The combine harvesting technology which has become common in the rice-wheat system in India leaves behind large quantities of straw in the field for open residue burning, and Punjab is one such region where this is regularly happening. This becomes a source for the emission of trace gases, resulting in perturbations to regional atmospheric chemistry. The study attempts to estimate district-wise burned area from agriculture residue burning. The feasibility of using low resolution (MODIS) and moderate resolution (AWiFS) satellite data for estimation of burned areas is shown. It utilizes thermal channels of MODIS and knowledge-based approach for AWiFS data for burned area estimation. A hybrid contextual test-fire detection and tentative-fire detection algorithm for satellite thermal images has been followed to identify the fire pixels over the region. The algorithm essentially treats fire pixels as anomalies in images and can be considered a special case of the more general clutter or background suppression problem. It utilizes the local background around a potential fire pixel, and discriminates fire pixels and avoids the false alarm. It incorporates the statistical properties of individual bands and requires the manual setting of multiple thresholds. Also, a decision-tree classification based on See5 algorithm is applied to AWiFS data. When combined with image classification using a machine learning decision tree (See5) classification, it gives high accuracy. The study compares the estimated burned area over the region using the two algorithms.