Estimating wheat biomass from GF-3 data and a polarized water cloud model

Wheat is one of the staple crops of the world. With the wide application of remote-sensing methods in agriculture, the use of data from synthetic aperture radar has attracted increasing attention for monitoring wheat growth. Most of previous studies estimated wheat biomass based on a water cloud model (WCM). However, when no data are available on soil moisture content, the applicability of such models is greatly reduced because of insufficient parameters. Thus, this study proposed a new polarized water cloud model (PWCM) called APWCM, which is a physical model and no auxiliary ground data was needed including soil moisture data. APWCM and WCM model was compared to estimate the above ground biomass of wheat in two different study areas. The results revealed that the WCM has a slightly lower root mean squared error (RMSE = 131.63 g m(-2), and 645.17 g m(-2)) in two different study areas. However, the APWCM has lower relative error (RE = 17.91%, and 13.53%) for wheat biomass estimation in areas with higher biomsass. The final result indicates that the APWCM can replace the WCM for estimating wheat biomass based on Gaofen-3 (GF-3) data when soil-moisture data are not available.