Session Lead: J. Derek Loftis (Virginia Institute of Marine Science)

Co-Lead(s): Navid Tahvildari, Patrick Taylor

Session Format: Oral presentations

Session Description: 

The hallmark of a well-rounded inundation prediction model is the ability to successfully characterize flood depths, extents, and duration before an impending catastrophe. The ability of a predictive model to identify these three facets of flooding in advance can be most challenging during compound flooding scenarios, due to the inherent complexities involved in calculating the combined flood impacts of rainfall, storm surge, and tides. Sensitivity of different meteorological inputs (wind speed and direction, air pressure, precipitation) converging upon varying conditions on the ground (initial soil moisture and saturation, infiltration through ground surfaces of varying permeability, spatially varying vegetation cover, and storm water drainage impoundment), quickly translate into a seemingly insurmountable modeling problem to address. Yet, technological advancements in big data and real-time water monitoring systems on the ground coupled with remote sensing data can now deliver enhanced predictive inputs for hydrodynamic, hydrologic, and machine learning models to provide realistic forecasts for compound flooding applications.

This special session invites authors to present their recent original research advancements in flood modeling and monitoring for compound flooding applications leveraging hydrodynamic, hydrologic, or machine learning models. This session does not require submitted abstracts to directly address 2016 Hurricane Matthew, Chesapeake Bay’s largest compound flooding event in recent history, but it is anticipated that over half of the submitted talks for this special session submitted to the 2026 Chesapeake Community Research Symposium will address it. Presentations addressing anticipated impacts of storms similar to 2016 Hurricane Matthew in the future (in the context of projected sea level rise and more precipitative storms correlated with climate change in the Mid-Atlantic Bight) are also appropriate for this session.

Presentations (Session 16 Abstracts):

  1. Joseph Zhang: A 25-year reanalysis of compound flooding hazard in US east and Gulf coast
  2. HaoCheng Yu, Lars Nerger, Fei Ye, Y. Joseph Zhang, Hyungju Yoo, Saeed Moghimi, Gregory Seroka, Zizang Yang, Edward Myers, Carsten Lemmen, S. Chin: Elevation skill enhancement from an efficient ensemble-based assimilation method in a large application STOFS-3D-Atlantic
  3. Jon Derek Loftis: Hydrodynamic Modeling of Compound Flooding During 2016 Hurricane Matthew:  Then, Now, and Storms Like It In The Future
  4. Zanko Zandsalimi, Mehdi Taghizadeh, Majid Shafiee-Jood, Negin Alemazkoor: Explicit Interdomain Learning of Rainfall–Tide Coupling for Compound Flood Forecasting Using Graph Neural Networks
  5. Hyungju Yoo, Y. Joseph Zhang, Zhengui Wang, Fei Ye, Haocheng Yu: Enhancing Thermal Process Representation in Intertidal Areas through Soil-Air-Water Heat Exchange: A Case Study of Charleston Harbor
  6. Jon Derek Loftis, Yash Kishor Sanap, Sridhar Katragadda, Russ Lotspeich: High-Precision River Stage Estimation via Passive Video Imagery Using Deep Learning and Image Segmentation
  7. Jon Derek Loftis: Spatial Evaluation of Flood Resilience Solutions Combining Real Time Water Level Sensors, Hydrodynamic Modeling, and High-Resolution Aerial Inundation Observations