Geometric anisotropic Semi-variogram analysis of ozone levels in Daerah Istimewa Yogyakarta, Indonesia

Abstract

Identification of areas with high O3 levels that pose a risk to public health is necessary. Ordinary co-kriging is a geostatistical method that determines the value of primary variables at specific locations using weighted values of secondary parameters. A Semi-variogram is required to demonstrate the spatial correlation between the observations measured using this method. This study aims to determine the best Semi-variogram model and produce a map of the predicted O3 level interpolation results using the ordinary co-kriging method with a geometric anisotropic Semi-variogram. Data from the first quarter of 2018’s air quality monitoring in Daerah Istimewa Yogyakarta (DIY) were used to interpolate O3 levels, with 72 points for CO levels and 53 points for O3 levels. The results showed that the Semi-variogram model with the lowest mean error (ME) value is a gaussian model that differs from the spherical model by only 0.003. The Gaussian model has the lowest root mean squared error (RMSE), but it is only 0.002 different from the spherical model. However, by comparing the mean squared deviation ratio (MSDR) values of the three models, the spherical model's MSDR value is the lowest. A comprehensive analysis showed that the spherical geometric anisotropic Semi-variogram model performed superior, resulting in the smallest minimum mean error (ME), root mean square error (RMSE) and minimum squared deviation ratio (MSDR) values. These findings highlight the potential of this approach to accurately map the spatial distribution of O3 and support evidence-based decision-making related to public health.