Answers are grounded in this paper and cite their sources.
Abstract Particulate matter in particular poses a serious risk to both human health and the environment. For efficient pollution management and the deployment of appropriate mitigation techniques, accurate air quality forecasting is essential. The objective of the proposed work is to forecast the concentration of PM10 (particulate matter with a diameter of 10 micrometres or less) using a variety of air pollutants as predictors, such as PM2.5, NO2, CO, and other pertinent chemical species. Proposed system makes use of a sizable dataset that includes historical measurements of PM10, PM2.5, NO2, CO, and numerous other air contaminants. It is common knowledge that these pollutants are significant sources of air pollution, and metropolitan areas regularly monitor these pollutants' concentrations. The accuracy of PM10 prediction is increased by using numerous air pollutants as predictors because the intricate interactions and synergistic effects among various pollutants can be captured. This study proposes the development of a system, inspired by existing machine learning and deep learning techniques, capable of forecasting the concentration of PM10 in the atmosphere, using data about other pollutants, as mentioned previously, in the air in the region of prediction, along with meteorological factors defining the weather in the region of prediction, including but not limited to relative humidity, wind speed, wind direction, average temperature and barometric pressure. Before using data from trusted sources (elicited in the references) to build machine learning and deep learning models, the relationships of various features within the dataset were studied in comparison with the target variable – concentration of PM10. In the work done, it has been observed that the Machine Learning model based on Decision-Tree architecture, Extra Trees Regressor, delivered a 93% R2 score, with an RMSE of less than 20 parts per million with respect to PM10 concentration in the air. With this proposed system, multiple advantageous technologies emerging in the domain of air pollution control and mitigation can be built, such as early warning systems for administrative and meteorological departments; and improvement of this system to expand the forecast horizon and integrate the forecasting framework are some valuable future improvements which can be added to this work.
No takes yet. Share an insight, caveat, or question.
Pandita et al. (2024) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: