Optimizing Remote Sensing Workflows for Stubble Burning Detection Using Google Earth Engine and Cloud-Based Geospatial Analysis

Abstract

Agriculture is a cornerstone of India's economy, particularly in the Indo-Gangetic Plains (IGP). However, the rise in mechanized harvesting has led to increased crop residue, with many farmers resorting to stubble burning as a quick and economical solution for field clearance. This practice remains prevalent due to inadequate residue management infrastructure and limited alternatives, despite regulatory efforts and awareness campaigns. This study leverages satellite data from NASA's Terra and Aqua platforms to estimate burnt areas in the IGP from 2000 to 2020, focusing on key states including Bihar, Haryana, Punjab, Uttar Pradesh, and West Bengal. The methodology involves mapping spatial and temporal patterns to identify stubble-burning hotspots, seasonal trends, and peak fire activity. Monthly quantification of cropland burning provides insights into seasonal peaks, while integrating satellite-derived data with agricultural statistics enhances the accuracy of fire estimates and sheds light on the influence of farming practices on burning trends. The analysis reveals that stubble burning is strongly correlated with the Rabi and Kharif crop cycles, with the highest incidences occurring during harvest periods—specifically in April, May, September, and November. Punjab and Haryana are identified as the most affected regions. While stubble burning peaked between 2013 and 2015, recent years have seen a gradual decline, likely due to increased regulatory measures and growing environmental awareness. These findings underscore the critical need for sustained interventions to mitigate the environmental and health impacts of stubble burning and promote sustainable agricultural practices.