Analysis of sentinel-1 data for regional crop classification: a multi-data approach for rabi crops of district Hisar (Haryana)

Abstract

Generation of spatio-temporal information such as land use system and management practices is one of the key ingredients for carrying out the regional level agro-ecosystem modelling. However, at the regional level availability of such data is scarce, where analysis of a cropping system is essential and a pre-requisite for studying the overall sustainability of the agricultural production system. The present investigation was carried out to identify the actually practised cropping pattern and their mapping in Hisar district of Haryana (India) using Multi-Data Approach (MDA). Multi-date sentinel-1 for the rabi season of 2019 was classified using multi-phase unsupervised classification approach and classes of interest were identified. Finally, classified images were subjected to logical combinations which helped in generating crop classification maps and statistics. Results showed that cropping pattern of the district exhibited huge variations and area under wheat was observed to be highest (204.76 thousand ha) in comparison to mustard crop (64.42 thousand ha) and least was under the sugarcane crop (0.97 thousand ha). Some other crops like vegetables and horticultural crops were also identified during this period, but the major crops that were identified during rabi 2019 were wheat and mustard. Hence, regional crop classification using sentinel-1 data appears to be a valuable tool for predicting a specific regions cropping pattern, which is considered to be the most significant aspect of an agricultural production system.