Global Calibration of the GEOS-5 L-Band Microwave Radiative Transfer Model over Nonfrozen Land Using SMOS Observations

• Published in: Journal of hydrometeorology 2013-06, Vol.14 (3), p.765-785
• Main Authors: De Lannoy, Gabriëlle J. M., Reichle, Rolf H., Pauwels, Valentijn R. N.
• Format: Article
• Language: English
• Subjects: Angle of incidence; Calibration; Forest soils; Hydrometeorology; Marine; Mathematical independent variables; Microwaves; Opacity; Soil water; Standard deviation; Vegetation
• ISSN: 1525-755X; 1525-7541
• DOI: 10.1175/jhm-d-12-092.1

A zero-order (tau-omega) microwave radiative transfer model (RTM) is coupled to the Goddard Earth Observing System, version 5 (GEOS-5) catchment land surface model in preparation for the future assimilation of global brightness temperatures (Tb) from the L-band (1.4 GHz) Soil Moisture Ocean Salinity (SMOS) and Soil Moisture Active Passive (SMAP) missions. Simulations using literature values for the RTM parameters result in Tb biases of 10–50 K against SMOS observations. Multiangular SMOS observations during nonfrozen conditions from 1 July 2011 to 1 July 2012 are used to calibrate parameters related to the microwave roughnessh, vegetation opacityτand/or scattering albedoωseparately for each observed 36-km land grid cell. A particle swarm optimization is used to minimize differences in the long-term (climatological) mean values and standard deviations between SMOS observations and simulations, without attempting to reduce the shorter-term (seasonal to daily) errors. After calibration, global Tb simulations for the validation year (1 July 2010 to 1 July 2011) are largely unbiased for multiple incidence angles and both H and V polarization [e.g., the global average absolute difference is 2.7 K for TbH(42.5°), i.e., at 42.5° incidence angle]. The calibrated parameter values depend to some extent on the specific land surface conditions simulated by the GEOS-5 system and on the scale of the SMOS observations, but they also show realistic spatial distributions. Aggregating the calibrated parameter values by vegetation class prior to using them in the RTM maintains low global biases but increases local biases [e.g., the global average absolute difference is 7.1 K for TbH(42.5°)].