|CAI, XITIAN - Princeton University|
|PAN, MING - Princeton University|
|CHANEY, N. - Princeton University|
|COLLIANDER, A - Jet Propulsion Laboratory|
|MISRA, S. - Jet Propulsion Laboratory|
|WOODD, E. - Princeton University|
Submitted to: Water Resources Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/13/2017
Publication Date: 4/30/2017
Citation: Cai, X., Pan, M., Chaney, N., Colliander, A., Misra, S., Cosh, M.H., Crow, W.T., Jackson, T.J., Woodd, E. 2017. Validation of SMAP soil moisture for the SMAPVEX15 field campaign using a hyper-resolution model. Water Resources Research. 53(4):3013-3028.
Interpretive Summary: This research uses a very high resolution land surface model-HydroBlocks-for the validation of the SMAP (Soil Moisture Active Passive) soil moisture products. At 30 m x 30 m resolution, the model can represent the spatial heterogeneity of soil. Therefore, the model is used as a tool to upscale in situ measurements to obtain areal mean soil moisture that is comparable to satellite retrievals. This research demonstrated that hyper-resolution land surface models (e.g., HydroBlocks) are very promising in solving the spatial mismatch between satellite footprint (~40 km) and the point scale of in situ measurements. Particularly they are very useful in the validation of satellite-derived soil moisture products. This specific study tests this concept on a field experiment which occurred in 2015 in Tombstone, Arizona. Our findings regarding the weaknesses of the SMAP retrievals are useful for improving the accuracy of SMAP soil moisture products and thus facilitate its usage in studying the water, energy, and carbon cycles.
Technical Abstract: Accurate global mapping of soil moisture is the goal of the Soil Moisture Active Passive (SMAP) mission, which is expected to improve the estimation of water, energy, and carbon exchanges between the land and the atmosphere. Like other satellite products, the SMAP soil moisture retrievals need to be validated, with the validation relying heavily on in situ measurements. However, a one-to-one comparison is ill-advised due to the spatial mismatch of the large SMAP footprint (~40 km) and the point scale in situ measurements. This study uses a recently developed hyper-resolution land surface model—HydroBlocks—as a tool to upscale in situ soil moisture measurements for the SMAPVEX15 (SMAP Validation Experiment 2015) field campaign during August 2-18, 2015. Calibrated against in situ observation, HydroBlocks shows a satisfactory Kling-Gupta efficiency (KGE) of 0.817 and RMSE of 0.019 m3/m3 for the calibration period. These results indicate that HydroBlocks can be used to upscale in situ measurements for this site. Different from previous studies, here in situ measurements are upscaled using a land surface model without bias correction. The upscaled soil moisture is then used to evaluate SMAP (passive) soil moisture products. The comparison of the upscaled network to SMAP shows that the retrievals are generally able to capture the areal-averaged soil moisture temporal variations. However, SMAP appears to be over sensitive to summer precipitation. We expect these findings can be used to improve the SMAP soil moisture product and thus facilitate its usage in studying the water, energy, and carbon cycles.