Optimization of a radiative transfer forward operator for simulating SMOS brightness temperatures over the Upper Mississippi Basin

Hans Lievens, Ahmad Al Bitar, Niko EC Verhoest, Francois Cabot, Gabrielle JM De Lannoy, Matthias Drusch, Gift Dumedah, Harrie-Jan Hendricks Franssen, Yann Henry Kerr, Sat Kumar Tomer, Brecht Martens, Olivier Merlin, Ming Pan, Martinus Johannes van den Berg, Harry Vereecken, Jeffrey Phillip Walker, Eric F Wood, Valentijn Rachel Noel Pauwels

Research output: Contribution to journalArticleResearchpeer-review

28 Citations (Scopus)


The Soil Moisture Ocean Salinity (SMOS) satellite mission routinely provides global multiangular observations of brightness temperature TB at both horizontal and vertical polarization with a 3-day repeat period. The assimilation of such data into a land surface model (LSM) may improve the skill of operational flood forecasts through an improved estimation of soil moisture SM. To accommodate for the direct assimilation of the SMOS TB data, the LSM needs to be coupled with a radiative transfer model (RTM), serving as a forward operator for the simulation of multiangular and multipolarization top of the atmosphere TBs. This study investigates the use of the Variable Infiltration Capacity model coupled with the Community Microwave Emission Modelling Platform for simulating SMOS TB observations over the upper Mississippi basin, United States. For a period of 2 years (2010-11), a comparison between SMOS TBs and simulations with literature-based RTM parameters reveals a basin-averaged bias of 30 K. Therefore, time series of SMOS TB observations are used to investigate ways for mitigating these large biases. Specifically, the study demonstrates the impact of the LSM soil moisture climatology in the magnitude of TB biases. After cumulative distribution function matching the SM climatology of the LSM to SMOS retrievals, the average bias decreases from 30 K to less than 5 K. Further improvements can be made through calibration of RTM parameters related to the modeling of surface roughness and vegetation. Consequently, it can be concluded that SM rescaling and RTM optimization are efficient means for mitigating biases and form a necessary preparatory step for data assimilation.
Original languageEnglish
Pages (from-to)1109 - 1134
Number of pages26
JournalJournal of Hydrometeorology
Issue number3
Publication statusPublished - 2015

Cite this