Download the paper:
Abstract:
Global retrievals of surface soil moisture from the Scanning Multichannel Microwave Radiometer for the period 1979-87 are assimilated into the NASA Catchment land surface model as it is driven with surface meteorological data derived from observations. Since model and satellite soil moisture data have substantially different climatologies, bias-reducing scaling techniques are applied as part of the assimilation system. Validation against ground-based measurements in Eurasia and North America from the Global Soil Moisture Data Bank demonstrates a long assumed (but rarely proven) property of soil moisture fields derived from data assimilation - that the assimilation product is superior to either satellite data or model data alone. We show that assimilation provides improved estimates of the average seasonal cycle and the anomalies. An analysis of the innovations reveals that the filter is only partially operating within its underlying assumptions and offers clues how spatially distributed model error parameters could further enhance filter performance.