On May 1, the LOEWE Center emergenCITY welcomed two researchers as research associates. While Mahshid Khazaeiathar was previously employed at emergenCITY as a research assistant in the Cyber-Physical Systems program area, Mohsen Dehghani Darmian is a new addition to the team.

Mahshid Khazaeiathar has been a doctoral student at TU Darmstadt in the Department of Engineering Hydrology and Water Resources Management (ihwb) and a research assistant in the emergenCITY team since 2022. Her research focuses on machine learning, data analysis, statistical modeling, time series prediction and hybrid modeling approaches in hydrology. As part of emergenCITY, she is involved in the “Smart Flood and Low Flow Warning System” project, which aims to develop an advanced environmental monitoring and warning system. This system uses real-time data and artificial intelligence to provide early warnings of flooding.

“I hope to improve resilience strategies and increase decision-making capacity for disaster preparedness,” explains Mashid Khazaeiathar.

To this end, she is working on developing data-driven solutions to improve early warning systems for extreme hydrological events, particularly flooding.

Ultimately, the integration of real-time data and hybrid modeling techniques should enable better forecasting and management of flood and low water events.

How much water runs through the river during floods?

Mohsen Dehghani Darmian, the Humboldt Foundation’s first postdoctoral climate protection fellow at TU Darmstadt, worked on the effects of climate change on the water quality and quantity of reservoirs as part of his sponsorship. His areas of expertise are water resource management, hydrodynamic and water quality modeling as well as applications of machine learning in hydrology.

“My aim is to improve the prediction of river discharge, especially for flood events under changing climate conditions,” explains Mohsen Dehghani Darmian.

To this end, he wants to develop advanced AI-based tools within the LOEWE center emergenCITY, in particular using genetic programming, which is a machine learning method inspired by principles of evolution. It aims to find the best program with regard to a predefined function.

His work is thus located at the interface between hydrology, climate adaptation and AI modeling, enabling a more accurate prediction of extreme weather events.