Answers are grounded in this paper and cite their sources.
Classification models improve emission estimates for climate-active gases and aerosols in Eastern Siberia and the Far East.
Using machine learning methods based on attribute data from satellite information products, a classification of thermal anomalies associated with wildfires and gas flares in Eastern Siberia and the Far East was conducted. Based on the analysis of seven different machine learning models, three models with the best performance metrics were selected (Random Forest, Extreme Gradient Boosting, and Categorical Boosting). Introduction of additional features such as meteorological data and spatiotemporal characteristics of thermal anomalies improved the training performance of the selected models by 12–25% for MODIS satellite data and by 12–21% for VIIRS satellite data. Ensemble modeling approaches were used to improve classification performance. The best two ensemble models were tested on new data. The model for MODIS data correctly classified 88.6% of wildfires and 86.4% of gas flares, while the model for VIIRS data correctly classified 97.6% of wildfires and 97.2% of gas flares. The obtained results allow for the exclusion of thermal anomalies caused by gas flares, which are falsely identified as wildfires when analyzing fire activity, and to improve the accuracy of emission estimates for climate-active gases and aerosols associated with fires.
No takes yet. Share an insight, caveat, or question.
Bondur et al. (2025) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: