Estimating cropland fire emissions with remote sensing: A literature review and data strategy

Abstract

Cropland fires are a major source of greenhouse gases and air pollutants, with consequences for climate change, air quality, and agricultural sustainability. Traditional estimation methods, such as ground measurements and statistical models, often face limitations related to cost, spatial coverage, and temporal resolution. Remote sensing provides a powerful alternative by enabling consistent and large-scale monitoring of fire dynamics. This review synthesizes advances in remote sensing approaches for cropland fire emission estimation, including burned area mapping, active fire detection, fire radiative power (FRP), and integrated multi-sensor techniques. Particular attention is given to their strengths, limitations, and the persistent challenges of detecting small fires, discriminating burned areas from tilled soils, and addressing uncertainties in emission factors. Building on this analysis, we outline a data strategy that leverages multi-source satellite observations, crop type classification, and emission modelling to enhance accuracy. The review highlights the potential of remote sensing to improve emission inventories, inform mitigation strategies, and support climate policy, while identifying opportunities for innovation through machine learning, hyperspectral data, and integrated monitoring frameworks.