This study analyzes the spatial-temporal dynamics of PM2.5 air pollution in DKI Jakarta using the Geographically-Temporally Weighted Regression (GTWR) approach. The objectives of this study are to evaluate the influence of meteorological factors, compare the performance of the global OLS model with GTWR, and identify spatial-temporal heterogeneity in Central Jakarta and North Jakarta using daily data from 2024. The results show that PM2.5 concentrations in both areas exceed WHO and SNI quality standards. The OLS model has limited explanatory power (R² = 0.303), while GTWR significantly improves accuracy (R² = 0.661–0.723) with a decrease in RMSE and MAE. Local estimates show that wind speed and humidity have a negative effect, while air temperature and sunshine duration have a positive effect, with temperature as the dominant factor. Differences in optimal bandwidth between regions confirm the existence of spatial-temporal heterogeneity. This study confirms the superiority of GTWR as an analytical framework to support location- and time-based air pollution control policies.
Copyrights © 2026