Data scientists are using new techniques to identify lakes and reservoirs around the world

A team of data scientists led by the University of Minnesota’s Twin Cities has published the first comprehensive global dataset of lakes and reservoirs on Earth showing how they have changed over the past 30 years.

The data will provide environmental researchers with new information about land and freshwater use and how lakes and reservoirs are affected by humans and climate change. This research is also a major advance in machine learning techniques.

A paper highlighting the Reservoir and Lake Surface Area Timeseries (ReaLSAT) data set was recently published in Scientific Data, a peer-reviewed open access journal published by Nature.

Study highlights include:

• The RealLSAT dataset contains location and surface area variations of 681,137 lakes and reservoirs greater than 0.1 square kilometers south of 50 degrees north latitude. The previous most comprehensive database, called HydroLAKES, has identified only 245,420 lakes and reservoirs for this part of the world and the minimum size considered in this study.
• RealLSAT provides data on the surface area of ​​each body of water for each month from 1984 to 2015. This makes it possible to measure changes in the area of ​​lakes and reservoirs over time, which is key to understanding how climate change and land use change water bodies. freshwater. The HydroLAKES data contains only static shapes for each body of water.
• The RealLSAT dataset is the culmination of eight years of research. This is a major milestone in the adoption of new knowledge-guided machine learning for use in environmental science. Unlike other existing efforts, this data set can now be extended almost automatically through machine learning and can be rapidly replicated for a wide variety of available Earth observation data with increasingly better resolution.

“Around the world, we see lakes and reservoirs changing rapidly with seasonal rainfall patterns, climate change and human management decisions over the long term,” said Vipin Kumar, senior author of the study and Regents Professor and William Norris Chair in the U of M Department of Science. and Computer Engineering. “This new dataset greatly enhances scientists’ ability to understand the impact of climate change and human actions on our clean water around the world.”

Building a global data set of lakes and reservoirs and how they are changing requires a new kind of machine learning algorithm that combines knowledge of the physical dynamics of water bodies with satellite imagery.

“ReaLSAT is a shining example where environmental challenges are motivating a new class of knowledge-guided machine learning algorithms that are now being used in a variety of scientific applications,” said Kumar.

Scientists who study the environment agree that ReaLSAT will improve their work.

“The availability and quality of surface freshwater is critical to the sustainable use of our planet,” said Paul C. Hanson, Distinguished Research Professor at the University of Wisconsin-Madison Limnology Center and co-author of the study. “Because RealLSAT shows changes in lakes and their boundaries, not just water pixels across the landscape, we can now link ecosystem processes about water quality with hundreds of thousands of lakes around the world.”

This research was supported by the US National Science Foundation and NASA. Access to computing facilities is provided by the University of Minnesota Supercomputing Institute.

/ Public Release. Material from this original organization/author may be timely, edited for clarity, style and length. The views and opinions expressed are those of the author. See more here.

#Data #scientists #techniques #identify #lakes #reservoirs #world