Sugarcane acreage estimation using Sentinel-2A satellite data and time series approach

Sugarcane is one of the most important commercial crops thus timely and precise information about area estimates of sugarcane is useful for policymakers and government planners. Remote sensing technologies have been developed for precise estimation of area and prediction of crop yield. The prospective of Sentinel-2A includes a wide range of applications in agriculture. Current study explored a remote sensing-based approach for estimating sugarcane area using Sentinel 2A in Navsari district, Gujarat. Sentinel 2A data in monthly interval (Dec-2019 to Dec- 2020) were downloaded from the website of Copernicus Open Access Hub (https://scihub.copernicus.eu/). The ground truth data was collected at different location sites in Navsari district for real time identification of sugarcane crop and validation of the image classification. The signature profiles for sugarcane crop were generated and Sentinel 2A satellite image was classified by maximum likelihood (ML) supervised classification. The Remote Sensing (RS) approach for estimation, the area estimates to 169.91 (00'ha). The time series approach was also utilized for estimating sugarcane area by analyzing 23 years of time series data from the year 1996 to 2019. To forecast future trends, employed an ARIMA (1,1,1) model, which provided reliable acreage estimation at 114.012 (00'ha). Additionally, it was compared the accuracy of combined approach with individual methods, signifying that the combined approach yielded superior results, the result stands at 141.961 (00'ha) with a lower absolute percent error of 6.57%.