Abstract
Information about the accurate soil moisture is crucial for agriculture. In the present work the Landsat-8, Sentinel-1 satellite data and Modified Water Cloud Model (MWCM) have been used to retrieve soil moisture from an agriculture area. The effect of crop canopy was incoporated in MWCM. The crop coverage was mapped from LANDSAT-8. Further, an account of the V1 and V2 vegetation and crop canopy’s primary and secondary descriptor were also estimated, respectively. Afterwards, a retrieval scheme of soil moisture was developed using dual polarizations (VH (Vertical characteristics signal send by radar/sensor, Horizontal characteristics signal received by radar/sensor) and VV (Vertical characteristics signal send by radar/sensor, Vertical characteristics signal received by radar/sensor)) available from the quad polarization of Sentinel-1 data. The model based soil moisture was compared with the observed soil moisture obtained from calibrated Time Domain Reflectometer (TDR). The model performance was evaluated with the help of statistical tests. The statistical tests result shows a good correlation between observed and estimated soil moisture (R2 = 0.86, 0.96, and 0.91, RMSE = 0.06, 0.03, and 0.02, RMSE% =2.0, 1.9, and 1.5) for different dates as: 17 January, 6 March and 15 April-2018, respectively. Therefore, this method has an operational potential for estimating soil moisture from area under crop.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call