Air quality index modelling in Indonesia based on spatial regression approach with queen contiguity weighting

National development is one of the efforts made by the state to improve all aspects of life within it. One of these aspects related to the environment that must be considered is the quality of the environment, in accordance with the SDGs. The quality of the living environment can be seen from Environmental Quality Index (EQI), which contains components of Air Quality Index (AQI), Water Quality Index (WQI), Land Quality Index (LQI) and Sea Water Quality Index (SWQI). One thing that needs to be paid more attention to is the AQI. This is because the PM2.5 (one of the air pollutants that affects air quality) in Indonesia is 34.3, which puts Indonesia in 1st place with the most polluted countries in the Southeast Asia region. The aim of this research is to model AQI in Indonesia based on a spatial regression approach. Regression modeling is used in this research because to determine the factors that influence the AQI. Furthermore, this research was approached using spatial regression because it was suspected that there would be the influence of spatial effects. In previous research, it was stated that due to Indonesia’s geographical location, there is dependence between provinces in Indonesia. Based on the results of modeling the AQI using classical regression, spatial lag regression, and spatial error regression, it was found that the best model was the spatial error regression with a coefficient of determination of 85.29% and an AIC of 169.55. Furthermore, the results of the individual spatial error regression tests show that the variables that influence the AQI in Indonesia are average air humidity, average wind speed and percentage of the number of motorized vehicles. Suggestions that can be given are for the government to improve policies in an effort to increase the AQI in Indonesia.