Back to top

SFS Annual Meeting

Poster Details

<< Back to Posters

REMOTE SENSING AS A TOOL FOR IDENTIFYING NON-POINT SOURCES OF NUTRIENTS IN THE LAKE SIMCOE BASIN, ONTARIO, CANADA

Non-point sources of nutrients pose one of the greatest threats to water quality in agricultural basins. By definition, non-point sources are difficult to identify and manage. Through a pilot study conducted in the Lake Simcoe basin, Ontario, Canada we looked to identify whether multispectral satellite imagery from LandSat 8 and Sentinel 2A, could be used to map nutrient hotspots at a watershed scale. By utilizing terrestrial plant growth (chlorophyll abundance) as a proxy for nutrient availability, random forest modelling was used to infer phosphorus deposits using multispectral analysis of chlorophyll across the landscape. Significant correlation was observed between inferred soil phosphorus and soil samples collected throughout the watershed. Once a map of inferred phosphorus has been established, we used hydrological flow mapping to assess the relationship between soil phosphorus and water total phosphorus samples collected throughout the watershed. Being able to identify non-point sources of phosphorus at a landscape scale, provides a powerful tool for targeting resource limited nutrient reduction efforts and best management practices to areas where they will be most beneficial in protecting water quality.

MacKenzie Waller (Co-Presenter/Co-Author), River Institute, mwaller@riverinstitute.ca;


Mark MacDougall (Primary Presenter/Author), River Institute, mmacdougall@riverinstitute.ca;