Improved SMAP dual-channel algorithm for the retrieval of soil moisture

Mario Julian Chaubell, Simon H. Yueh, R. Scott Dunbar, Andreas Colliander, Fan Chen, Steven K. Chan, Dara Entekhabi, Rajat Bindlish, Peggy E. O'Neill, Jun Asanuma, Aaron A. Berg, David D. Bosch, Todd Caldwell, Michael H. Cosh, Chandra Holifield Collins, Jose Martinez-Fernandez, Mark Seyfried, Patrick J. Starks, Zhongbo Su, Marc ThibeaultJeffrey Walker

Research output: Contribution to journalArticleResearchpeer-review

63 Citations (Scopus)


The soil moisture active passive (SMAP) mission was designed to acquire L-band radiometer measurements for the estimation of soil moisture (SM) with an average ubRMSD of not more than 0.04 m3 m-3 volumetric accuracy in the top 5 cm for vegetation with a water content of less than 5 kg m2. Single-channel algorithm (SCA) and dual-channel algorithm (DCA) are implemented for the processing of SMAP radiometer data. The SCA using the vertically polarized brightness temperature (SCA-V) has been providing satisfactory SM retrievals. However, the DCA using prelaunch design and algorithm parameters for vertical and horizontal polarization data has a marginal performance. In this article, we show that with the updates of the roughness parameter $h$ and the polarization mixing parameters Q, a modified DCA (MDCA) can achieve improved accuracy over DCA; it also allows for the retrieval of vegetation optical depth (VOD or τ). The retrieval performance of MDCA is assessed and compared with SCA-V and DCA using four years (April 1, 2015 to March 31, 2019) of in situ data from core validation sites (CVSs) and sparse networks. The assessment shows that SCA-V still outperforms all the implemented algorithms.

Original languageEnglish
Pages (from-to)3894-3905
Number of pages12
JournalIEEE Transactions on Geoscience and Remote Sensing
Issue number6
Publication statusPublished - Jun 2020


  • Dual-channel algorithm (DCA)
  • soil moisture (SM) retrieval
  • soil moisture active passive (SMAP)
  • vegetation optical depth (VOD) retrieval

Cite this