Surface water temperature observations and ice phenology estimations for 1.4 million lakes globally

Picture
Figure 6 from Korver et al., 2024. The yearly mean LSWT (a), the yearly range in LSWT (b), and the yearly duration of ice cover (c) for the center points of ∼1.4 million lakes, based on Julian day averages over 2013–2021 (left) and averaged by one-degree latitude bins (right).
Background

​Lakes play a critical role in climate regulation, freshwater availability, and ecosystem functioning, yet global information on lake temperature and ice cover remains incomplete. Most existing observations rely on in-situ monitoring networks that are geographically sparse and biased toward large, well-studied lakes. Small and medium-sized lakes, which make up the majority of lakes worldwide, are particularly underrepresented, despite their importance for carbon cycling, biodiversity, and regional climate feedbacks. This lack of consistent global data has limited our ability to assess long-term trends in lake warming and ice phenology at broad spatial scales.

Approach

We assisted in the development of LakeTEMP, a global dataset of lake surface water temperature and ice phenology derived from Landsat 8 thermal infrared observations collected between 2013 and 2021. The analysis targeted more than 1.4 million lakes larger than 0.1 km² identified in the HydroLAKES database. Temperature observations were extracted from lake center points to minimize land contamination and processed using a globally applicable split-window algorithm. To support broad-scale analysis, individual observations were combined into monthly and annual summary products using seasonal trend modeling that accounts for data gaps caused by cloud cover and seasonal ice.

Key Findings
  • Global coverage and accuracy: LakeTEMP provides quality-controlled temperature observations for approximately 1.4 million lakes, validated against in-situ measurements with strong agreement across continents.
  • New insights into ice phenology : Analysis shows that more than 86% of lakes freeze annually, with an average ice duration exceeding half the year for much of the global lake surface area.
  • Thermal extremes identified : The dataset reveals global patterns in lake warming, including extremely warm lakes primarily associated with industrial and artificial water bodies.

Impact
​
LakeTEMP fills a major gap in global environmental monitoring by providing analysis-ready temperature and ice data for nearly every mapped lake on Earth. By enabling consistent comparisons across regions and lake sizes, the dataset supports climate research, large-scale limnology, ecosystem assessment, and integration with other global hydro-environmental databases.

Resources

Published Paper : Korver, M. C., Lehner, B., Cardille, J.A., & Carrea, L. (2024). Surface water temperature observations and ice phenology estimations for 1.4 million lakes globally. Remote Sensing of Environment, 308, 114164. DOI : https://doi.org/10.1016/j.rse.2024.114164.

Data Product Repository :
The LakeTEMP dataset is publicly available under a CC-BY-4.0 License at the following URL: https://doi.org/10.6084/m9.figshare.23844660