ANALISIS SPASIAL-TEMPORAL PENCEMARAN UDARA PM2.5 DI PROVINSI JAWA BARAT DENGAN PENDEKATAN GEOGRAPHICALLY-TEMPORALLY WEIGHTED REGRESSION

Abstract

Air pollution remains a critical environmental issue in West Java Province, particularly due to elevated PM2.5 concentrations that pose serious risks to public health. This study aims to examine the spatial–temporal variability of PM2.5 concentrations using the Geographically Temporally Weighted Regression (GTWR) approach by incorporating wind speed, air humidity, and average temperature as predictor variables. The analysis utilizes air quality monitoring data from Bandung City and Bogor City collected between March and May 2024. GTWR is employed to capture local spatial and temporal variations that cannot be adequately explained by global Ordinary Least Squares (OLS) regression. The results indicate a substantial improvement in model performance when using GTWR, with the coefficient of determination increasing from 0.066 to 0.718 in Bandung City and from 0.068 to 0.694 in Bogor City. Wind speed demonstrates a significant negative effect on PM2.5 concentrations, while air humidity and average temperature exhibit negative but statistically insignificant relationships. These findings confirm that GTWR provides a more effective framework for modeling spatial–temporal air pollution patterns and offers enhanced insights into localized PM2.5 dynamics in urban areas..