Evaluation of the Relationships Between Aerosol Optical Depth and Environmental Factors Based on Geographic and Temporal Weighted Regression Model (Case Study: Sulaymaniyah Province, Iraq) (2020-2025).

Abstract

Aerosol optical depth (AOD) is one of the important indicators for assessing aerosol load and dust conditions in arid and semi-arid regions. Considering the topographic, climatic, and environmental heterogeneity of Sulaymaniyah Province, Iraq, and its exposure to dust transport, the aim of the present study was to estimate AOD and investigate the spatiotemporal variability of the effect of environmental factors using the geographical and temporal weighted regression (GTWR) model. In this study, AOD was modeled using five explanatory variables including soil moisture, wind speed, normalized difference vegetation index (NDVI), land surface temperature (LST), and precipitation over a six-year period (2020-2025). The performance of ordinary least squares (OLS), geographical weighted regression (GWR), temporally weighted regression (TWR), and GTWR models was compared using statistical indices. Also, the spatial and temporal coefficients of GTWR were analyzed to identify heterogeneous patterns of variable effects. In comparing the models, GTWR showed the best performance with R² (0.912), RMSE (0.078), and AICc (-15376.3) and had a higher power in explaining AOD changes than other models. Spatial and temporal coefficient analysis showed that soil moisture and NDVI were the most stable factors reducing AOD, while LST was the most important factor increasing AOD. Wind speed and precipitation also showed variable spatiotemporal effects, indicating their dependence on local and temporal conditions. This pattern can be considered the result of the interaction of moisture conditions and surface vegetation, global warming, Zagros ruggedness, and dust transport processes. Overall, the findings show that GTWR, in addition to improving the accuracy of the estimation, provides the possibility of identifying controlling factors and dust-sensitive areas, and can be a scientific basis for aerosol monitoring and air quality management in Sulaymaniyah province and similar areas.