Ecological risk assessment of land use change in the North China Plain: key data and characteristic analysis

Ecological risk assessment of land use change in the North China Plain: key data and characteristic analysis Wei Song¹, *, Xuyang Zhang² 1. Key Laboratory of Surface Pattern and Simulation, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101; 2. School of Earth and Environment, Anhui University of Science and Technology, Huainan 232001, Anhui Province Abstract: This dataset is based on multi-source data of land use, meteorology, and social economy in the North China Plain from 2000 to 2020. It integrates core information including land use change (LUC) probability, ecosystem services (ES) gains and losses, and the Sharpe ratio-based ecological risk index (ERI), providing fundamental support for precision governance of regional ecological risks. From 2000 to 2020, cultivated land in the North China Plain was the dominant land use type (accounting for 56.52%), while construction land saw the largest increase (+4.01%). Under three scenarios (natural development, urban expansion, and ecological protection) in the North China Plain in 2030, the average ERI values are 0.4718, 0.4719, and 0.4703 respectively, with the urban expansion scenario carrying the highest risk. High-risk areas in the North China Plain are concentrated in the Beijing-Tianjin-Hebei region and the coastal areas of the Bohai and Yellow Seas, while low-risk areas are distributed in the foothills of the Yanshan and Taihang Mountains. This dataset can be used for the delimitation of urban development boundaries, ecological protection planning, and research on land use optimization under the "dual-carbon" goals. Key words: North China Plain; land use change; Sharpe ratio; ecological risk index; CA-Markov model; ecosystem services 1. Data Sources and Processing 1.1 Basic Data (1) Core driving data Land use data: Derived from the Resource and Environment Science Data Center, including three periods (2000, 2010, and 2020) with a resolution of 30m, classified into 6 types: cultivated land, forest, grassland, water area, construction land, and unused land. Digital Elevation Model (DEM): Obtained from the Geospatial Data Cloud. Meteorological data: Data on precipitation, temperature, and evapotranspiration from 2000 to 2020 provided by the China Meteorological Science Data Sharing Service Network. (2) Auxiliary data Socio-economic data: 1 km resolution GDP and population grid data, as well as vector data of roads and railways, reflecting the intensity of human activities. Soil data: Data on soil types and erosion types provided by the International Institute for Applied Systems Analysis (IIASA) with a resolution of 1km, used for calculating soil conservation services. 1.2 Data Processing Methods (1) Prediction of land use change probability LCM (Land Change Modeler) model: Via the sensitivity module of the TerrSet platform, combined with the multi-layer perceptron (MLP) algorithm, it generates transition potential maps for 2020-2030. CA-Markov model: Couples Cellular Automaton (CA) and Markov chain to simulate LUC under three scenarios (with a Kappa coefficient of 0.84 and an accuracy of 0.90). The formula is as follows: (1) where An denotes the land use state in the n-th period, and Pij represents the transition probability matrix from land type i to j. (2) Ecosystem service assessment The InVEST model was used to calculate 4 key services (water yield (WY), habitat quality (HQ), carbon storage (CS), and soil conservation (SC)). (3) Calculation of ecological risk index Based on the modified Sharpe ratio formula, it couples LUC probability (P) with ES gains and losses. 1.3 Scenario setting S1 (natural development): continues the LUC trend from 2000 to 2020. S2 (urban expansion): increases the probability of transition from cultivated land and grassland to construction land, and restricts the reverse transition of construction land. S3 (ecological protection): protects forests and grasslands, and restricts the unordered expansion of cultivated land and construction land. 2. Core Data Results 2.1. Land Use Change Data of the North China Plain Table 1 Changes in Area and Proportion of Land Use Types in the North China Plain from 2000 to 2020 Land Use Type 2000 2010 2020 Percentage change (%) Cultivated land 329313.10km²(59.66%) 321026.58km²(58.16%) 311960.99km²(56.52%) -3.14 Forest 83608.08km²(15.15%) 83184.94km²(15.07%) 83638.35km²(15.15%) 0.00 Grassland 60194.25km²(10.91%) 52732.67km²(9.55%) 52770.13km²(9.56%) -1.35 Water 17294.17km²(3.13%) 18890.62km²(3.42%) 22198.61km²(4.02%) +0.89 Construction land 56876.22km²(10.30%) 74078.33km²(13.42%) 78978.47km²(14.31%) +4.01 Unused land 4652.56km²(0.84%) 2023.41km²(0.37%) 2373.13km²(0.43%) -0.41 2.2. Multi-scenario Land Use Prediction Data for 2030 Table 2 Core Characteristics of Multi-scenario Land Use Transition in the North China Plain in 2030 Scenario Main Transformation Types Proportion of Key Type Change (%) High-Transition Region S1 (Natural Development) Cultivated land → Construction land, Grassland → Construction land Construction land + 0.46, Cultivated land - 0.21 Northern part of the North China Plain, Southwestern part of the North China Plain S2 (Urban Expansion) Cultivated land → Construction land, Forest → Grassland Construction land + 1.31, Forest - 0.24 Beijing-Tianjin-Hebei, Jinan-Zhengzhou cities S3 (Ecological Conservation) Construction land → Forest, Unused land → Cultivated land Forest + 0.41, Grassland + 0.32 Foothills of the Taihang Mountains and Dabie Mountains 2.3. Multi-scenario Ecological Risk Data Table 3 Characteristics of Multi-scenario Ecological Risk (ERI) in the North China Plain in 2030 Scenario ERI Mean Value Risk Level Proportion (%) High-Risk (HR) Core Zone S1 (Natural Development) 0.4718 HR:13.44,MHR:20.50,MR:30.25 Yellow Sea and Bohai Sea coast, Yantai S2 (Urban Expansion) 0.4719 HR:14.40,MHR:17.30,MR:32.94 Beijing - Tianjin, Jinan - Qingdao S3 (Ecological Conservation) 0.4703 HR:13.48,MHR:23.20,MR:27.45 Only in parts of the Yellow Sea and Bohai Sea Note: MHR = medium-high risk, MR = medium risk, MLR = medium-low risk, LR = low risk 3. Data Description and Application 3.1. Data Limitations (1) LUC prediction relies on the CA-Markov model, which is only based on historical data and does not incorporate dynamic factors such as climate policies and extreme climates. (2) The ecological risk assessment does not cover non-LUC-driven risks such as soil pollution and biological invasion, which may underestimate the comprehensive risk. (3) The meteorological data has a resolution of 1 km, making it difficult to capture the impact of local microclimate on ecosystem services. 3.2. Data Application Directions (1) Support the delimitation of urban development boundaries: combine the high-risk areas (HR areas) in S2 (e.g., Beijing-Tianjin-Hebei region) to restrict the expansion of construction land and avoid encroachment on ecological space. (2) Serve the zoned governance of ecological risks: conduct centralized restoration in HR areas (Bohai and Yellow Seas), promote the ecological transformation of agriculture in MR areas (central and southern Shandong), and construct ecological corridors in LR areas (Yanshan foothills). (3) Assist in cultivated land protection and ecological coordination: based on the S3 scenario, optimize the layout of "cultivated land-ecological land" to balance food security and ecological security.