Spatio-temporal dynamics of soil moisture in tallgrass prairie using ERS-1 SAR image time series: first year results

摘要

Radar remote sensing of soil moisture has been demonstrated in croplands and rangelands. The authors are interested in synoptic estimates of soil moisture to aid long term ecological investigation of tallgrass prairie and have initiated a three year study to develop this technology. The authors' study area, Konza Prairie Research Natural Area (KPRNA), exhibits a variety of factors to complicate soil moisture estimation via synthetic aperture radar (SAR) including significant terrain variation, heterogeneous soil types and depths, and watershed-level experimental treatments including prescribed burning, grazing by large ungulates, and water supplementation. Several indicators of the moisture status of soil-water-vegetation continuum were gathered during the 1995 growing season including soil moisture via neutron probe and TDR, xylem pressure,potential values, and aboveground biomass (wet and dry weights). Due to orbital track overlap, eleven ERS-1 SAR images of KPRNA were acquired from mid-April to early October. Spring through mid-summer were unusually cool and wet, thereby delaying phenological development and attenuating the water stress typical of the later growing season. Backscattering values show some consistent agreement with ground data. Watersheds burned in April showed markedly higher backscattering at the end of the growing season due to a lack of litter layer from previous seasons' growth. Grazed units that were also burned showed the highest backscattering caused by no litter layer and more exposed soil.