A dataset for ecological protection and restoration zoning in the Qilian mountains

Abstract: This dataset provides a comprehensive spatial framework for the ecological protection and restoration of the Qilian Mountains, an essential ecological security barrier in northwestern China. The dataset is formulated based on a multi-dimensional assessment of the region's primary ecological issues, integrating 'Mountain-Water-Forest-Farmland-Lake-Grass' (Shan-Shui-Lin-Tian-Hu-Cao) elements. It utilizes multi-source data, including remote sensing imagery (Landsat, MODIS), meteorological records, soil databases, and geospatial data on human activities (e.g., mining locations), to construct a detailed evaluation index system. This system quantifies key challenges such as soil erosion, glacier retreat, grassland degradation, mining-induced damage, and agricultural land suitability. The dataset is a composite "Ecological Restoration Index" (ERI) derived from a weighted synthesis of these indicators. Based on the spatial heterogeneity of the ERI and dominant ecological problems, the Qilian Mountains planning area is categorized into distinct restoration zones. This dataset provides support for implementing targeted restoration projects, optimizing land use management, and enhancing regional ecosystem services. Key words: Qilian Mountains; ecological restoration; data-driven zoning; land use change; ecological security; indicator system Data Sources and Processing The dataset was developed through a systematic integration of multi-source geospatial and statistical data, processed to diagnose ecological issues and define restoration priorities. Basic data inputs included land use classifications for 1990 and 2015, meteorological records from the China Ground Climate Data Daily Value Dataset (V3.0) for precipitation and temperature analysis, and soil properties from the Harmonized World Soil Database (HWSD). Vegetation dynamics were assessed using 2001-2015 Landsat imagery to calculate inter-annual NDVI change rates for forests and grasslands. Human activity data was derived from statistical yearbooks (GDP) and web-scraped Point of Interest (POI) data from the Gaode Maps API to determine mine distribution density. The processing methodology involved calculating eight key indicators, including soil retention (modeled via USLE), vegetation change rates (modeled via linear trend analysis of NDVI time series), mine density, extreme precipitation, and agricultural production potential. These indicators were standardized, weighted according to their relevance to 'Mountain-Water-Forest-Farmland-Grass' components, and synthesized to produce a composite Ecological Restoration Index. This index was then used to spatially delineate homogenous restoration zones. Data results The analysis identified significant historical ecological changes and established a quantitative basis for restoration zoning. From 1990 to 2015, the planning area experienced notable land use transitions. While the total forest area remained relatively stable, shrubland and sparse forest areas declined significantly. Total grassland area increased by 3.36%, primarily driven by a large expansion of low-coverage grassland (+28.80 million ha), while medium-coverage grassland decreased (-6.58 million ha). Cropland and construction land both expanded, particularly in the eastern regions, with a 5.40% increase in cultivated land. Diagnostically, key issues include widespread mining activities causing surface fragmentation, significant glacier retreat (-37.5% from 1990-2010), and measurable grassland degradation (3.98% showing significant NDVI decline) concentrated in river valleys. Table 1 details the indicator system used for the composite assessment. Table 2 summarizes the land use conversions between 1990 and 2015, highlighting the shift in grassland quality. Table 3 presents the final 8-zone restoration plan, derived from the composite index, detailing the primary ecological function and key issues for each zone. Table 1. Ecological restoration assessment indicator system Target element Indicator Abbreviation Weight Data Source Mountain, Water Soil Retention Amount SE 0.3 USLE Model (Rainfall, Soil, DEM) Mountain Mine Distribution Density MD 0.1 Gaode POI Data Water Extreme Precipitation Days EP 0.1 Climate station data Water Mean Annual Precipitation PRE 0.1 Climate station data Forest Forestland Vegetation Change Rate FS 0.1 Landsat NDVI time series Grass Grassland Vegetation Change Rate GS 0.1 Landsat NDVI time series Farmland Agricultural Production Potential AEZ 0.1 resource and environment data center Human Gross Domestic Product (GDP) GDP 0.1 Nighttime lights & statistics Table 2. Land Use Change in the Qilian Mountains Planning Area (1990-2015) Land Use Type 1990 Area (ha) 2015 Area (ha) Net Change (ha) Key Transition Characteristics Cropland ~1,373,500 ~1,447,700 +74,200 Expansion into areas of high production potential [cite: 634, 636] Forest (Total) ~1,449,000 ~1,445,900 -3,100 Stable total, but decline in shrubland and sparse forest Grassland (Total) ~8,729,700 ~9,033,300 +303,600 Decrease in medium-cover; significant increase in low-cover Water Area ~873,000 (1990) ~869,000 (2010) -4,000 Decrease in rivers; sharp increase in lakes/reservoirs post-2000 [cite: 590-592] Construction Land (6% of 1990 mix) (32% of 2015 mix) Significant Increase Major conversion from cropland and grassland [cite: 566-568, 569] Unused Land (22% of 1990 mix) (13% of 2015 mix) Significant Decrease Converted primarily to grassland and cropland [cite: 564, 566-568] Glacier/Snow ~293,000 (1990) ~183,000 (2010) -110,000 -37.5% reduction over two decades Note: Area figures for 1990/2015 are estimated from charts; percentages for Construction/Unused Land refer to their proportion of the total transition mass. Glacier/Water data is for 1990-2010. Table 3. Ecological restoration primary zones and characteristics Zone ID Primary Zone Name Area (10k km2) Area (%) ERI (mean) Key issues & restoration focus 1 Forest-Grassland Water Conservation Zone 2.75 11.61 0.54 Vegetation degradation, extreme precipitation. Focus: wetland/headwater protection. (e.g., Datong River) 2 Forest-Grassland Water Conservation Zone 2.80 11.83 0.65 Water resource scarcity, low-cover grassland. Focus: grassland quality, wetland protection. (e.g., Dang River) 3 Desert-Grassland Ecological Governance Zone 6.11 25.78 0.56 Wind erosion, sparse vegetation, desertification. Focus: windbreak, sand fixation. (e.g., Aksai Basin) 4 Farmland-Grassland Ecological Restoration Zone 3.43 14.45 0.68 High-intensity agriculture, overgrazing, mining. Focus: mine remediation, oasis control. (e.g., North Qilian Oasis) 5 Forest-Grassland Water Conservation Zone 3.28 13.84 0.62 Soil erosion, vegetation degradation. Focus: restore forest/grass vegetation. (e.g., Qinghai Lake North) 6 Forest-Grassland Water Conservation Zone 3.62 15.26 0.50 Vegetation degradation, severe soil erosion. Focus: increase vegetation cover, protect glaciers. (e.g., Shule River Upstream) 7 Farmland-Grassland Ecological Restoration Zone 0.66 2.77 0.73 Severe soil erosion, agricultural intensification. Focus: soil conservation, improve farming efficiency. (e.g., Huangshui Valley) 8 Desert-Grassland Ecological Governance Zone 1.06 4.46 0.57 Arid, sparse vegetation, low production potential. Focus: restore vegetation, halt degradation. (e.g., NE Qaidam Basin) Data description and application This dataset represents a systematic, large-scale assessment for ecological management; however, certain limitations should be noted. The dataset's core indicators are derived from data sources with varying resolutions (e.g., 30m Landsat, 1km GDP) and time periods (e.g., 1990-2015 land use vs. 2001-2015 NDVI trends), which introduces potential scale mismatches. The mine distribution data, while innovative, is based on 2015 POI scraping and may not capture all historical or illicit mining sites. The indicator weighting scheme is based on expert knowledge for this specific restoration plan and may require adjustment for other applications. The dataset is explicitly designed to guide the implementation of targeted ecological restoration projects. For example, in Zone 4 (North Qilian Oasis), the focus is on controlling oasis scale, improving water-use efficiency, and mine remediation. In Zone 1 (Datong River), projects prioritize forest and grassland restoration, headwater protection, and glacier preservation. The dataset serves as a foundational tool for spatially allocating resources and defining specific engineering and policy interventions (e.g., grazing exclusion, afforestation, wetland restoration) across the Qilian Mountains.