First Results from the SMAP Level 4 Surface and Root Zone Soil Moisture (L4_SM) Data Product

Observations collected by the NASA Soil Moisture Active Passive (SMAP) mission are directly related to surface soil moisture (0-5 cm). Several of the key applications targeted by SMAP, however, require knowledge of root zone soil moisture (0-100 cm). The foremost objective of the SMAP Surface and Root Zone Soil Moisture (L4_SM) data product is to fill this gap and provide estimates of root zone soil moisture that are informed by and are consistent with SMAP observations. Such estimates are obtained by merging SMAP brightness temperature observations with estimates from a land surface model in the NASA Goddard Earth Observing System, version 5 (GEOS-5) land data assimilation system. The land surface model component of the system, the Catchment model, is driven with observations-based surface meteorological forcing data, including precipitation, which is the most important driver for soil moisture. The model also encapsulates knowledge of key land surface processes, including the vertical transfer of soil moisture between the surface and root zone reservoirs. A radiative transfer model to simulate brightness temperatures was calibrated using L-band observations from the Soil Moisture and Ocean Salinity (SMOS) mission.

The horizontally distributed ensemble Kalman filter update step considers the respective uncertainties of the model estimates and the observations, resulting in a soil moisture and soil temperature analysis at 9 km resolution that is, in theory, superior to satellite or model estimates alone. Moreover, error estimates for the L4_SM product are generated as a by-product of the data assimilation system. Early mission results indicate that the L4_SM product meets the root-mean-square error target of 0.04 m3/m3 (after removal of the long-term mean differences). An investigation of the observation-minus-forecast residuals further reveals modest biases in the assimilation system. It also demonstrates where the assimilation system overestimates or underestimates the actual errors, suggesting avenues for future calibration and algorithm enhancements.