Soil physical and hydraulic properties are essential inputs to the soil-vegetation-atmosphere transfer (SVAT) schemes currently found in regional- to global-scale climate and hydrology models. Increasing the availability and accuracy of these soil parameters will improve the physical realism of SVAT parameterizations and may lead to a better understanding of the water and energy budget at different scales as well as soil moisture distribution patterns measured in situ and inferred from remotely sensed observations.
As a part of the 1997 Southern Great Plains Hydrology Experiment (SGP97) conducted in Oklahoma, soil core sampling was conducted over quarter sections in each of the intensive research areas (Central Facility, El Reno, and Little Washita) within the SGP region. Sampling was based on a-priori information and concurrent inspection of different representative soil, landscape, and vegetation combinations. The cores were used to characterize the soil physical (texture, bulk density, organic matter content), hydraulic (water retention and hydraulic conductivity functions) and thermal properties (volumetric heat capacity, heat diffusivity, and thermal conductivity).
The SGP97 soil characterization effort has yielded a very unique data set, which we are using to infer important soil (e.g., soil type, texture, porosity, bulk density), topographic (slope, aspect, elevation, depth to water table, etc.) and vegetation (type, vegetation density, management practice, etc.) parameters with neural networks. This work will establish pedo-topo-vegetation transfer functions (PTVTFs) describing soil hydraulic/thermal properties and soil moisture contents at different (hierarchical) spatial scales. This paper summarizes our field and laboratory data and the subsequent development of spatial data coverages in a GIS environment.