Abstract:
The NASA Soil Moisture Active Passive (SMAP) mission has a projected launch in 2014 and will provide global mapping of surface soil moisture and landscape freeze/thaw (F/T) status using L-band (1.26 GHz) active and passive microwave remote sensing. Primary science objectives for SMAP include reducing uncertainty regarding terrestrial carbon (CO2) uptake and release and the purported missing carbon sink on land. An operational level 4 carbon (L4_C) product is planned under SMAP to quantify surface soil organic carbon (SOC) stocks, soil moisture and temperature controls for heterotrophic respiration and the net ecosystem exchange of CO2 (NEE) using model assimilation based soil moisture, temperature and F/T inputs from SMAP retrievals with ancillary information on global land cover and vegetation productivity (GPP). We conducted an initial global implementation and evaluation of the SMAP L4_C algorithms using MODIS (MOD17) GPP inputs and MERRA reanalysis based daily surface air temperature and soil moisture fields. The resulting model simulations are generally consistent with the distribution and magnitude of SOC stocks available from global soil inventories, while estimated carbon fluxes also correspond (R2 > 0.6; RMSE < 1.5 g C/m2/day) with CO2 flux measurements from the global tower network (FLUXNET). A model uncertainty analysis indicates an anticipated L4_C product accuracy for NEE within 30 g C / m2 / yr or 1.6 g C / m2 / day, and similar to accuracies attained from tower eddy covariance measurements. The resulting NEE calculations are used as a land surface constraint within an atmospheric transport model assimilation framework (CarbonTracker) to quantify terrestrial source-sink activity for atmospheric CO2.