土地承载了各种人类活动与自然生态系统功能，其组成与结构对于碳排放与碳吸收效应具有重要影响。快速城市化地区土地利用结构变化剧烈，建设用地扩张、生态用地减少与破碎化等土地利用变化影响了地区碳排放与碳吸收的空间布局，导致城市生态系统总体碳排放增长。目前，关于土地利用碳效应的研究中，多关注于土地利用碳效应机理、碳核算方法、影响因素分析等，但在碳效应基础上的土地利用结构优化研究还处于起步阶段。鉴于土地利用结构对区域碳效应的影响，从低碳视角进行土地利用结构优化研究，进一步探索更具碳减排效益与经济效益的土地利用结构，对城市低碳可持续发展具有重要意义。

本文构建了一套低碳视角下“碳效应核算-影响因子识别-碳生态承载/碳经济贡献耦合评价-土地利用结构优化”的土地利用碳效应核算与结构优化分析框架。以快速城市化地区珠三角城市群为案例，系统分析了2010~2020年城市群土地利用碳排放与碳吸收时空变化特征，并通过多目标规划方法探索了2030年不同土地利用结构下城市碳减排潜力与土地利用经济效益，具体研究结果如下：

（1）基于土地利用碳效应原理，构建了城市群土地利用碳核算框架。通过从能源消费、工业生产、废弃物和生物呼吸活动四个方面建立土地利用碳排放清单，结合陆地生态系统遥感模型模拟生态系统NEP，对研究区2010~2020年耕地、林地、草地、水域、建设用地、道路交通用地与未利用地的碳排放/吸收进行估算。结果表明，珠三角城市群2010~2020年期间土地利用碳排放贡献呈城乡建设用地＞道路交通用地＞耕地的特点，碳排放总量呈波动上升趋势；在空间分布特征上，广州、东莞、佛山与深圳的建设用地，以及中山与佛山的道路交通用地属于碳排放强度高值区；从时间角度分析珠三角城市碳排放变化，可将九市分为三类：稳步下降型、波动下降型和整体上升型。城市群西北部、西南部与东北部山地丘陵的林地属于碳吸收强度高值区，碳吸收总量经历了先上升后下降的过程，2020年碳吸收总量较2010年提升了1.86%，肇庆、惠州、江门与广州是城市群内主要的碳汇提供城市。基于碳排放与碳吸收差值得到城市群净碳排放，2010、2015到2020年珠三角城市群净碳排放变异系数缩小，在空间上呈现“中心倒U”型分布，惠州市净碳排放增长显著。

（2） 从土地利用视角，构建了城市群土地利用碳效应影响因素识别模型。根据改进的LMDI模型，从城市群与城市层面分解碳排放与碳吸收影响因素，并进一步分析不同时期各影响因素对碳排放/吸收的作用效果。结果表明，在研究期内碳排放影响因素作用强度排序为土地利用效率＞经济发展水平＞人口规模＞土地利用结构＞土地碳排放强度；在“十二五”到“十三五”时期，深圳、中山对城市群碳排放总量作用效果由抑制变为促进，惠州始终呈现碳排放促进特征，但各城市主要的碳排放促进作用因素不同。研究期内生态系统碳汇能力对城市群碳吸收总量的作用效果大于生态系统面积；林地碳汇能力在2016年后有显著碳吸收促进作用，草地面积与碳汇能力对碳吸收量在研究期内整体呈现积极影响。

（3）建立了城市群碳生态承载/碳经济贡献耦合协调度评价体系。从经济发展与土地利用碳效应角度构建碳生态承载力与碳经济贡献系数两个指标，并基于耦合协调度理论评价城市群碳生态承载力与碳经济贡献系数的耦合协调性。结果表明，在空间分布格局上，珠三角城市群2010年、2015年与2020年三期碳生态承载力呈现四周高中心低的空间分布特点，碳经济贡献系数则反之；根据耦合协调度理论， 2010、2015和2020年珠三角城市群整体碳生态承载/碳经济贡献耦合协调度大多呈下降趋势，整体处于轻度失调状态。

（4）构建了典型城市土地利用结构多目标优化模型，为低碳视角下城市土地利用结构调控提供参考。以广州为例，以土地利用类型为决策变量，根据广州城市发展相关规划设置约束条件，以土地利用经济效益最大化与土地利用减排潜力最大化构建目标函数，设置惯性发展、低碳优先与经济优先三种土地利用结构发展模式，探索低碳视角下的土地利用结构优化模式。结果表明， 2030年，在三种土地利用模式下，广州土地利用碳排放均有不同程度下降；经济优先模式下，2030年单位GDP碳排放强度最低，表明此模式中广州可以实现低碳-经济效益的协同最优化。基于PLUS模型模拟不同模式下2030年广州土地利用空间分布结果显示，在惯性发展模式下，从化区、增城区水域面积有被侵占的风险，在土地利用低碳优先、经济优先模式中广州生态空间得到了有效保护。

最后，在土地利用碳效应影响因素分析与结构优化模拟结果的基础上，从土地利用结构与土地利用碳效应两个维度提出土地利用管理建议，在城市功能分区规划调控、加强植被管理，提升生态系统碳汇能力，促进能源结构转型，降低土地利用碳排放强度等方面采取措施，促进城市土地利用的低碳可持续发展。

The composition and structure of land use has significant impacts on urban carbon emissions and carbon sequestration processes, as land bears various human activities and natural ecosystem functions. The expansion of urban construction land, reduction and fragmentation of ecosystem land in rapidly urbanizing areas have affected the spatial distribution of carbon emissions and carbon uptake in the region, leading to the growth of overall carbon emissions in urban ecosystems. At present, the research on the carbon effect of land use mainly focuses on the mechanism of the carbon effect of land use, carbon accounting methods, analysis of influencing factors, etc. However, the research on the optimization of land use structure based on the carbon effect is still in its infancy. Therefore, it is of great significance for urban low-carbon sustainable development to conduct research on land use optimization from the perspective of low-carbon and further explore more low-carbon and economic land use patterns.

Based on the land use characteristics in the Pearl River Delta urban agglomeration(PRDUA), this paper constructed a set of urban land use optimization framework “land use carbon accounting – factor decomposition – coupling coordination degree evaluation – land use structure optimization” from the low-carbon perspective, and discussed the carbon emission reduction potential and economic benefits under different land use modes, aiming to provide the basis for urban land use low-carbon economy collaborative optimization and urban low-carbon development. Specific research contents are as follows:

(1) Carbon accounting of land use in PRDUA: By establishing land use carbon emission inventory from energy consumption, industrial production, waste and biological respiration activities, and simulating ecosystem NEP based on remote sensing model of terrestrial ecosystem, the carbon emission/absorption of cultivated land, forest land, grassland, water area, construction land, road traffic land and unused land in the study area were estimated from 2010 to 2020. The results showed that the contribution of land use carbon emissions in the PRDUA from 2010 to 2020 was in the order of urban and rural construction land > road traffic land > cultivated land, and the total carbon emissions fluctuated and increased. The construction land in Guangzhou, Dongguan and Shenzhen, and the road traffic land in Zhongshan and Foshan belong to the areas with high carbon emission intensity. From the perspective of time, the carbon emission change of cities in PRDUA can be divided into three categories: steady decline, fluctuation decline and overall increase. The mountainous and hilly forest lands in the northwest, southwest and northeast of the urban agglomeration belong to the high-value area of carbon sequestration intensity, and the total carbon uptake volume experienced a process of first rising and then declining. In 2020, the total carbon sequestration volume increased by 1.86% compared with 2010. Zhaoqing, Huizhou, Jiangmen and Guangzhou are the main carbon sequestration providers in the urban agglomeration. Carbon emission minus absorption gives net carbon emissions, of which the variance coefficient shrank from 2010 to 2020, and it was mainly related to the changing of carbon emissions from construction land. The net carbon emission showed a “central inverted U” type distribution in spatial perspective, and the net carbon emission of Huizhou increased significantly.

(2) Identification of influencing factors of land use carbon effect in PRDUA. Based on the improved LMDI model from the perspective of land use, the influencing factors of carbon emission and carbon absorption were decomposed from the urban agglomeration and city level, and the effects of various influencing factors on carbon emission/absorption in different periods were further analyzed. The results showed that in the study period, the intensity of carbon emission was ranked as land use efficiency > economic development level > population size > land use structure > land carbon emission intensity. During the "12th Five-Year Plan" to "13th Five-Year Plan" period, the effect of Shenzhen and Zhongshan on carbon emissions changed from inhibition to promotion. Huizhou always showed the characteristics of promoting carbon emissions, but the main factors of promoting carbon emissions were different in each city. The effect of ecosystem carbon sink capacity on total carbon uptake in PRDUA was greater than that of ecosystem area. The carbon sink capacity of forest land had a significant promoting effect on carbon uptake after 2016, and the grassland area and carbon sequestration capacity had a positive effect on carbon uptake during the study period.

(3) Coupling coordination degree analysis of Ecological Support Coefficient (ESC) and Economy Contributive Coefficient (ECC) in PRDUA. Two indexes of ECC and ESC were constructed from the perspective of carbon effect of economic development and land use, and the coupling coordination between the two indicators was evaluated based on the coupling coordination degree theory. The results show that the ESC of PRDUA in 2010, 2015 and 2020 presents the spatial distribution characteristics of high around and low in the center, while the ECC is vice versa. According to the coupling coordination degree theory, the overall ESC/ECC coupling coordination degree of PRDUA in 2010, 2015 and 2020 mostly showed a downward trend, and was in a mild disequilibrium state.

(4) Study on Optimization of urban land use structure. To explore the future land use optimization model, the paper taking the core city Guangzhou as an example, the land use type was taken as the decision variable, the constraint conditions were set according to the relevant planning of urban development in Guangzhou, and the objective function was constructed by maximizing the economic benefit of land use and maximizing the emission reduction potential of land use. Three development modes of land use structure, namely inertia development, low-carbon priority and economic priority, were set up to explore the optimization mode of land use structure from the perspective of low carbon. The results showed that in 2030, under the three land use modes, land use carbon emissions in Guangzhou will decrease to varying degrees. In the economy-first mode, the carbon emission intensity per unit GDP in 2030 is the lowest, indicating that Guangzhou can realize the collaborative optimization of low-carbon and economic benefits in this mode. Based on the PLUS model, the spatial distribution of land use in Guangzhou in 2030 under different models shows that the water area of Conghua District and Zengcheng District is at risk of encroachment under the inertial development model, and the ecological space of Guangzhou is effectively protected under the low-carbon and economic land use priority models.

Finally, based on the analysis of influencing factors of land use carbon effect and the simulation results of structure optimization, it is suggested that measures should be taken to promote the low-carbon sustainable development of land use from the aspects of urban functional zoning planning and regulation, strengthening vegetation management and promoting the transformation of energy structure.