Interpolation biases in assessing spatial heterogeneity of outdoor air quality in Moscow, Russia

Air quality monitoring is challenging in countries where public access to data is not enabled by government agencies along with open access policies. This is especially true for Russia, where the Federal environmental monitoring network is outdated, and those systems operated by municipalities are generally under-funded. In fact, only three municipal agencies in Russia make real-time information on pollutant concentrations available online. Utilizing data from MosEcoMonitoring, we structure analytical models to infer spatial heterogeneity of nitrogen dioxide in Moscow. Model results are validated using non-governmental air quality data. The developed approach represents an open-source analytical framework for air quality evaluation and population exposure assessment on a city and district level. Both visualizations and predictive performance suggest governmental data is biased in a number of ways, highlighting the importance of ancillary data in uncovering socio-environmental implications.

Automated summary

Air quality monitoring is challenging in Russia due to the lack of public access to government data. The existing monitoring network is outdated and under-funded. The study shows that governmental data is biased and highlights the importance of using ancillary data to uncover socio-environmental implications.