Sun et al. (2026) Dataset of typical terminal lake and surrounding oasis outlines in arid/semi-arid endorheic basins based on remote sensing data

Identification

• Journal: Scientific Data
• Year: 2026
• Date: 2026-02-05
• Authors: Z. T. Sun, Shijin Wang, Xinggang Ma, Yao Li, Li Zhou, Wenju Cheng, Zijian Ren
• DOI: 10.1038/s41597-026-06671-z

Research Groups

• Yulong Snow Mountain Cryosphere and Sustainable Development National Field Science Observation and Research Station, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences
• Midui Glacier-Guangxie Glacial Lake Disaster Field Science Observation and Research Station in Tibet Autonomous Region
• University of Chinese Academy of Sciences
• State Key Laboratory of Cryospheric Science and Frozen Soil Engineering, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences
• Yantai Meteorological Bureau
• Ludwig-Maximilians-University Munich
• State Key Laboratory of Ecological Safety and Sustainable Development in Arid Lands, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences
• College of Earth and Environmental Sciences, Lanzhou University

Short Summary

This study developed a comprehensive, long-term dataset of terminal lake and surrounding oasis outlines in 12 arid/semi-arid endorheic basins worldwide from 1985 to 2022, using optical remote sensing and manual corrections. The dataset provides a valuable tool for analyzing spatiotemporal variations and supporting sustainable water resource management under climate change and human activities.

Objective

• To construct a comprehensive, long-term dataset of typical terminal lake and surrounding oasis outlines in arid/semi-arid endorheic basins globally, to monitor their changes, understand water resource sustainability, and facilitate rational water resource management policies.

Study Configuration

• Spatial Scale: Global, focusing on 12 typical terminal lakes and their surrounding oases within arid/semi-arid endorheic basins (e.g., Aral Sea, Lake Chad, Ulungur Lake, Ebi Lake, East Juyan Sea, Taitema Lake, Abhe Bid Lake, Dead Sea, Humboldt Lake). Data resolution is 30 meters.
• Temporal Scale: 1985 to 2022, with data points for 1985, 1990, 1995, 2000, 2005, 2010, 2015, 2020, and 2022.

Methodology and Data

• Models used:
  • Optical remote sensing indices: Normalized Difference Water Index (NDWI), Modified Normalized Difference Water Index (mNDWI), Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), Fractional Vegetation Cover (FVC).
  • Software: ArcGIS Professional (version 3.0), ENVI (version 5.6).
  • Platform: Google Earth Engine (GEE) for image acquisition and preprocessing.
  • Algorithms: ee.Algorithms.Landsat.simpleComposite for cloud removal and composite image generation, landsat_gapfill.sav for Landsat 7 gap-filling.
• Data sources:
  • Satellite imagery: Landsat 5 Thematic Mapper (TM), Landsat 7 Enhanced Thematic Mapper Plus (ETM+), Landsat 8 Operational Land Imager (OLI) (from GEE and USGS).
  • Ancillary data: Global Aridity Index (AI) in 2000 (Version 3), WMO Basins and Sub-Basins dataset, HydroBASINS dataset, National 1:250000 three-level River Basin data set (for China), GLC_FCS30D global 30 m land-cover dynamics monitoring product (1985-2022).
  • Validation data: Google Earth Professional historical imagery.

Main Results

• A dataset containing outlines and surface areas of 12 typical terminal lakes and their surrounding oases in arid/semi-arid endorheic basins worldwide from 1985 to 2022 was constructed.
• Verification accuracy against Google Earth historical images: 89.18% for terminal lakes and 87.53% for surrounding oases.
• Kappa coefficients: 0.8569 for terminal lakes and 0.8614 for surrounding oases, indicating excellent accuracy.
• Uncertainty due to image resolution (buffer method): Area error for the vast majority of outlines does not exceed 10% of the total area, with none exceeding 25%.
• Comparison with existing lake datasets: The dataset showed good agreement with GDLS (R² = 0.9520, PBIAS = 26.07%) and GLATS (R² = 0.8794, PBIAS = 33.00%), and moderate agreement with HydroLAKES (R² = 0.6679, PBIAS = -28.55%).
• Comparison with HDCO oasis dataset (Northwest China, 2020): Kappa coefficient of 0.5965, increasing to 0.7746 after accounting for misclassification of lake areas as oases in HDCO.

Contributions

• Fills significant data gaps in existing global lake databases by providing comprehensive, long-term (1985-2022) spatiotemporal dynamics of 12 typical terminal lakes and their surrounding oases, including previously unrecorded or sparsely recorded ones (e.g., Taitema Lake, Humboldt Lake, East Juyan Sea).
• Offers a foundational dataset for investigating environmental changes, ecological impacts, and water resource sustainability in arid/semi-arid endorheic basins globally.
• Provides direct visualized data to evaluate the effectiveness of ecological restoration projects and water resource management strategies (e.g., Tarim River Basin, Heihe River Basin in China).
• Supports various research applications, including constructing sample datasets for machine learning models, serving as explanatory variables for biological population dynamics, investigating co-variation mechanisms between lake water volume and quality, and evaluating climate change impacts on water resources and oasis ecosystems.
• Serves as foundational driving data for hydrological models and for validating simulation results of hydrological processes in endorheic basins.

Funding

• Science and Technology Projects of Xizang Autonomous Region, China (grant no. XZ202502JD0004)
• Strategic Priority Research Program of Science and Technology program of Gansu Province (24JRRA104)

Citation

@article{Sun2026Dataset,
  author = {Sun, Z. T. and Wang, Shijin and Ma, Xinggang and Li, Yao and Zhou, Li and Cheng, Wenju and Ren, Zijian},
  title = {Dataset of typical terminal lake and surrounding oasis outlines in arid/semi-arid endorheic basins based on remote sensing data},
  journal = {Scientific Data},
  year = {2026},
  doi = {10.1038/s41597-026-06671-z},
  url = {https://doi.org/10.1038/s41597-026-06671-z}
}