Multiscale spatiotemporal analysis of carbon dioxide emissions in China from 2012 to 2021 based on kernel normalized...

To achieve carbon peaking and carbon neutrality, it is crucial to accurately understand the spatial distribution of carbon dioxide emissions (CE). However, current research primarily focuses on macro perspectives at global and national scales, making it challenging to analyze the detailed spatiotemporal changes in CE. This study combines remote sensing technology with statistical models and, for the first time, integrates the kernel normalized difference vegetation index (kNDVI) and NPP-VIIRS nighttime light data (NTL) to establish the kernel vegetation nighttime light index (kVNTL). Using this foundation, we utilized the geographically and temporally weighted regression (GTWR) model to generate 250-m resolution CE raster maps of China for the period from 2012 to 2021. The model's fit was exceptional, with an adjusted R² of 0.96. Furthermore, we employed kernel density analysis, trend analysis, hotspot analysis, and directional distribution to unveil the spatiotemporal dynamics of CE in China. The results indicate that (1) the proposed kVNTL significantly improved the accuracy of CE estimates compared to NTL. (2) From 2012 to 2021, China's CE showed uneven distribution, manifesting significant clustering. (3) Regional interactions and geographical location significantly influenced China's CE. (4) The northwestern regions of China have gradually become high CE areas.