How can scientific methods improve the accuracy of burnt-area estimates in stubble burning cases?

There is ongoing debate about the reliability of burnt-area estimates related to stubble burning. I want to explore how technology and scientific analysis can address anomalies and improve reporting.