LakeTEMP: Global Lake Surface Water Temperature and Ice Phenology
Figure 1. from Korver et al., 2024. Overview of the LakeTEMP dataset and its two subsets.
|
Surface water temperature observations and ice phenology estimations for 1.4 million lakes globally
Global monitoring of lake temperature and ice cover has been limited by sparse in-situ observations, especially for small and remote lakes that dominate global lake numbers. As presented in this paper, we helped develop LakeTEMP, the first quality-controlled global dataset of lake surface water temperature and ice phenology for approximately 1.4 million lakes using Landsat 8 thermal imagery. This dataset enables consistent, planet-scale assessment of lake thermal dynamics and ice cover, providing a critical foundation for climate, hydrology, and ecosystem research.
|
Canadian Lake Health Assessment Using Satellite Remote Sensing
Graphical abstract from Deutsch et al., 2022.
|
Assessing the current water clarity status of ~ 100 000 lakes across southern Canada: A remote sensing approach
Monitoring lake health across Canada is constrained by the sheer number of lakes and by traditional methods that require field measurements to coincide exactly with satellite overpasses. In this paper, we showed that aggregating Landsat 8 imagery using a robust median-based approach improves satellite estimates of lake water clarity and dissolved organic matter, even without strict timing alignment. This advance enabled national-scale assessment of water quality for ~100,000 lakes across southern Canada, directly supporting freshwater management and policy. |
Conservation Prioritization in African Lakes
Figure 3 from Danaher C, 2022. Percentage of species that are classified as threatened (vulnerable, endangered, and critically endangered obligate freshwater species and freshwater‐dependent vertebrate species) per watershed.
|
Prioritizing conservation in sub-Saharan African lakes based on freshwater biodiversity and algal bloom metrics
Freshwater biodiversity in sub-Saharan Africa is increasingly threatened by eutrophication and harmful algal blooms, yet many lakes remain poorly monitored and underrepresented in conservation planning. In this paper, we combined high-resolution satellite monitoring with biodiversity data to identify lakes where elevated algal bloom risk overlaps with high freshwater species richness, identifying where ecological value and environmental stress intersect across Ghana, Ethiopia, and Zambia. We identified 169 priority conservation areas, providing a scalable, data-driven framework to guide freshwater protection and conservation planning in remote and data-scarce regions.
|
Improving Lake CDOM Algorithms
Figure 1. Locations of calibration (grey), validation (blue), and the additional 128 lakes without clear imagery taken within 30 days of sampling (purple) plotted on a map of Canada.
|
Multiple Images Improve Lake CDOM Estimation: Building Better Landsat 8 Empirical Algorithms across Southern Canada
Monitoring Coloured Dissolved Organic Matter (CDOM) is essential for managing lake health, yet traditional satellite methods are limited by the rare chance of a cloud-free satellite pass perfectly coinciding with field sampling. We recognized that this "temporal matching" requirement restricts our ability to monitor remote or rarely visited water bodies. In this paper, we prove that using the median value of multiple Landsat 8 images over several summers creates significantly more accurate models than relying on a single, perfectly timed image. This methodology allows us to monitor nearly 40% more lakes across Canada by removing the restrictive requirement for contemporaneous sampling, providing a vital tool for national-scale environmental assessments.
|
Large-scale Calibration and Validation of Lake Water Clarity Algorithms
Graphical abstract
|
Landsat 8 Lake Water Clarity Empirical Algorithms: Large-Scale Calibration and Validation Using Government and Citizen Science Data from across Canada
Monitoring water clarity across Canada’s nearly 900,000 lakes is difficult because traditional satellite methods rely on locally calibrated models and tightly matched field, satellite timing, limiting national-scale assessments. In this paper, we show that a single, universal algorithm can estimate Secchi disk depth (standard field measure of water clarity based on the depth at which a black-and-white disk remains visible), across diverse Canadian lake types. To do so, we combined Landsat imagery with long-term median filtering to reduce atmospheric noise, even when observations are months apart. Regression residuals were used as a spatial diagnostic to evaluate whether our national lake-clarity model behaves differently in different ecozones. Model performance was independently validated using high-quality field data from the NSERC Canadian Lake Pulse Network. This work enables flexible, reliable national-scale monitoring of lake water clarity, strengthening long-term freshwater management and environmental assessment.
|
