BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
METHOD:PUBLISH
PRODID://CWSP//521419
BEGIN:VEVENT
DTSTAMP:20260528T133910
VTIMEZONE:America/New_York
DTSTART:20260617T170000Z
DTEND:20260617T183000Z
UID:521419
SUMMARY:Webcast 5. Machine Learning in Watershed Modeling for Stormwater Planning and Flood Forecasting
LOCATION:Online via Zoom
DESCRIPTION:Webcast 5. Machine Learning in Watershed Modeling for Stormwater Planning and Flood Forecasting\n\n06/17/26 01:00 PM EST\n - 06/17/26 02:30 PM EST\Description:\nSpeakers:\n\nWalter McDonald, Marquette University\nUsing machine learning and remote sensing for urban stormwater BMP inspection and maintenance\nDescription: Urban stormwater best management practices (BMPs)—including both green and gray systems such as bioretention basins, detention ponds, permeable pavements, and swales—are critical for mitigating flooding, improving water quality, and restoring more natural hydrologic function. However, maintaining these systems over time remains a persistent challenge, as traditional field-based inspections are labor-intensive, costly, and difficult to scale across large urban areas. This study investigates how emerging remote sensing technologies and machine learning algorithms can enhance the efficiency of BMP inspection and maintenance. By integrating drone and high-resolution satellite imagery with machine learning classification models, we developed an approach to detect key land cover indicators of maintenance needs, including vegetation stress, sediment buildup, and surface deterioration. Applying this framework to dozens of BMP sites in Milwaukee, Wisconsin, we demonstrate that remote sensing and data-driven methods can serve as a scalable and cost-effective screening tool for stormwater program managers, helping to prioritize maintenance and optimize inspection resources.\n\nCatherine Riihimaki,  2NDNATURE Software Inc\nThe Evolution of Stormwater Modeling: From Hydraulic Models to AI-Enhanced Management Tools\nIn this talk, 2NDNATURE will take us on a journey through the evolution of stormwater modeling and discover how modern tools are making sophisticated analysis accessible to program managers at scale.\n\nStormwater management has long relied on complex custom hydraulic models that are powerful but difficult to implement across entire watersheds or municipalities. This webinar traces the evolution from those traditional models to the Tool to Estimate Load Reductions (TELR)—a fit-for-purpose solution designed specifically for stormwater program managers. TELR provides practical guidance on where runoff requires mitigation and quantifies how management efforts reduce volumes and pollutant loads, without the complexity that has historically limited widespread application. Now, AI is extending TELR's capabilities further by automatically determining critical inputs like structural BMP locations and street sweeper curb access, while improving code flexibility for long-term sustainability.\n \n\Location:\nOnline via Zoom\n\n,
X-ALT-DESC;FMTTYPE=text/html:Webcast 5. Machine Learning in Watershed Modeling for Stormwater Planning and Flood Forecasting<br /><br />06/17/26 01:00 PM EST - 06/17/26 02:30 PM EST<br />Description:<br /><p>Speakers:</p>

<p class="elementtoproof" style="background:white">Walter McDonald, Marquette University<br />
<strong><em>Using machine learning and remote sensing for urban stormwater BMP inspection and maintenance</em></strong><br />
<strong>Description</strong>: Urban stormwater best management practices (BMPs)&mdash;including both green and gray systems such as bioretention basins, detention ponds, permeable pavements, and swales&mdash;are critical for mitigating flooding, improving water quality, and restoring more natural hydrologic function. However, maintaining these systems over time remains a persistent challenge, as traditional field-based inspections are labor-intensive, costly, and difficult to scale across large urban areas. This study investigates how emerging remote sensing technologies and machine learning algorithms can enhance the efficiency of BMP inspection and maintenance. By integrating drone and high-resolution satellite imagery with machine learning classification models, we developed an approach to detect key land cover indicators of maintenance needs, including vegetation stress, sediment buildup, and surface deterioration. Applying this framework to dozens of BMP sites in Milwaukee, Wisconsin, we demonstrate that remote sensing and data-driven methods can serve as a scalable and cost-effective screening tool for stormwater program managers, helping to prioritize maintenance and optimize inspection resources.<br />
<br />
Catherine Riihimaki,&nbsp; 2NDNATURE Software Inc<br />
<em><strong>The Evolution of Stormwater Modeling: From Hydraulic Models to AI-Enhanced Management Tools</strong></em><br />
In this talk, 2NDNATURE will take us on a journey through the evolution of stormwater modeling and discover how modern tools are making sophisticated analysis accessible to program managers at scale.<br />
<br />
Stormwater management has long relied on complex custom hydraulic models that are powerful but difficult to implement across entire watersheds or municipalities. This webinar traces the evolution from those traditional models to the Tool to Estimate Load Reductions (TELR)&mdash;a fit-for-purpose solution designed specifically for stormwater program managers. TELR provides practical guidance on where runoff requires mitigation and quantifies how management efforts reduce volumes and pollutant loads, without the complexity that has historically limited widespread application. Now, AI is extending TELR's capabilities further by automatically determining critical inputs like structural BMP locations and street sweeper curb access, while improving code flexibility for long-term sustainability.<br />
&nbsp;</p>
<br />Location:<br />Online via Zoom<br /><br />,  
PRIORITY:3
TRANSP:TRANSPARENT
BEGIN:VALARM
TRIGGER:-PT5M
ACTION:DISPLAY
DESCRIPTION:Reminder
END:VALARM
END:VEVENT
END:VCALENDAR