Landsat Reveals Reservoir Changes and Bathymetry
Communities around the world rely on reservoirs for drinking water, hydroelectric power, irrigation and more. These critical freshwater resources are affected by seasonal and long-term changes; Water levels in reservoirs may drop during hot summer months or due to prolonged drought, or may flood after a particularly severe storm. Despite their importance, there are significant gaps in our knowledge of reservoir structure and dynamics. Two recent papers use Landsat data to fill these gaps.
Researchers at the University of Southampton used Landsat data to identify where water advanced or retreated between 1984 and 2022, creating the first global dataset identifying the exact year of permanent changes in surface water, such as when a reservoir formed or a stream dried up. The study can track changes in streams as narrow as 30 m and lakes as small as 900 m.2. In a separate study, researchers at Texas A&M University used Landsat data to create a global bathymetric dataset called “3D-LAKES” that allows water managers to estimate the storage capacity of reservoirs.
The animation above shows Amistad Reservoir on the Texas-Mexico border. It uses a 1985 natural color Landsat image overlaid with a Copernicus digital elevation model (DEM) and bathymetric data from the 3D-LAKES dataset. Vertical relief is exaggerated by a factor of four to emphasize topographic features and landforms. The reservoir is jointly managed by the United States and Mexico through the International Boundary and Water Commission (IBWC) for flood control, recreation, and hydroelectric power. Despite its importance to both countries, the reservoir is slowly shrinking. The Surface Water Transitions dataset shows water levels receding over recent decades, with significant recessions between 2012 and 2016. The 3D-LAKES dataset reveals the underwater shape of the reservoir. Together, these datasets complement in situ data on water levels and conditions collected throughout the year.
Human communities shape and are shaped by water. We divert rivers, build reservoirs and build artificial islands, while natural forces – storms, winding rivers and rising seas – reshape our waterways and coastlines. With satellite data being an important tool for studying ecosystem dynamics, researchers have begun to gain a more comprehensive global understanding of where water is and how it changes over time. In their study of aquatic transitions, the University of Southampton team focused specifically on permanent changes in lakes, rivers, coasts and other bodies of water around the world.
Looking at long-term changes in surface waters can help scientists understand the drivers of change, said Gustavo Willy Nagel, the study’s principal investigator. Knowing when a lake began to recede helps water managers determine whether drought, irrigation or other forces caused the decline.
Scientists, policymakers, and water managers can explore the interactive dataset that Nagel and his team created to visualize changes close to home as well as serious global impacts such as the drying of the Aral Sea, lakes created by melting glaciers in Tibet, and the construction of the Palm Islands in Dubai.
Assessing long-term changes in surface waters is a major challenge because surface waters are extremely dynamic. Seasonal fluctuations and climatic forces mean that rivers, lakes and coastlines are constantly changing. To identify permanent water changes while excluding seasonal fluctuations, the researchers ran two algorithms. The first detected whether the water body was advancing or retreating during the study period using the modified normalized difference water index (mNDWI), which uses the shortwave infrared (SWIR) band instead of the near-infrared (NIR) band. The second algorithm used the Green_Red Normalized Difference Water Index (grNDWI) – an index proposed by the research team – to identify the precise year of transition of the water body. A change was considered “permanent” if it did not return to its previous state during the study period from 1984 to 2022.
“The dataset shows, for every location on the planet, the areas where water has advanced or retreated and the year of this change,” Nagel said.
Landsat can help us monitor surface waters. But what about below the surface area?
In a study published in Scientific data in October 2025, researchers at Texas A&M University merged Landsat and ICESat-2 data to create bathymetric maps of half a million lakes and reservoirs around the world. The research team, led by Huilin Gao, used Landsat imagery to calculate the surface area of water bodies, delineate where water meets land, and track changes in water extent over time. Next, they combined laser altimetry from the ICESat-2 satellite to infer the underwater bathymetry of the water bodies. Using these measurements, scientists refined surface-elevation relationships, a key metric for understanding how water storage changes with water level.
The resulting dataset, dubbed 3D-LAKES, is static because the bathymetry does not tend to change significantly from year to year. “This dataset can support many applications, from monitoring water storage to refining hydrological models,” said Chi-Hsiang Huang, the lead author of the study.
3D-LAKES can be used in combination with Landsat-based maps, such as Surface Transition Research or the popular Global Surface Water dataset, to help water resource managers assess the volume of water contained in a reservoir or lake. This allows them to assess flood risks, map habitat or calculate the amount of water available during a particularly dry season. Researchers can also track changes in water volume over time, helping to understand long-term trends in water storage.
Measuring underwater topography has historically been expensive and impractical on a global scale. The 3D-LAKES dataset now provides researchers and managers with crucial bathymetric data on lakes and reservoirs around the world. “With this new dataset, we can achieve a more comprehensive understanding of the impacts of lakes and reservoirs on regional climatology, water security and ecosystem services,” Gao said. Both studies provide water and land managers with unprecedented tools for resource management and planning, from the Amistad Reservoir to the Australian Outback to the Brazilian Amazon.
