Development of a PM2.5 Forecasting Dashboard for Bang Pa-In District, Phra Nakhon Si Ayutthaya using Power BI PM2.5 Forecasting and Management in Bang Pa-In District, Phra Nakhon Si Ayutthaya Province Using Statistical Models and Dashboard Development with Power BI

Abstract

This study researched the PM2.5 levels and developed an appropriate model that can predict both hourly and 24-hour average concentrations of PM2.5; it also applies descriptive epidemiological methods to study monitoring with respect to the PM2.5 health problems in Bang Pa-In District, Phra Nakhon Si Ayutthaya Province. The dataset includes hourly PM2.5 in Phra Nakhon Si Ayutthaya from August 1, 2024, through November 30, 2024, and the number of PM2.5-related diseases obtained from the Epidemiological Surveillance Database, Bang Pa-In District Public Health Office and Health Data Center, Ministry of Public Health in seven types within the period. It applies the Box-Jenkins forecasting method to select the best model, comparing it with smoothing techniques. Descriptive statistics such as proportions, minimum and maximum values, and correlation metrics like incidence rate and incidence rate ratio are used for disease surveillance.In order to identify the best models for forecasting, the Box-Jenkins method was used: ARIMA(1,1,1)(1,1,1) for hourly PM2.5 concentration forecasts was quite accurate up to 48 hours, while ARIMA(0,1,3) for the 24-hour average PM2.5 showed up to 31 days of accuracy. Moreover, exposure to high levels of PM2.5 significantly increases the risk of respiratory diseases such as influenza, pneumonia, and other respiratory diseases, as can be seen from the analysis of PM2.5 levels and associated disease surveillance, with statistically significant findings.