Vol. 12, Issue 1 (2023)

Author(s):
MP Adawadkar, SB Wadatkar, AR Pimpale and IK Ramteke

Abstract:
Crop acreage data and crop distribution in different districts are essential for the planning and implementation of various governmental policies. The current study was conducted for three districts of North Maharashtra growing wheat and onion crop using multi-spectral, multi-date Sentinel 2A satellite data for Rabi season of 2020-21. The Ground Truth of 54 sites was taken during the study. Time series NDVI stack of multi-date Sentinel 2A data was prepared in ERDAS Imagine software. The ERDAS Imagine version 20 and ArcGIS version 10.2 software were used for all remote sensing and GIS analysis. Hybrid classification method using K-means clustering and visual analysis was used for identification and acreage estimation. Unsupervised classification was carried out on resultant multilayer dataset and a classified image with 100 classes of clusters was generated. Reference Temporal Spectral Profile (RTSP) based on ground truth data was used for supervised classification. Acreage of wheat and onion estimated using remote sensing was then compared with statistics given by Department of Agriculture, Government of Maharashtra. Wheat and onion acreage for the study area using this approach is found 172903 ha and 190404 ha which is deviating 3.66 and 9.08 percent from than statistics given by the Department of Agriculture.