Event is held virtually on Zoom.
https://umn.zoom.us/j/98889082768
Password: CSFall21
Advanced Remote Sensing Methods for Water Quality Monitoring and Spatial/Temporal Trend Analysis in 10,000+ Minnesota Lakes
Using satellite imagery, we have been assessing lake water quality in Minnesota, USA for over 20 years. These assessments at around five year intervals were used for spatial and temporal trends and causative factors. Recent advances in satellite technology (improved spectral, spatial, radiometric and temporal resolution) and atmospheric correction, along with cloud and supercomputing capabilities have enabled the use of satellite data for automated regional scale measurements of water resource characteristics. These new capabilities provide opportunities to improve lake and fisheries management by measuring more variables (chlorophyll, colored dissolved organic matter (CDOM) and total suspended matter, the main determinants of water clarity) more often.
To utilize these capabilities we have developed field-validated methods and implemented them in an automated water quality monitoring system on University supercomputers. This system enables near real-time monitoring of water quality variables at regional scales, which will enhance our understanding of spatial and temporal variability and responses of surface waters to environmental change.
Combining these new capabilities with earlier assessments, we created a 35-year (1985-2020) satellite-derived late summer water clarity database for long-term trends. To explore seasonal patterns we used the system to create monthly water quality (clarity, chlorophyll, CDOM) data using all available May-Oct Landsat 8 and Sentinel 2 imagery from 2017 to 2020 for 10,000+ lakes. These maps and auxiliary data were used for spatial/temporal analysis to explain regional differences in water quality. Areas dominated by forest/wetland had higher water clarity than agricultural/developed areas. Changes in water clarity throughout the state were attributed to land use intensification, best management practices and changes in precipitation patterns and increasing temperatures. Differences in CDOM were related to rainfall amounts and predominant land cover/use, with wetland/forested area associated with higher CDOM than agricultural areas.
Leif Olmanson is a Researcher at the University of Minnesota with over 20 years experience developing remote sensing applications to create temporally and spatially rigorous datasets of water and land resources for large area ecosystem characterization. He is particularly interested in developing field validated image processing methods implemented in automated geospatial analysis systems such as Google’s Earth Engine and Minnesota Supercomputing Institutes supercomputers to gain a better understanding of the natural environment. He currently leads a team of researchers and computer scientists to build a system capable of near real-time water quality monitoring for Minnesota’s >10,000 lakes using satellite imagery to providing critical water quality information for lake and fisheries management.