Abstract:

Land use and land cover change (LUCC) are the primary drivers of changes in terrestrial ecosystem carbon storage. Rapid urbanization typically induces rapid ecosystem carbon storage loss, particularly in metropolitan areas. A Thorough understanding of the driving mechanisms and design of cost-effective land-use planning and ecosystem management strategies is required. The proliferation of studies has quantified ecosystem carbon storage, for example, using the InVEST model, uncovering the spatiotemporal dynamics, investigating the driving factors, and simulating the spatiotemporal change under different LULC scenarios using various models (e.g., integrating CLUE-S with a system dynamics (SD) model (SD-CLUE-S) and Future Land Use Simulation model (FLUS)) at multiple scales (e.g., province, watersheds, and coastal zones). However, significant knowledge gaps remain, as few studies have simultaneously investigated the spatiotemporal patterns, driving forces, and management strategies of ecosystem carbon storage in urban metropolises experiencing rapid urban expansion. It is unclear how ecosystem carbon storage will respond to different scenarios of spatially explicit land-use management zoning, considering rapid urban expansion. This understanding is of great importance for better urban expansion planning and ecosystem management, but remains scarce. Focusing on the Changsha-Zhuzhou-Xiangtan Metropolitan Area, China, and employing a multi-model integration framework, this study: 1) quantified the spatiotemporal evolution patterns of ecosystem carbon storage from 1990 to 2020 using the carbon storage module of the InVEST model with carbon densities of different LULC types and LULC maps for 1990, 2000, 2010, and 2020. 2) Identified the dominant driving factors of the spatial heterogeneity of ecosystem carbon storage and their interactive impacts using the Geodetector to quantify the explanatory power (q statistic) of ten potential factors, which belong to three aspects of natural conditions (i.e., elevation, slope, surface roughness, annual average temperature, and annual precipitation), ecological factors (i.e., normalized difference vegetation index (NDVI) and habitat quality), and socioeconomic factors (i.e., GDP per unit area, population density, and nighttime light intensity). 3) We simulated land use in 2035 using the Patch-Generated Land Use Simulation (PLUS) model for three scenarios of land use and management strategies, and the quantified corresponding ecosystem carbon storage. For the Natural Development Scenario (S1), land use change that may decrease carbon density will be prohibited in the permanent basic farmland and areas within the ecological conservation redline. The construction land that usually decreases carbon density is mainly distributed within the urban growth boundary, and land use change in other regions follows the historical land use change pattern (e.g., with the same land use change pattern and land use transition probability). The Ecological-Development Coordination Scenario (S2) is similar to S1 but includes an additional policy for the Green Heat Area surrounding the three cities of Changsha, Zhuzhou, and Xiangtan. The Green Heat Area is an ecologically protected area established by the government. It has an area of 529.79 km² and two zones with different protection levels (i.e., a core protection zone and an integrated development zone). Land use change that may decrease carbon density is prohibited in the core protection zone of the Green Heat Area and is restricted to the integrated development zone outside the urban growth boundary. The Strict Protection Scenario (S3) implemented the strictest protection level, with any land use change that may decrease carbon density not allowed in the core protection zone, and the probability of land cover change that may decrease carbon density was further decreased in the integrated development zone. The results showed that: 1) 7.33×10⁶ tons of ecosystem storage carbon were lost in the study area from 1990 to 2020, mainly because of the land cover change from forest with high carbon density to urban land (i.e., residential area, industrial land, and transportation land). Ecosystem carbon storage loss displayed a dispersed spatial pattern in the 1990s, an aggregated spatial pattern in the 2000s, and a dispersed spatial pattern in the 2010s. 2) Geographic detector analysis identified habitat quality, nighttime light intensity, NDVI, and GDP per unit area as the four most important driving factors influencing the spatial heterogeneity of ecosystem carbon storage, whereas topographical factors (elevation, slope, and surface roughness) showed limited effects. The role of climate-related factors (i.e., annual average temperature and annual precipitation) in regulating ecosystem carbon storage temporally increased. Different driving factors interactively impact ecosystem carbon storage, with the strongest interactive effects between habitat quality and nighttime light intensity. 3) Ecosystem carbon storage in 2035 was simulated to decrease for all three scenarios compared to that in 2020, but with different magnitudes (i.e., 5.80×10⁶ tons for S1, 1.79×10⁶ tons for S2, and 2.01×10⁶ tons for S3). This demonstrates that targeted differentiated spatial zoning management, particularly the protection of small but critically important green core areas, can effectively mitigate carbon stock decline and generate a “spatial restructuring-functional enhancement-carbon sink gain” effect. However, the carbon storage loss in the S3 scenario was slightly higher than that in the S2 scenario, indicating that a rigid protection policy may not be the best choice, highlighting the necessity of adopting a balanced and coordinated management strategy. Based on our findings, we divided the study area into three zones (Key Protection Zone, Buffer Regulation Zone, and Low-Carbon Transition Zone) and suggested ecosystem management strategies to mitigate ecosystem carbon storage loss, considering rapid urbanization. The Key Protection Zone includes areas within the ecological red line and forests with high carbon density, where destructive activities are prohibited using a negative-list management approach. The Buffer Regulation Zone is mainly composed of permanent basic farmland and agricultural production spaces, where carbon stocks are increased by enhancing the NDVI and soil carbon sequestration capacity through constructing farmland forest networks. The Low-Carbon Transition Zone includes urban and rural residential areas, where low-impact development technologies and multidimensional land greening strategies can be implemented to enhance ecosystem carbon storage. The findings of this study can provide valuable insights for better land use planning and management to realize the goal of “Carbon Peaking Carbon Neutral” for the Changsha-Zhuzhou-Xiangtan Metropolitan Area and similar regions with rapid urbanization worldwide.

Key words: land use change, carbon storage, influencing factors, scenario simulation, ecological green heart, Chang-Zhuzhou-Xiangtan metropolitan area

摘要：

土地利用变化是影响陆地生态系统碳储量空间分布变化的主要因素，研究成长型都市圈碳储量时空格局和情景模拟，对于优化区域规划和管理、实现“双碳”战略目标具有重要意义。以长株潭都市圈为研究对象，基于1990－2020年土地利用、碳密度、生境质量等多源数据，运用InVEST、OPGD和PLUS模型，分析长株潭历史时期碳储量时空分异特征和影响因素，探究在国土空间“三线”和生态绿心实施差异化管控政策约束的前提下，预测未来3种情景碳储量的响应趋势。 结果发现，30年来，长株潭都市圈碳储量共流失7.33×10⁶t，主要由于高碳密度的生态/农业用地转移至低碳密度的城乡建设用地；碳储量空间分异呈现“中部北部低洼－四周圈层递增”特征，湘江沿岸带状洼地、生态绿心局部高地；碳储量变化区域呈“块状分散－片状集中－块状、点状分散”演变特征。生境质量指数、夜间灯光强度、地均GDP、NDVI等生态环境和社会经济要素是影响研究区域碳储量的主要驱动因子，自然气候要素的影响力呈增强趋势；“自然条件－生态环境－社会经济”各因素通过双因子增强和非线性增强影响碳储量变化，交互作用日趋复杂多元，其中生境质量与人口密度、夜间灯光强度的交互作用持续高位产生双因子增强效应。与2020年相比，生态-发展协调情景碳储量流失1.79×10⁶ t，是自然增长碳储量流失量（5.80×10⁶ t）的30.86%，严格保护情景碳储量流失（2.01×10⁶ t）量的89.05%。研究结果可为缓解长株潭都市圈碳储量流失提供决策依据和参考。

关键词: 土地利用变化, 碳储量, 影响因素, 情景模拟, 生态绿心, 长株潭都市圈