Clustering and prediction of carbon balance in production-living-ecological spaces: a county-level study in the Longzhong Loess Hilly Region, China
摘要
The Production-Living-Ecological Space (PLES) is a fundamental component of territorial spatial planning and integrated urban-rural development. Investigating the carbon balance within PLES at the county level is crucial for coordinating regional elements and optimizing spatial configurations under carbon neutrality goals. The Longzhong Loess Hilly Region (LLHR)serves as a typical ecologically fragile and agriculturally dominated area, making it highly representative for studying carbon balance dynamics in China and similar regions. This study examined 34 counties in this region, quantifying carbon sources and sinks within the PLES and analyzing spatiotemporal evolution patterns of carbon balance. Static and dynamic indicators were integrated, followed by cluster analysis with the MOS neural network model and an ARIMA–NAR hybrid model to project the carbon balance for each county by 2025. The results indicate that all counties exhibited increasing net carbon emissions, with the proportion from production space rising from 53% to 71%. Although emission reduction pressure persists across all counties, it has gradually alleviated over time. The center of carbon balance potential shifted from southeast to northwest. Carbon balance patterns were categorized into type I (High-carbon developed), type II (High-carbon traditional), type Ⅲ (Low-carbon developing), and type IV (Low-carbon transitioning), with type II being the most widespread, covering 56.5% of the total area, and type I contributing the largest share, at 37.1%, of the total emissions. Projections suggest that by 2025, while all counties will still face emission reduction challenges, most will show declining carbon emissions and improved carbon compensation rates (CCR), demonstrating positive progress toward carbon neutrality.