干旱区生态环境知识资源中心

Traditionally, effective soil temperature (T-eff) has been considered to be a secondary intermediate variable in microwave radiative transfer theory. However, its impact on microwave radiometry is comparable to that of vegetation cover, soil surface roughness and dielectric constant. T-eff is defined as the weighted temperature of the emitting layers, where the weighting involved depends on the soil moisture profile. In this study, we evaluate the suitability of various models for estimating T-eff using temperature and moisture profiles obtained from a land surface model. MERRA-Land (the Modern-Era Retrospective analysis for Research and Applications-Land) soil moisture profiles and temperature profiles were used to reproduce global T-eff data sets with single parameter schemes (e.g. the Beta Soil Moisture Active Passive, SMAP scheme), two-parameter Choudhury’s schemes (e.g. the current SMAP’s scheme), two-parameter Wigneron schemes (e.g. the current Soil Moisture and Ocean Salinity scheme, SMOS), and multilayer T-eff schemes (e.g. Lv’s scheme). The results show that differences in T-eff between these schemes are usually <5 K. The comparison between Wigneron’s and Lv’s schemes indicates that, the difference is small (RMSD, root mean squared difference <1 K). In exceptional cases (<1%), the RMSD between Choudhury’s T-eff scheme and Lv’s scheme can reach around 5 K. The Beta SMAP T-eff has a difference of around 5 K, compared to Lv’s scheme. Such a change in T-eff could lead to an emissivity difference of around 0.015. The most extreme emissivity difference is only found in desert areas, at 42.5 degrees (the incidence angle used by the SMAP mission is about 40 degrees). (C) 2016 Elsevier Inc. All rights reserved.