A method to downscale soil moisture to fine-resolutions using topographic, vegetation, and soil data
Date
2014
Authors
Ranney, Kayla J., author
Niemann, Jeffrey D., advisor
Green, Timothy R., committee member
Kampf, Stephanie K., committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Various remote-sensing and ground-based sensor methods are available to estimate soil moisture over large regions with spatial resolutions greater than 500 m. However, applications such as water management and agricultural production require finer resolutions (10 - 100 m grid cells). To reach such resolutions, soil moisture must be downscaled using supplemental data. Several downscaling methods use only topographic data, but vegetation and soil characteristics also affect fine-scale soil moisture variations. In this thesis, a downscaling model that uses topographic, vegetation, and soil data is presented, which is called the Equilibrium Moisture from Topography, Vegetation, and Soil (EMT+VS) model. The EMT+VS model assumes a steady-state water balance involving: infiltration, deep drainage, lateral flow, and evapotranspiration. The magnitude of each process at each location is inferred from topographic, vegetation, and soil characteristics. To evaluate the model, it is applied to three catchments with extensive soil moisture and topographic data and compared to an Empirical Orthogonal Function (EOF) downscaling method. The primary test catchment is Cache la Poudre, which has variable vegetation cover. Extensive vegetation and soil data were available for this catchment. Additional testing is performed using the Tarrawarra and Nerrigundah catchments where vegetation is relatively homogeneous and limited soil data are available for interpolation. For Cache la Poudre, the estimated soil moisture patterns improve substantially when the vegetation and soil data are used in addition to topographic data, and the performance is similar for the EMT+VS and EOF models. Adding spatially-interpolated soil data to the topographic data at Tarrawarra and Nerrigundah decreases model performance and results in worse performance than the EOF method, in which the soil data are not highly weighted. These results suggest that the soil data must have greater spatial detail to be useful to the EMT+VS model.