Monitoring transplanting operation of rice crop using passive microwave radiometer data

This study highlights a methodology to detect the transplanting operation of wetland rice at a regional scale using SSM/I brightness temperature in frequencies like 19, 37 and 85 GHz with vertical polarization. A cloud removal algorithm was used to make weekly composites of the brightness temperature, which were used to estimate the soil wetness index (SWI). Flooding of rice fields with water for transplanting induces very high change in SWI due to contrasting dielectric constant of water (80) and soil (4). Different weather conditions and fractional wet area under the footprint of sensor affect SWI, hence absolute value of SWI was not adequate. Therefore, multi-year SWI anomaly was used to generate a threshold value of SWI change to detect when SWI change between two consecutive weeks was an indicator of transplanting. A minimum SWI increment of 6 between two successive weeks was determined to be an optimal threshold value to detect transplanted fields. Based on the SWI threshold, five transplanting periods (standard meteorological week 21-25) were observed over a large area (Punjab and Haryana states) in the Indo-Gangetic plains of India. The cumulative areas under different rice transplanting periods were of 14%, 19%, 37%, 53% and 100% for the standard meteorological weeks 21, 22, 23, 24 and 25, respectively. The results were validated for the 2007 wet season rice crop. Estimated rice transplanting periods using SSM/I data were found to be in good agreement (90%) with the ground truth observations in the study area. (C) 2010 IAgrE. Published by Elsevier Ltd. All rights reserved.