Abstract
Land-use change alters terrestrial carbon storage, yet reproducible workflows for evaluating scenario-based changes remain limited in tourism-oriented World Heritage landscapes. This protocol integrates a custom Markov cellular automaton, four-pool carbon accounting equivalent to the Integrated Valuation of Ecosystem Services and Tradeoffs framework, and an optimal parameters-based geographical detector to assess land-use and carbon-storage changes across Yixian and adjacent areas of Huangshan in southern Anhui, China. China Land Cover Dataset maps from 2005, 2010, and 2015 were used for calibration and out-of-period validation. Four exploratory scenarios-Business As Usual, Tourism Expansion and Development, Ecological Conservation Priority, and Village Revitalization and Activation-were simulated for 2030 and 2050. Validation across 4,632,329 valid pixels yielded an overall accuracy of 96.61%, a Kappa coefficient of 0.850, and a Figure of Merit of 0.107. Baseline carbon storage was 59.505 teragrams of carbon, with forests contributing 95.9%. Projected carbon losses by 2050 ranged from 5.01% under Ecological Conservation Priority to 13.49% under Tourism Expansion and Development. Matched one-at-a-time perturbations supported the same scenario ordering. The optimal parameters-based geographical detector identified slope, relief, and elevation as the strongest assessed topographic associations. The supplied inputs, parameters, outputs, and scripts support reproducibility; however, the scenarios should be interpreted as comparative stress tests rather than calibrated forecasts.