Bai et al. (2025) The Shrinkage of Lakes on the Semi-Arid Inner Mongolian Plateau Is Still Serious

Identification

• Journal: Water
• Year: 2025
• Date: 2025-10-24
• Authors: Juan Bai, Yue Zhuo, Naichen Xing, Fuping Gan, Yi Guo, Baikun Yan, Yichi Zhang, Ruoyi Li
• DOI: 10.3390/w17213056

Research Groups

• China Aero Geophysical Survey and Remote Sensing Center for Natural Resources, China Geological Survey
• Key Laboratory of Aerial Geophysics and Remote Sensing Geology, Ministry of Natural Resources
• State Key Laboratory of Remote Sensing Science, School of Geography, Beijing Normal University

Short Summary

This study developed a remote-sensing-based framework to monitor long-term lake water storage (LWS) changes in the semi-arid Inner Mongolian Plateau (IMP), revealing a net decline of 1.21 Gt between 2000 and 2021 with distinct regional and temporal shifts driven by both climatic factors and anthropogenic activities.

Objective

• To develop a novel remote-sensing-based framework for reconstructing historical lake water levels and estimating water level changes in lakes without altimetry data.
• To quantify the long-term spatiotemporal variations in lake water storage (LWS) for 109 lakes (≥5 km²) on the Inner Mongolian Plateau (IMP) from 2000 to 2021.
• To identify and analyze the key climatic and anthropogenic drivers of observed LWS changes across different hydrological subregions of the IMP.

Study Configuration

• Spatial Scale: Inner Mongolian Plateau Lake Zone (IMP), covering 1.96 million km². The study focused on 109 lakes with a historical maximum area of 5 km² or more, categorized into three hydrological subregions: western and southern IMP, central and eastern IMP, and northeastern IMP.
• Temporal Scale: 2000–2021 (22 years), with monthly data for unfrozen seasons (May–October).

Methodology and Data

• Models used:
  • Novel lake monitoring framework (reconstructs historical lake level time series and estimates water level variations in lakes without altimetry data).
  • Logarithmic water level–surface area correlation (for reconstructing missing water levels).
  • Similarity-based parameter regionalization (for estimating water levels in unobserved lakes).
  • Trapezoidal cylinder volume equation (for calculating lake water storage variation).
  • Water balance method (for estimating indirect recharge).
  • Model simulation (Complementary relationship principle) for Evapotranspiration (Eta).
• Data sources:
  • Satellite Imagery: Landsat-5/7/8, Sentinel-2 (for lake surface area and Land Use/Land Cover Change - LUCC).
  • Satellite Altimetry: ICESat/GLAH14, ICESat-2/ATL13, CryoSat-2 (SIN+LRM) (for water level changes).
  • Ancillary Datasets:
    • China Lake Dataset (1960s–2020) (for lake boundary information).
    • Hydroweb data (for validation of water level and water storage changes).
    • In-situ observations of lake level and water storage (from Cetian Reservoir, for validation).
    • Precipitation data (fusion of Global Precipitation Measurement (GPM) and in situ data).
    • GRACE/GRACE-FO (for Terrestrial Water Storage Anomalies - TWSA).
    • National Statistical Yearbook (for cultivated land area and groundwater reserves).

Main Results

• The developed framework accurately reconstructed water levels (Root Mean Square Error (RMSE): Hulun Lake = 0.46 m; Dalinor Lake = 0.22 m; Cetian Reservoir = 0.56 m) and lake water storage (LWS) changes (RMSE: Hulun Lake = 0.63 Gt; Dalinor Lake = 0.02 Gt; Cetian Reservoir = 0.009 Gt).
• Among the 109 studied lakes (≥5 km²), 49 exhibited water level increases, 40 showed decreases, and 20 experienced minimal changes (below ±0.20 m). Decreasing trends dominated central and eastern regions, while increasing levels prevailed elsewhere.
• The total LWS on the IMP experienced a net decline of 1.21 Gt from 2000 to 2021, averaging 0.06 Gt/yr. This trend included a distinct shift: a loss of 10.82 Gt from 2000 to 2012, followed by a recovery with an increase of 9.61 Gt from 2013 to 2021.
• Spatially, significant LWS reductions were concentrated in the central and eastern IMP (total loss of 1.31 Gt), primarily driven by intensive water diversion, groundwater exploitation, and agricultural expansion.
• LWS increases were observed mainly in the western and southern IMP (total gain of 0.65 Gt), attributed to enhanced precipitation, reduced aridity, and ecological conservation projects.
• The northeastern IMP experienced an initial decline (9.63 Gt loss from 2000–2012) followed by a recovery (9.08 Gt gain post-2012), influenced by precipitation deficits and ecological replenishment projects.

Contributions

• Developed a novel remote-sensing-based framework for continuous monitoring of lake water levels and storage, particularly for lakes lacking direct altimetry data, by reconstructing historical time series and estimating levels based on similarity-based regionalization.
• Provided the first comprehensive, long-term (2000–2021) spatiotemporal quantification of LWS variations for 109 lakes (≥5 km²) across the entire Inner Mongolian Plateau, addressing a previous data gap.
• Identified and attributed the complex, regionally distinct drivers of LWS changes (climate vs. anthropogenic activities) across the IMP, offering improved understanding of lake dynamics in semi-arid China.
• Offers technical guidance and theoretical foundations for sustainable water resource management and ecological protection in semi-arid regions.

Funding

• China Geological Survey: "Remote Sensing Quantitative Survey and Monitoring of Water Cycle Elements and Natural Resources in a River Basin" (Grant No. DD20230500106).

Citation

@article{Bai2025Shrinkage,
  author = {Bai, Juan and Zhuo, Yue and Xing, Naichen and Gan, Fuping and Guo, Yi and Yan, Baikun and Zhang, Yichi and Li, Ruoyi},
  title = {The Shrinkage of Lakes on the Semi-Arid Inner Mongolian Plateau Is Still Serious},
  journal = {Water},
  year = {2025},
  doi = {10.3390/w17213056},
  url = {https://doi.org/10.3390/w17213056}
}