Spatiotemporal simulation of sustainable development based on ecosystem services under climate change

2. SSP245 natural growth scenario. In the SSP245 (moderate forcing, moderate emissions for intermediate development pathways) climate change model, LUCs from 2000–2020 were followed without setting conversion probabilities between different land types [29]. This scenario maintains the existing development trend without considering future planning or special restrictions, allowing all land types except construction and water bodies to convert into each other. This scenario represents a balanced state that neither excessively empha- sizes ecological protection nor completely disregards economic development, reflecting the actual development patterns and realities of many countries today. 3. SSP585 urban expansion scenario [30]. Under the SSP585 (high forcing, high emissions for unbalanced development pathways) climate change model, the emphasis is on economic development, with high-economic-benefit land designated as high-grade. Conversion fol- lows a principle from low to high, prioritizing the needs of urban residents for production, business, residence, and recreational land when conflicts arise between ecological and pro- duction/living land uses, aiming for maximum economic benefits. This scenario represents a pathway that is highly dependent on fossil fuels and rapid urbanization, reflecting the pressures that may be placed on the environment and ecosystems in the context of eco- nomic prioritization. 2.4. PLUS model The PLUS model [31], an improved cellular automata model built upon the FLUS model [32].

This approach allows for the prediction of LUCs in the study area by the year 2035 [33]. Using the Markov chain based on land use data from 2010 and 2015, combined with driving factor data, the areas of various land uses in the study area for 2020 were predicted [34]. The weight- ing settings for land use projections in 2035 across the three scenarios are shown in Table 1. The Kappa value of the PLUS model is 0.71, and the overall accuracy is 0.69, indicating good simulation performance. By integrating land use and driving factor data, the model provides a basis for predicting future LUC and for assessing ES and SDGs under climate change scenarios. 2.5. ESs assessment This study utilized the InVEST model to calculate four types of ESs: water yield, carbon stor- age, habitat quality, and soil retention. The detailed parameters involved in the calculations can be found in the supplementary materials, and the simplified calculation formulas are as follows. The specific formula for water yield is as follows [35, 36]: Yij ¼ 1 AETij Pi � � � Pi ð1Þ Table 1. Weighting settings for land use projections across multiple scenarios. Land use type SSP126 SSP245 SSP585 Farmland weight 0.2 0.25 0.15 Forest weight 0.4 0.3 0.2 Grassland weight 0.2 0.2 0.15 Water weight 0.05 0.05 0.05 Built-up land weight 0.1 0.15 0.4 Unused land weight 0.05 0.05 0.05 https://doi.org/10.1371/journal.pone.0316605.t001 PLOS ONE Sustainable development simulation under climate change: A spatiotemporal ecosystem services approach

PLOS ONE | https://doi.org/10.1371/journal.pone.0316605 February 4, 2025 5 / 19