First application of regression analysis to retrieve soil moisture from SMAP brightness temperature observations consistent with SMOS soil moisture
Langue
en
Autre communication scientifique (congrès sans actes - poster - séminaire...)
Ce document a été publié dans
MicroRad 2016 - 14. Specialist Meeting on Microwave Radiometry and Remote Sensing of the Environment, 2016-04-11, Espoo. 2016, vol. 4, p. np
Aalto University publication series
Résumé en anglais
Two dedicated soil moisture (SM) spaceborne missions, ESA’s Soil Moisture and Ocean Salinity (SMOS) and NASA’s Soil Moisture Active Passive (SMAP) satellites, were launched in 2009 and 2015, respectively. Both satellites ...Lire la suite >
Two dedicated soil moisture (SM) spaceborne missions, ESA’s Soil Moisture and Ocean Salinity (SMOS) and NASA’s Soil Moisture Active Passive (SMAP) satellites, were launched in 2009 and 2015, respectively. Both satellites have been providing microwave brightness temperature (TB) observations and SM retrievals at L-band since then (Entekhabi et al., 2010; Kerr et al., 2012). A recent study demonstrated the efficiency of physically-based multiple-linear regression equations (Wigneron et al., 2004) to retrieve SM from AMSR-E TB observations. The regression equations were derived analytically from the radiative transfer model. The purpose of that initial study was to extend the SMOS SM product into the past i.e., 2003-2009, using AMSR-E TB observations. The current study follows the same strategy to retrieve SM from SMAP TB observations with a purpose to extend the SMOS SM product into the future at the global scale. Regression coefficients were calibrated using SMOS horizontal and vertical TB observations and SM level 3 (SMOSL3 as a training data), over the 2013 - 2014 period. Based on these calibrated coefficients, global SM maps were produced from the SMAP TB observations during the 31/03-08/09/2015 period (referred here to as SMAP-reg). The SM data set obtained from SMAP TBs using the regression equations has been compared to the SMAP SM dataset computed withthe single channel algorithm and both exhibit the same temporal dynamics. For instance, figure 1 shows the (Pearson) correlation coefficient (R) between SMAP-reg and SMAP original SM over 31/03-08/09/2015. A remarkable agreement, R (mostly > 0.8), was obtained between the SMAP-reg and SMAP original SM products.Ongoing evaluations of the SMAP-reg SM product, with comparison to the SMAP original SM, against the global MERRA-Land SM simulations and in situ measurements will be presented. The main interest in the SMAP-reg SM product is that it is fully consistent with the SMOS Level 3 SM product. One of the key remaining tasks is toensure the consistent relative calibration between SMOS and SMAP TBs.< Réduire
Mots clés
projet smos
télédétection
donnée satellite
équation de régression
modèle de transfert radiatif
Mots clés en anglais
remote sensing
Origine
Importé de halUnités de recherche