Can satellite imagery provide insights into water quality?

Sushant Mehan, assistant professor and South Dakota State University Extension water resource specialist, is conducting a research project that will utilize satellite imagery to assess water quality in eastern South Dakota. 

Lake Mitchell is a man-made reservoir that once served as Mitchell's primary source of drinking water. Today, Lake Mitchell is often polluted with harmful algae blooms, and local officials are determining the best way to clean up the lake.

The lake has also been the site of many studies, and with a large swathe of data collected, it is an ideal spot for a South Dakota State University water research project. Sushant Mehan, assistant professor and SDSU Extension water resource specialist, has received a grant from the South Dakota Water Resources Institute to assess surface water quality in eastern South Dakota through integrated remote sensing. 

The Water Resources Institute at SDSU is one of the U.S.'s 54 water resources research institutes. Authorized by Congress, the institute connects research expertise at South Dakota's universities to water-related problems at the local level. It is aimed at conducting research to aid in the resolution of state and regional water problems, promotes technology transfer and the dissemination and application of age-related information, and provides competitive grants for students and researchers.

The focus of Mehan's work will utilize integrated remote sensing — satellite imagery — to assess surface water quality. He will use data acquired by Landsat satellites from the U.S. Geological Survey's Earth Resources Observation and Science Center near Sioux Falls. Mehan will validate the findings from the satellites with ground-level measurements at Lake Mitchell. 

"The goal is to achieve a comprehensive understanding of both surface and subsurface water trends by reconstructing historical water quality data for a water body," Mehan said.

First, Mehan and a doctoral student will use remote sensing data to develop algorithms and models for analyzing water quality parameters. Common water quality measurements include temperature, dissolved oxygen, pH, conductivity and turbidity. Then, the research team will use machine learning techniques — artificial intelligence tools — to build vertical profiles of the water bodies. This helps the team assess water quality beyond surface-level concentrations.

"This novel approach has the potential to revolutionize water quality assessment and management, providing more accurate and timely data for decision-making," Mehan said.

After the model is validated from measurements taken at Lake Mitchell, Mehan will be able to explore the relationship between water quality parameters and environmental factors. This analysis will be crucial in better understanding the underlying reason for Lake Mitchell's challenges.

"The proposed model will incorporate spectral indices from remote sensing alongside climate variables to predict continuous water quality data," Mehan explained.

While the project will be focused on the validation efforts at Lake Mitchell, Mehan believes if the project is successful, the model can be applied to other lakes in eastern South Dakota and beyond.

"This research aims to leverage remote sensing and statistical analysis to develop a comprehensive, integrated model for estimating and understanding the trends of water quality and changes in eastern South Dakota," Mehan added.

Overall, the project will help enhance the water quality evaluations conducted by the South Dakota Department of Agriculture and Natural Resources and will help also mitigate water quality degradation.