A Nomograph to Incorporate Geophysical Heterogeneity in Soil Moisture Downscaling

摘要

Hydrological applications require robust and periodic spatially distributed soil moisture data. Radiometer-based soil moisture (similar to 30-60-km resolution), after being appropriately downscaled (5-km resolution), can be a valuable resource for providing such data globally. However, the accuracy of available downscaling algorithms is severely affected by subgrid variability in geophysical factors and precipitation within a satellite footprint. In this work, we introduce a scaling nomograph that incorporates the scale and site specific dependence of soil moisture on geophysical heterogeneity and antecedent wetness conditions to overcome this limitation. We developed functional scaling relationships to estimate the semivariogram of downscaled soil moisture change without any available fine-scale soil moisture data. The nomograph enables these relationships to be specific to the geophysical heterogeneity and antecedent wetness within a radiometer-based satellite footprint through footprint specific heterogeneity and wetness indices. The heterogeneity index quantifies the subgrid scale variability and covariability of soil, vegetation, and topography within the footprint, and the wetness index is a measure of antecedent precipitation. The nomograph was developed for Arizona, Iowa, and Oklahoma and can enable downscaling to scales varying between 0.8 and 6.4km. The true power of the nomograph is to enable the use of static dominant factors like soil to define dynamic scale specific scaling relationships for soil moisture for different kinds of land use and land cover in a data driven yet scientific approach, thus providing spatial transferability to the downscaling scheme. The spatial transferability of the nomograph was validated by downscaling Soil Moisture Ocean Salinity data in Manitoba, Canada.Plain Language Summary The launch of numerous satellites like National Aeronautics and Space Administration's (NASA) Soil Moisture Active Passive satellite and European Space Agency's (ESA) Soil Moisture Ocean Salinity satellite has opened avenues for the use of remotely sensed soil moisture in operational modeling scenarios. However, these data sets require some degree of downscaling before being effectively incorporated into models for operational use. In this study, we devised a new technique to downscale satellite based remotely sensed soil moisture data to useful operational scales. The technique proposed in this study is a first demonstration for a data-driven method to incorporate subgrid variability of land-surface heterogeneity and precipitation into the scaling technique which have extensively been established as limiting factors for the performance of scaling algorithms. The technique is based on a novel nomograph (look-up graph) introduced in this study that enables the scaling algorithm to be dynamic based on the local heterogeneity and prevalent wetness conditions and has the potential for spatial transferability.