Analysis of spatiotemporal variations in carbon storage and driving mechanisms in Chongqing's karst regions under different grades of rock desertification

This study analyzed the spatio-temporal dynamics and driving mechanisms of ecosystem carbon storage in the ecologically fragile Karst Rocky Desertification (KRD) region of Chongqing, China, spanning the period from 2001 to 2020. Our core objectives were threefold: to clarify the spatiotemporal evolution patterns of ecosystem carbon storage, to identify the key driving factors and their interaction mechanisms, and to reveal the non-linear relationship and critical threshold characteristics between vegetation restoration and carbon accumulation.

To achieve these aims, a systematic methodological framework was established by integrating multi-source spatial data and advanced analytical techniques. Ecosystem carbon storage, serving as the dependent variable, was estimated using the Carbon module of the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model. This process incorporated multi-temporal Land Use/Land Cover (LULC) data and region-specific carbon density coefficients, derived from field sampling and literature review, covering four pools: aboveground biomass, belowground biomass, soil, and litter. Independent variables included spatial data layers such as the Normalized Difference Vegetation Index (NDVI), topographic factors (slope, elevation), climatic variables (precipitation, temperature), and socioeconomic indicators (GDP, population density). All data were standardized into comparable raster layers with consistent spatial resolution. For temporal trend analysis, a combination of Sen’s Slope estimation and the Mann–Kendall (M-K) test was employed. For exploring spatial driving mechanisms, an innovative integrated approach was utilized: the Optimal Parameter Geographic Detector (OPGD) was applied to objectively determine the optimal discretization of continuous variables, while Random Forest-based Partial Dependence Analysis (PDP) was used to capture non-linear relationships and threshold effects. This comprehensive framework was designed to provide a robust foundation for analyzing the complex dynamics of carbon sequestration in the region.