GHENT UNIVERSITY
item HOUSER, PAUL - GEORGE MASON UNIVERSITY
item VERHOEST, NIKO - GHENT UNIVERSITY
item Gish, Timothy

Submitted to: International Union of Geodesy and Geophysics Meeting
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/1/2007
Publication Date: 7/2/2007
Citation: De Lannoy, G.J.M., Pauwels, V.R.N., Houser, P.R., Gish, T.J., Verhoest, N.E.C. 2007. Representativeness of point soil moisture observations, upscaling and assimilation, quantification and reduction of predictive uncertainty for sustainable water resources management. International Union of Geodesy and Geophysics Symposium. 313:249-257.

Interpretive Summary: Soil water contents are complex and dynamic but influence crop production, chemical transport, chemical persistence, and the hydrologic cycle in general. As a result, the capacity to model soil water behavior is important to a number of disciplines. To estimate the temporal evolution of the spatial mean soil moisture patterns, the relationship between point measurements and the average at the field scale was investigated. The best results for the prediction of the spatial mean soil moisture were obtained through the assimilation of observations from soil moisture probes with close to zero time-mean differences between their recorded point values and the spatial mean values. Consequently, improved soil moisture predictions can be obtained by upscaling the point data, e.g. after matching the point observations cumulative density function to that of the spatial mean soil moisture. This approach will help quantify field scale soil water dynamics and will be used to quantify remote sensing algorithms of soil moisture.

Technical Abstract: To estimate the temporal evolution of the spatial mean soil moisture in the Optimizing Production Inputs for Economic and Environmental Enhancement (OPE3) field, the relationship between point measurements and the average behavior of the field scale soil moisture has been investigated. In a simple variational assimilation experiment with the Community Land Model (CLM2.0), it has been shown that the soil moisture information from a representative site was much more appropriate for estimating the spatial mean soil moisture profile than the information from other sites. The best results for the reanalysis as well as for the prediction of the spatial mean soil moisture were obtained through the assimilation of observations from probes with close to zero time-mean differences between their recorded point values and the spatial mean values. Further improved results can be obtained by upscaling the point data, e.g. after matching the point observations cumulative density function to that of the spatial mean soil moisture.