Assessment of multi-date Sentinel-2 data for field-level monitoring of isabgol (Plantago ovata Forsk) cropping practices in India

Abstract

Accurate identification and mapping of isabgol fields help in macro-level planning in the arid and semi-arid regions, where variability is very high due to erratic weather conditions, besides providing the production estimates of the crop. Isabgol is an important medicinal crop cultivated in western India. This study aims to accurately identify isabgol growing area at field level with help of progressive remotely sensed satellite data. Sentinel-2 data was used for the first crop season (2020) and the second crop season (2021) for the isabgol crop classification. Cluster to cluster comparison between satellite driven data and ground control point has been done for accuracy assessment. The producer accuracy ranged from 63.80 to 88.00% for the first crop (2020) and 70.84 to 88.89% for the second crop (2021). Our results were in sync with revenue records data (0.95 and 0.99 correlation for the first and second crop seasons, respectively). We found improved producer accuracy for the first crop over the second crop. The results shown that the time series Sentinel-2 data could be used for isabgol identification in various regions of India. The remote sensing-based methods could be used for precise estimation of isabgol crop acreage will help predict demand and supply. This information is valuable to the researchers, policy makers, pharmaceutical industries, and agronomists to accurately address issues related to import/export of isabgol and price fixation.