Satellites miss majority of stubble fires, Delhi air pollution underestimated: Report

India’s official satellite-based system for monitoring stubble burning is missing the majority of incidents, creating large gaps in assessing the impact of fires on air quality, according to a new study by Delhi-based think tank the International Forum for Environment, Sustainability and Technology (iFOREST).

The Stubble Burning Status Report 2025, released on Monday, highlights the government’s current monitoring relies mainly on two polar-orbiting satellite sensors: the Moderate Resolution Imaging Spectroradiometer (MODIS) and the Visible Infrared Imaging Radiometer Suite (VIIRS).

The report says these satellites capture only a small fraction of actual fires in Punjab and Haryana. They observe the land between 10.30 am and 1.30 pm, while farmers have increasingly shifted burning to late afternoon and evening, after the satellites have passed.

Using 15-minute data from SEVIRI, a sensor on the Meteosat-8 and 9 satellites, iFOREST found that over 90 per cent of large farm fires in Punjab in 2024 and 2025 occurred after 3 pm. By comparison, only 3 per cent of large fires occurred after that time in 2021. In Haryana, most large fires have taken place after 3 pm since 2019.

The study notes this shift has widened the gap between actual fires and those detected by MODIS and VIIRS. As a result, official datasets show a steep decline in active fire counts — over 90 per cent in Punjab and Haryana since 2021 — even though the actual burnt area has fallen at a much slower rate.

The report combines data from MODIS, VIIRS, Sentinel-2’s burnt-area mapping, and SEVIRI’s 15-minute observations.

Chandra Bhushan, CEO of iFOREST, said the mismatch between fire timing and satellite overpass times has created a structural limitation.

“Farmers have shifted burning to the late afternoon, while our monitoring relies on satellites that capture active fires only during a narrow time window — 10.30 am to 1.30 pm. The result is a massive underestimation of fires, emissions and their contribution to air pollution in Delhi,” he said.

Despite the monitoring gaps, burnt-area mapping indicates a real, albeit uneven, decline in stubble burning. In Punjab, the burnt area fell from a peak of 31,447 sq km in 2022 to about 20,000 sq km in 2025. In Haryana, it dropped from 11,633 sq km in 2019 to 8,812 sq km in 2025, though trends have fluctuated in recent years.

The report warns that relying solely on active fire counts gives an incomplete picture and has led to underestimation of stubble burning’s contribution to Delhi’s pollution this year, as most fires occurred outside satellite observation hours.

The findings suggest India’s stubble-burning management policies need adjustments, not just in Punjab and Haryana but in other states as well. Active fire-count data shows a rise in burning in Uttar Pradesh and Madhya Pradesh.

The study recommends that the Consortium for Research on Agroecosystem Monitoring and Modeling from Space (CREAMS), which runs the national monitoring system, begin publishing burnt-area data alongside fire counts. It also suggests that Delhi’s Decision Support System for air-quality forecasting revise its methodology to account for missed fires.

Ishaan Kochhar, programme lead at iFOREST, said India needs a monitoring system that integrates high-resolution burnt-area mapping with geostationary observations. “Policy decisions are being shaped by incomplete information. Monitoring must be updated to match the changing pattern of stubble burning,” he said.