Bias correction of satellite soil moisture and assimilation into the NASA Catchment land surface model

Surface soil moisture data from different sources (satellite retrievals, ground measurements, and land model integrations of observed meteorological forcing data) have been shown to contain consistent and useful information in their seasonal cycle and anomaly signals even though they typically exhibit very different mean values and variability. At the global scale, in particular, it is currently impossible to determine which soil moisture climatology is more correct. The biases pose a severe obstacle to exploiting the useful information contained in satellite retrievals of soil moisture in a data assimilation algorithm. A simple method of bias removal is to match the cumulative distribution functions (cdf) of the satellite and model data. Cdf estimation typically requires a long data record. By using spatial averaging with a 2 degree moving window we can obtain statistics based on a one-year satellite record that are a good approximation of the desired local statistics of a long time series. This key property opens up the possibility for operational use of current and future soil moisture satellite data. Introduction and Approach Accurate knowledge of the state of the land surface is important for many applications. For example, there is increasing evidence that accurate land initialization contributes to skill in subseasonal climate forecasts of summer mid-latitude precipitation and air temperature (Koster et al., 2003, 2004). Our ability to accurately characterize global soil moisture fields relies on (1) retrievals of surface soil moisture from satellite, and (2) land surface models that integrate meteorological forcing data (such as precipitation and radiation from observations or atmospheric data assimilation) and land surface parameters (such as soil hydraulic or vegetation properties). It has long been argued that a land data assimilation system that merges these two sources of information will improve our knowledge of the state of the land surface. Such a data assimilation system must, however, address severe biases that have been identified in surface soil moisture.