Elucidating carbon emission responses to land-use transition using the Kaya–LMDI model: a case study of Hainan, China

Land-use transition (LUT), a pivotal vector for anthropogenic intervention in the carbon cycle, profoundly influences the formation and evolution of regional carbon emission patterns. This study focuses on Hainan, China's sole tropical island, and establishes a model accounting for carbon emissions associated with LUT based on related remote-sensing data, socioeconomic statistics, and energy consumption-related data between 2000 and 2025. We combine spatial autocorrelation analysis, an extended logarithmic mean Divisia index decomposition model, and the Tapio decoupling model to systematically elucidate the spatiotemporal features of LUT-associated carbon emissions, their driving factors, and their decoupling relation with economic growth. Notably, Hainan Province features an LUT involving decreasing and increasing proportions of carbon-sink land and carbon-source land, respectively, with construction land expansion being the primary transition mode driving carbon emission growth. The associated carbon emission response features a spatial differentiation pattern of high values concentrated in the north and west and low values localized in the south and east. In addition, carbon sources and sinks demonstrate considerable spatial agglomeration. Economic output is the core driver promoting carbon emission growth, with improvements in land-use efficiency and energy intensity being critical for carbon emission mitigation. During the examined period, the correlation between LUT-associated carbon emissions and economic growth evolves from weak to strong decoupling, demonstrating the remarkable efficacy of peak carbon and carbon neutrality goals in guiding emission reduction-focused LUT. Overall, this research provides a scientific basis for coordinating LUT and low-carbon development in the Hainan Free Trade Port initiative.