HR: 1330h
AN: H12A-0959 [PDF]
TI: Soil Hydraulic Property Uniqueness as Determined From Inverse Modeling Using Surface Temperature : the
Role of Soil Type
AU: * Gutmann, E D
EM: gutmann@colorado.edu
AF: University of Colorado, Department of Geological Sciences
University of Colorado
Campus Box 399, Boulder, CO 80309-0399 United States
AU: Small, E
EM: Eric.Small@colorado.edu
AF: University of Colorado, Department of Geological Sciences
University of Colorado
Campus Box 399, Boulder, CO 80309-0399 United States
AB:
Knowledge of the spatial distribution of soil hydraulic properties (SHPs) is critical for a broad range of earth system
problems, from predicting the hydrologic implications of climate variability to understanding how land cover and land use
change modify water, energy, and carbon cycling at the earth's land surface. SHPs strongly control how water and energy flow
through soils, and therefore influence infiltration at the soil surface, redistribution of water within the soil column, and
the loss of water from soil via evaporation and transpiration. Soil hydraulic properties define two fundamental
relationships: (1) the water characteristic function or water retention curve, which describes how the water content of soil
varies with the potential of the medium; and (2) the unsaturated conductivity curve of the medium, which shows how
conductivity varies with water content. We have developed a method to determine SHPs from remotely sensed surface
temperature (T$_{s}$) via inverse modeling. Model simulations of T$_{s}$ are compared to observed values. If the observed
and modeled values do not match, the SHPs are modified, and the forward model run is repeated. This process continues until
the observed and modeled T$_{s}$ values are minimized. This method uniquely determines the SHP for some soil types, but is
much less accurate for others. We will discuss the limitations of this method with respect to soil type.
We present an example based partially on applying the inverse procedure to forward model output for bare soil. We used a
modified version of the NOAH land surface model with observed weather forcing data and a variety of soil types. The Ts
output from these forward model runs were used as input to the MOSCEM-UA inverse modeling framework. The accuracy with which
the inverse procedure is able to determine the soil hydraulic properties is controlled in part by the original soil type
itself. For example, when one assumes a 1K RMS sensor error, the van Genuchten "n" parameter for silt can only be determined
to fall within a range of values from 1.2 to 3, but for sand it can be determined to fall between 3 and 3.5. Additionally,
certain parameters are easier to estimate then others, the van Genuchten "alpha" parameter is less identifiable than the "n"
parameter for instance.
DE: 1833 Hydroclimatology
DE: 1866 Soil moisture
DE: 1875 Unsaturated zone
DE: 1878 Water/energy interactions
DE: 1894 Instruments and techniques
SC: Hydrology [H]
MN: 2003 Fall Meeting