Open AccessForecasting Model Validation of Particulate Air Pollution by Low Cost Sensors Data
Author(s) -
Nicoletta Lotrecchiano,
Filomena Gioiella,
Aristide Giuliano,
Daniele Sofia
Publication year - 2019
Publication title -
journal of modeling and optimization
Language(s) - English
Resource type - Journals
ISSN - 1759-7676
DOI - 10.32732/jmo.2019.11.2.63
Subject(s) - air quality index , environmental science , air pollution , particulates , pollution , context (archaeology) , pollutant , merge (version control) , air pollutant concentrations , particulate pollution , meteorology , computer science , air pollutants , geography , ecology , chemistry , organic chemistry , archaeology , information retrieval , biology
Environmental pollution in urban areas may be mainly attributed to the rapid industrialization and increased growth of vehicular traffic. As a consequence of air quality deterioration, the health and welfare of human beings are compromised. Air quality monitoring networks usually are used not only to assess the pollutant trend but also in the effective set-up of preventive measures of atmospheric pollution. In this context, monitoring can be a valid action to evaluate different emission control scenarios; however, installing a high space-time resolution monitoring network is still expensive. Merge of observations data from low-cost air quality monitoring networks with forecasting models can contribute to improving significantly emission control scenarios. In this work, a validation algorithm of the forecasting model for the concentration of small particulates (PM10 and PM2.5) is proposed. Results showed a satisfactory agreement between the PM concentration forecast values and the measured data from 3 air quality monitoring stations. Final average RMSE values for all monitoring stations are equal to about 4.5 µg/m3.