Advancing farm level crop monitoring through UAV–Sentinel 2 data integration under variable cropping scenarios of NER

Abstract Precision farming (PF) has emerged as a game-changer in agriculture, offering technological solutions to address the critical challenges of food security and climate change. However, the widespread adoption of PF faces hurdles due to the complexities of diverse cropping systems and the high costs associated with advanced ground-based instruments. To overcome this, an innovative approach was introduced in Bandia village of Assam, India by using UAV multispectral and Sentinel 2 data synergistically. The UAV imagery acquired on 17th March 2021, with ten multispectral bands (444–842 nm) was used for classifying different land use types using Object Based Image Analysis (OBIA) technique. The classification resulted into a moderately diverse cropping system with maize and rice cultivated as dominant crops occupying 45.56% and 40.87% of the total cultivated area. The diversity of the cropping pattern was further validated by ecological indices, with Shannon's Diversity Index (DI) at 1.09, Simpson's DI at 0.62, and Evenness Index at 0.78. Successively, crop above ground biomass, leaf area index and height were monitored based on the optimized Partial Least Square Regression (PLSR) model using vegetation indices from both the platforms. Cost analysis of this approach revealed a remarkable 99% cost reduction compared to traditional PF techniques. Our findings strongly suggest that the synergistic use of UAV and satellite data offers a more comprehensive view of agriculture lands, enabling high-precision monitoring of crop growth and development throughout the growth cycle and facilitating improved field level management.