Assessing the Seasonal Crop Acreage in the Ganges Delta Using Multi-Temporal Sentinel-2 Data: A Case Study in Gosaba CD Block

Abstract

The present study assessed the seasonal and inter-annual variation in cropping pattern and crop acreage during summer (mid-February to May) and winter (November to January) seasons of 2017-2018, 2018-2019, and 2019-2020 cropping years over the Gosaba Block of Indian Sundarbans. Multi-temporal Sentinel-2 images acquired during the critical crop growth stages during a particular season were taken together for classification instead of a single image classification approach. This study employed the random forest algorithm for image classification for seasonal crop-fallow mapping over three consecutive years. The estimated post-monsoon (winter) crop coverage was less than that in the pre-monsoon (summer) season. In the post-monsoon season crop coverage varied from 3% to 12% of the net cultivable area, whereas pre-monsoon season crop coverage varied from 23% to 42%. The estimated area under boro rice decreased in 2018-19 and 2019-20 with a simultaneous increase in non-rice crops. Seasonal fallow land showed wide inter-annual variation. The amount and distribution of rainfall during the last part of the monsoon season and in the following months had a strong influence on the cropping pattern of the Sundarbans region. Results clearly showed that winter crop coverage gradually decreased over the study period depending on the rainfall pattern in the monsoon season and in the following months. During summer season, the area under non-rice crops notably increased at the cost of the area under boro rice. These findings clearly indicated the farmers’ preferences for cultivating crops according to the climatic variability. The present study may be helpful for cropping intensification in the threatened regions for achieving sustainability.