From field observations to temporally dynamic soil surface roughness retrievals in the U.S. Corn Belt

• Published in: Remote sensing of environment Vol. 287; p. 113458
• Main Authors: Walker, Victoria A., Wallace, Vivian, Yildirim, Enes, Eichinger, William E., Cosh, Michael H., Hornbuckle, Brian K.
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 15.03.2023
• Subjects: L-band vegetation optical depth; SMAP; Soil surface roughness; L-band vegetation optical depth; Soil surface roughness; SMAP
• ISSN: 0034-4257; 1879-0704
• DOI: 10.1016/j.rse.2023.113458

Soil surface roughness in the U.S. Corn Belt varies outside of the growing season due to rainfall and soil management activities such as tillage. This is not consistent with the assumption that roughness is a static parameter in the Soil Moisture Active Passive (SMAP) soil moisture retrieval algorithm. SMAPVEX16-IA, an extensive field campaign focused on calibration of SMAP retrievals in croplands, occurred during a period when roughness is theoretically at a minima. Satellite-scale measurements of rms height and correlation length collected via pinboard, gridboard, and lidar methodologies resulted in a ensemble of model roughness coefficients that, at h = [0.13, 0.97], is significantly rougher than the original SMAP assumption in croplands of h = 0.108. Comparing simulated brightness temperatures using the ensemble h to SMAP observations resulted in a best-fit angular sensitivity coefficient of N=−2 for both horizontal and vertical polarizations. Satellite-scale h is then retrieved for a four-year study period (excluding growing seasons) via the single channel algorithm from SMAP brightness temperatures and in situ observations of soil moisture and temperature. Roughness retrievals are highly variable in the spring as rainfall events decrease h followed by an increase during soil dry-down; h steadily increases in the fall post-harvest as fields are tilled across the network. Retrievals of h are effectively independent of polarization and demonstrate the potential to simultaneously retrieve soil moisture and roughness in the absence of crops. These soil roughness retrievals could provide new information on tillage and other soil management activities. •Soil roughness sampled across a SMAP core validation network in the Corn Belt.•Temporally dynamic model roughness coefficient retrieved when crops are absent.•Both in situ and retrieved roughness is larger than assumed by SMAP in croplands.•Retrieved roughness increases from fall tillage then varies with spring rainfall.