Digitalization of environmental monitoring of urbanized territories: an integrated approach to air quality assessment based on heterogeneous data sources

Abstract

Rapid urbanization and the concomitant deterioration of atmospheric air quality determine the need to improve the methodology of environmental monitoring in order to make timely and informed management decisions. The present study is aimed at developing and validating a comprehensive air quality assessment system for urbanized areas that integrates data from diverse sources: stationary observation posts, mobile sensors, remote sensing systems, and crowdsourcing platforms. The methodology is based on a multiparametric approach that combines spatial and temporal modeling of the spread of pollutants, machine learning algorithms for integrating heterogeneous data, and geostatistical analysis for validating the results. The empirical base consists of data from 27 cities of the Russian Federation with different climatic conditions and pollution patterns, collected during 2019-2023. The results indicate a significant increase in the accuracy of predicting pollutant concentrations using the integration model (the average forecast error was reduced by 43.7% compared with traditional methods). The developed calibration algorithm ensures the correct interpretation of low-budget sensor data, which expands the possibilities for decentralized monitoring. The proposed visualization system makes it possible to generate dynamic pollution maps with high spatial resolution (up to 50 m), which provides detailed identification of problem areas and modeling of pollution propagation scenarios. The results obtained form the methodological and technological basis for the development of a digital environmental monitoring system that meets modern environmental quality management requirements and the principles of sustainable development.