A Synergistic Approach for Smoke Detection, Evaluation and Retracing

Abstract: Smoke pollutant, a mixture of dust, ash, carbon dioxide, and sulfur dioxide, is a major pollution which affects the air quality significantly and is harmful to human health. Remote sensing is one of most effective ways to identify fire characteristics and monitor the behavior of smoke emitted from fire. Satellite remote sensing provides efficient approaches for smoke detection, we already developed a multi-threshold algorithm based on the analysis of the spectral characteristics of smoke, in which both MODIS (MODerate resolution imaging spectro-radiometer) RSBs (Reflective Solar Band) and TEBs (Thermal Emissive Band) are selected. Modeling has advantages of quantitative analysis in smoke study. In this paper, we propose an approach to use remote sensing and modeling synergistically so as to take the advantages of both and improve the estimation of smoke and its characteristic. In our approach, land cover type and seasonal factor are also taken into account in response to fuel change. By combining remote sensing measurements and model simulation, smoke features can be identified in more details. For testing and validation, we select examples for detailed analysis.