Near Real-Time Monitoring of Flood Inundation in Africa using Remote Sensing and Machine Learning

Near Real-Time Monitoring of Flood Inundation in Africa using Remote Sensing and Machine Learning

Webinar Speaker: Dr. Subit Chakrabarti, Director of Technology at Cloud to Street, USA
Dr. Subit Chakrabarti

About the Webinar

In recent years, there has been an uptick in the frequency of climate disasters across the globe. African countries are generally poorly equipped to recover and adapt to climate-related shocks, most notably, disastrous flooding. According to Africa’s Pulse publication (Zeufack et al., 2021), within the last decade, storms have quadrupled in frequency while floods have also increased tenfold when compared to the 1970s. Since 2015, Cloud to Street has supported humanitarian organizations and governments of 28 countries across Africa with a near real-time multi-sensor flood monitoring system. In this talk, this system, utilizing Landsat and Sentinel constellations coupled with our deep learning based CNN flood detection algorithm, will be demonstrated. We will also show how this flood monitoring system provides the necessary historical data to assess and mitigate risk, improves situational awareness, and informs disaster response and recovery with medium to high resolution satellite observations.

About the Speaker

Subit is Director of Technology at Cloud to Street where he manages a team of scientists and engineers with the goal of producing high-quality maps of peak flood extent relevant to the needs of disaster responders, flood managers, and insurers. His technical expertise and interest is in developing novel spatio-temporal machine learning methods applicable for large-scale earth imagery. He has a PhD in Electrical Engineering from the University of Florida where his thesis focused on machine learning-based superresolution of microwave imagery for land surface and biophysical models. Prior to Cloud to Street, Subit worked as a data scientist at Indigo Agriculture and Telluslabs on machine learning-based mapping of crop types, regenerative farming practices and crop yields using satellite imagery.