About the Presentation:
The distribution and fluctuation of open surface water in lentic (standing) water bodies (i.e., lakes, reservoirs, ponds, dugouts, and wetlands) are shifting rapidly in response to climate change and human alterations to catchments. Monitoring lentic open surface water dynamics is both important and urgent—not just for long-term trends, but also to capture short-term fluctuations. The HydroPatterns pilot project, conducted in two pilot areas (one in the boreal region and one in the prairies), leveraged open-access Sentinel optical and radar satellite imagery, machine learning, and cloud-based technologies to map open surface water on Alberta’s landscape at a monthly scale and 10-meter resolution, achieving over 90% overall accuracy. Compared to existing surface water dynamics products, the fully automated HydroPatterns workflow offers improved edge delineation, detects a greater numbers of small water features (<1 ha), and identifies more non-permanent water features. By providing both intra- and inter-annual inundation frequency metrics, along with month-to-month variations in water extent, the HydroPatterns can be scaled up province-wide to support data-driven decision-making for water and wetland management and conservation efforts.