Multi-scenario Simulation Of Land Use For Land Spatial Planning

High-quality development has become a contemporary theme in view of China’ s practical problems that seriously affect the country’ s sustainable social and economic development,such as the increasing limitation of natural resources,the degradation of natural environment and the disorder of spatial development.Territorial spatial planning organically links land elements with high-quality development,takes the optimization of territorial spatial pattern as the main form,and promotes high-quality development by alleviating potential conflicts between resource and environmental diversity in the multifunctional utilization of land.Territorial space system is a complex and dynamic space giant system produced by natural ecological factors and social economic factors.The complexity of element types and the heterogeneity of spatial forms determine the characteristics of local adaptation and diversity of planning objectives.The multi-objective scenario simulation of land use is an important tool to understand the future complex territorial space change and provide a basis for decision-making of territorial space planning.Among them,how to reasonably predict future land use changes under multiple scenarios according to the social,economic and ecological attributes of territorial space,and on this basis construct scientific and reasonable spatial layout of urban,agricultural and ecological space,guide the demarcation of three districts and three lines,and form a sustainable spatial development layout of territorial space with coordinated population,resources and environment and unified social,economic and ecological benefits? It is an important link to solve the problems facing our social economic development at present,or an urgent task to adapt to the high quality development construction.After in-depth research and analysis of existing land spatial layout optimization theories and land use simulation models,this paper uses land use multi-objective scenario simulation as a research entry point,uses information gain to establish strong and weak relationships between different planning objectives and multiple factor indicators,and uses machine learning Develop a land use scenario simulation model based on the automatic parameterization process of system elements using spatiotemporal CA spatial simulation and other technical methods,explore the multi scenario dynamic evolution mechanism of land use,and construct a technical method system for land use spatial development and layout based on the results of land use scenario simulation.An empirical study was conducted on the spatial development technology system in Chengdu,setting up three spatial transformation scenarios: ecological protection,agricultural development,and urban construction,and simulating the conversion probabilities of seven land types for each spatial grid under the three scenarios.Guided by the multi objective scenario simulation results of Chengdu in 2025,the development layout of three types of space(ecological space,agricultural space,and urban space)in Chengdu in 2025 is obtained based on the connection system between land space classification and land use.Among them,Chengdu has 9.2 %ecological space,50.63% agricultural space,and 40.17% urban space,with ecological protection space accounting for 5.47% and farmland protection space accounting for 9.24%.The spatial development layout of the three types of spaces is also in line with the strategic development plan of Chengdu.The development of regional towns is more centralized,forming a spatial pattern of "one center,multiple clusters" for coordinated regional development.The research results provide a reasonable basis for the future development of Chengdu.This study mainly constructs a land space development method under multiple scenarios from the perspectives of land space transformation on a time scale and land space optimization on a spatial scale:(1)Based on the development of land space in the historical period,choose one of the most suitable machine learning models for the research area to predict the spatiotemporal transformation mechanism of future land space,supplemented by information gain,excavate the dynamic system network relationship between land spatial pattern and multivariate environmental factors,and output a differentiated model parameter index set under three scenarios: ecological protection,agricultural development,and urban construction.At the same time,based on the development dynamics of regional resources and environment,and using the improved spatiotemporal CA transformation rules,a land use multi scene spatial transformation model based on the automatic parameterization process of system elements is constructed.(2)Starting from the main functions of the three types of space and the connection system of land use,the forest land,grassland,and water grid are merged into ecological space,the cultivated land grid is merged into agricultural space,and the urban construction land,rural residential areas,and other construction land grids are merged into urban space.Three land space development plans for ecological protection,agricultural development,and urban construction scenarios are obtained.Starting from the main functions of different land spaces,based on the layout results of multi scene land spaces,with the ecological function space under the ecological protection scenario as the priority,and the agricultural production function space under the agricultural development scenario as the secondary priority,establish rules for the delimitation of ecological space,agricultural space,urban space,ecological protection areas,and farmland protection areas,and ultimately obtain a clear controlled development area within the region,Forming a method system for optimizing the spatial layout of land.