Integration of SEM and Nonparametric Spline in Spatial Data Modeling and Visualization for Analysis of CO2 Reduction Based on Green Space in Makassar City

This study aims to analyze the characteristics of green space in reducing carbon dioxide (CO2) levels in Makassar City by integrating Structural Equation Modeling (SEM) and Nonparametric Spline. Data were collected from 251 observation points mapped using ArcGIS, along with satellite imagery taken from September 27 to October 1, 2024, to identify land use fractions such as shrub vegetation, non-shrub vegetation, roads, residential areas, and water bodies. The analyzed variables include net CO2, green space characteristics (shrub and non-shrub fractions), non-green space characteristics (residence, industry, commerce, roads, sea, and drainage), and meteorological variables (temperature, humidity and solar radiation). The results show that the best model for green space characteristics was found at the knot point of 36 with a minimum GCV of 27644.53. This model divides the area into two regions: areas with less than 36% green space and areas with more than 36% green space. An increase in green space tends to reduce CO2 levels. Meanwhile, the best model for non-green spaces was found at the knot point of 66 with a minimum GCV of 27644.18. An increase in non-green space above 66% leads to a sharper rise in CO2 levels. This study provides data-driven recommendations for urban planning and green space management, utilizing statistical modeling and spatial data visualization to support CO2 emission reduction in Makassar City