A Geospatial dataset for systematic ecological problem diagnosis and comprehensive restoration zoning in Shaanxi province, China

Abstract: This dataset provides a systematic, multi-dimensional assessment of the ecological status of Shaanxi Province, China, based on the "Mountains, Rivers, Forests, Farmlands, Lakes, and Grasslands" (MRFFLG) "life community" concept. Derived from multi-source geospatial, meteorological, remote sensing, and socio-economic data centered around 2015, this dataset integrates 8 core diagnostic indicators: Soil Conservation, Mine Distribution Density, Extreme Precipitation, Annual Precipitation, Forest Vegetation Change, Cropland Quality, Grassland Vegetation Change, and Poverty Level. These indicators are synthesized into a composite Ecological Restoration Index (ERI) and used to delineate 8 distinct comprehensive restoration zones via spatial clustering. Key findings reveal the provincial average ERI was 0.39 in 2015, with watershed values ranging from 0.23 to 0.60. The ERI exhibits a significant spatial gradient, decreasing from the ecologically robust south (Shaannan) to the ecologically vulnerable north (Shaanbei). The dataset identifies differentiated ecological problems: the northern Loess Plateau (Zones 4 and 6) is dominated by soil erosion, vegetation degradation, and water scarcity, while the southern Qinba Mountains (Zones 1, 5, 8) are primarily affected by extreme precipitation and soil erosion. This dataset provides a scientific basis for moving beyond single-element management and supporting the targeted, systemic implementation of ecological protection and restoration projects. Key words: ecological problem diagnosis; systematic restoration of mountains, rivers, forests, farmlands, lakes, and grasslands; ecological conservation and restoration zoning; Shaanxi Province 1. Introduction The concept of "Mountains, Rivers, Forests, Farmlands, Lakes, and Grasslands" (MRFFLG) as an interconnected "life community" has become a guiding principle for ecological civilization in China. This approach necessitates a shift away from traditional, fragmented, single-element management—where forestry, water management, and agriculture operate in isolation—toward integrated and systemic governance. Effective ecological restoration under this paradigm requires a comprehensive diagnosis of interconnected problems and the scientific delineation of management zones based on these systemic issues. However, quantitative methods for such a comprehensive diagnosis and zoning remain relatively underdeveloped. This paper describes a geospatial dataset created to address this gap, using Shaanxi Province as a case study. The dataset provides a replicable methodology and a baseline for diagnosing regional ecological problems and partitioning the landscape into coherent zones for targeted restoration. 2. Methodology and Data Sources 2.1. Study area The study area is Shaanxi Province, located in the inland heartland of China. The province has a long, narrow geography and is divided into three distinct natural zones by the Beishan and Qinling Mountains: the northern Loess Plateau (Shaanbei), the central Guanzhong Plain, and the southern Qinba Mountains (Shaannan). This geography creates significant climatic and ecological heterogeneity, making it an ideal region for developing a differentiated zoning methodology. 2.2. Indicator system A diagnostic indicator system was constructed based on the MRFFLG framework, encompassing 8 key indicators to assess the status of mountains, water, forests, farmlands, grasslands, and human well-being (Table 1). Soil Conservation (SC), representing mountain, water, and farmland issues, was calculated using the Universal Soil Loss Equation (USLE). This involved modeling rainfall erosivity (R-factor) using the Wischmeier formula, soil erodibility (K-factor) using the EPIC formula, and calculating topographic (LS), vegetation cover (C), and engineering practice (P) factors. Mine Distribution Density (MD), representing a key pressure on mountain systems, was derived from Point of Interest (POI) data. A Python-based crawler was used to retrieve "mine" POIs from the GaoDe Map API, which were then cleaned to remove irrelevant points (e.g., "Mine Road") before density analysis. Extreme Precipitation (EP) and Annual Precipitation (Pre) were calculated from the China Ground Climate Data Daily (V3.0) dataset. EP was defined as the number of days with daily precipitation 50 mm. Forest Vegetation Change (FS) and Grassland Vegetation Change (GS) were assessed using the 2001-2015 MODIS NDVI (MOD13A3) 1km dataset. The interannual change rate () was calculated for each pixel using a linear regression, with an F-test to determine the significance of the trend. Cropland Quality (QC), representing farmland, was derived from the 1:1,000,000 Land Resource Map. Poverty Level (GDP), representing the human element, was estimated using 2013 DMSP-OLS nighttime light (NTL) data. To correct for NTL saturation in urban centers, a desaturated Enhanced Administrative Nighttime Light Index (EANTLI) was developed by integrating MODIS EVI data. This was then aggregated to a regional average index (ANLI) to represent economic status. Table 1: Diagnostic indicators of ecological problems in Shaanxi Province. Type Ecological problem Indicator name Indicator direction Data source Mountain Soil erosion Soil conservation + 1. China Ground Climate Data Daily (V3.0) 2. Soil organic matter content and soil texture data from the Harmonized World Soil Database (HWSD v1.1) Mine exploitation Mine distribution density – Water Extreme precipitation(Uneven precipitation distribution) Days of heavy rain – China Ground Climate Data Daily (V3.0) Annual precipitation + Forest Sparse vegetation(Poor forest quality) Interannual variability of forest vegetation + 1. 2001-2015 1km MODIS monthly composite VI product (MOD13A3) NDVI 2. 2015 forest distribution data Farmland Farmland quality Cropland quality – 1:1,000,000 Land Resource Map Grassland Grassland degradation Interannual variability of grassland vegetation + 1. 2001-2015 1km MODIS monthly composite VI product (MOD13A3) NDVI 2. 2015 grassland distribution data Economic status Poverty status GDP + 1. 2013 DMSP-OLS stable nighttime light data, 1km res. 2. 2013 1km MODIS monthly composite VI product (MOD13A3) EVI 2.3. Comprehensive evaluation and zoning A comprehensive Ecological Restoration Index (ERI) was constructed using a linear weighted function (). All 8 indicators were normalized and standardized. Weights ) were assigned based on the number of MRFFLG elements an indicator represents: Soil Conservation (0.3); all 7 other indicators (0.1). A lower ERI value indicates a poorer ecological condition and a higher necessity for restoration. Finally, the 8 distinct ecological restoration zones were delineated using the ArcGIS Grouping Analysis tool, which performed a spatial clustering analysis at the watershed level based on the 8 indicator values. 3. Data description and results The final dataset includes 7 periods of LUCC data (1970-2015), 8 core indicator raster layers (2015), the composite ERI raster, and a vector file of the 8 final restoration zones. 3.1. Diagnostic indicator characteristics The 8 core indicators reveal strong spatial heterogeneity across Shaanxi. Key quantitative and qualitative characteristics of these data layers are summarized in Table 2. There is a clear south-to-north gradient, with precipitation, soil conservation, and vegetation health generally decreasing towards the north. Vegetation change from 2001-2015 was positive overall, but the dataset identifies key areas of degradation. Table 2: Summary of core diagnostic indicator data layers. Indicator layer Metric(s) Value(s) Spatial distribution characteristic Soil Conservation (SC) Total Annual 75.42 Increases from Northeast to Southwest. Low in Shaanbei, high in Shaannan. Annual Precipitation (Pre) Range 300–1700 mm Decreases from South to North. Shaannan is wettest, northern Shaanbei is driest. Extreme Precipitation (EP) Range (days 50mm) 0–6 days Decreases from South to North. Highest in southern Shaannan (e.g., Zhenping). Forest Change (FS) Trend (2001-15) = 0.0044 0.20% of area declined. Decline mainly in low-elevation Guanzhong Plain. Grassland Change (GS) Trend (2001-15) = 0.0053 0.33% of area declined. Decline mainly in Guanzhong Plain & Hanjiang Valley. Cropland Quality (QC) Grades 1–4 Grade 1 (high quality) in Guanzhong Plain. Grade 4 (low quality) in Shaanbei & Shaannan. Poverty Level (GDP) County Classification 70.1% "poverty areas" Richest areas in Guanzhong (Xi'an, Xianyang). Poorest in high-altitude Shaannan. Mine Distribution (MD) (Density) - Concentrated in the central Guanzhong Plain; scattered in Shaanbei & Shaannan. 3.2. Ecological restoration index and zones The composite ERI for 2015 had a provincial average of 0.39, with watershed values ranging from 0.23 to 0.60. The spatial distribution of the ERI is distinct, showing a clear decrease from south to north, confirming that the most ecologically vulnerable areas requiring restoration are concentrated in the northern Loess Plateau. The clustering analysis produced 8 distinct zones, summarized in Table 3. Table 3: Spatial characteristics of the eight ecological restoration zones Zone ID Geographic location Area (10,000 km2) Area percentage (%) Avg. Eri 1 Shaannan Qinling eastern region 1.045 5.084 0.448 2 Guanzhong plain central urban area 0.557 2.710 0.382 3 Hanjiang river valley basin 0.626 3.046 0.433 4 Shaanbei Loess Plateau central and southern Region 3.953 19.232 0.309 5 Shaannan qinling mountains 4.406 21.436 0.453 6 Shaanbei loess plateau northern region 3.906 19.004 0.255 7 Guanzhong plain area 4.648 22.614 0.349 8 Shaannan daba mountains area 1.413 6.875 0.557 Each zone possesses a unique ERI average and area, and more importantly, a specific profile of dominant ecological problems, which is detailed in Table 4. Table 4: Dominant ecological problem profiles by restoration zone Zone ID Geographic location Dominant ecological problems identified 1 Shaannan Qinling eastern region High soil erosion; uneven precipitation; high poverty. 2 Guanzhong plain central urban area Forest/grassland degradation; high mining intensity; soil erosion. 3 Hanjiang river valley basin High susceptibility to extreme precipitation events. 4 Shaanbei Loess Plateau central and southern Region Drought; high soil erosion; poor cropland quality; vegetation degradation. 5 Shaannan qinling mountains Vegetation degradation; high impact from extreme precipitation; soil erosion. 6 Shaanbei loess plateau northern region (Lowest ERI) Drought; severe vegetation degradation; slow recovery; high mining intensity. 7 Guanzhong plain area Forest/grassland degradation; high mine distribution; soil erosion. 8 Shaannan daba mountains area (Highest ERI) High poverty; high susceptibility to extreme precipitation and soil erosion. 4. Data limitations and application This dataset provides a robust, quantitative, and spatially explicit baseline for ecological management in Shaanxi. However, several limitations should be noted. The assessment is primarily a static snapshot centered on 2015 and does not capture the dynamic, non-linear processes of ecological change. The ERI weighting scheme, while systematic and based on the MRFFLG framework, is one of several possible approaches; alternative weighting methods could produce different spatial ERI distributions. Furthermore, some indicators are based on 1km resolution data (e.g., NDVI, NTL), which may generalize or mask fine-scale ecological degradation. The POI-based mining data identifies hotspots but may not capture the full spatial extent of mining impacts, particularly from historical or informal sites. The primary application of this dataset is to support the transition from traditional, single-element ecological management to a systemic, integrated "life community" approach. The 8 distinct zones and their specific problem diagnoses (Table 4) are designed to directly inform "one-zone, one-policy" restoration strategies. The ERI map serves as a tool for prioritizing restoration investments, especially in low-ERI watersheds (e.g., in Zones 4 and 6). This dataset and its methodology provide a replicable scientific framework for other regions seeking to implement and operationalize the MRFFLG concept for targeted ecological restoration.