Poster Presentations:
Graduate Student Researcher
Stevens Institute of Technology
Simulating Climate Change Impacts on Urban Compound Coastal and Pluvial Flooding Using a Street-Scale Hydrodynamic Model
Co-Authors: Erfan Amini, Presenter (Stevens Institute of Technology), Shima Kasaei (Stevens Institute of Technology), Reza Marsooli (Stevens Institute of Technology), Philip Orton (Stevens Institute of Technology), and Muhammad Hajj (Stevens Institute of Technology)
Abstract: Coastal urban areas face increasing compound flood risks from high sea level and intense rainfall events, exacerbated by climate change. This has resulted in an increasing need for compound flood hazard assessment and climate adaptation strategies in the face of intensifying storms and rising sea levels. This research explores compound flood hazards in a changing climate by employing high-resolution computational flood modeling for a densely populated coastal area. We developed a street-scale hydrodynamic model for simulating compound flooding due to storm tide and heavy rainfall in the Hoboken-Jersey City urban area, New Jersey. We utilize the Super-Fast INundation of CoastS (SFINCS) model with a spatial grid resolution of 1.5 meters to accurately simulate compound flooding due to storm tide (tide and storm surge) and intense rainfall. The model is validated against post-tropical cyclone Ida using rainfall data, tide gauge measurements, and crowdsourced flood information and images. Numerical experiments are performed to assess the efficacy of current urban flood adaptation strategies under various climate change scenarios. By simulating multiple sea level rise and intensified rainfall event scenarios under a warmer future climate, the study evaluates the evolving hazards in the future. These simulations reveal potential impacts on flood extent, depth, and duration. The results aim to provide insights for coastal communities and decision makers, offering a quantitative basis for refining flood management strategies using Regional Resilience Determination (R2D) Tool.
Undergraduate Student Researcher
University of Rhode Island
Data-Enabled Validation of Regional Hurricane Damage Models with Variance-Based Sensitivity Analysis: A Case Study of Hurricane Ian
Abstract: This research aims to critically analyze the impact of several factors influencing hurricane damage and accurately replicate historical damage observations using advanced prediction models established in the literature. To this aim, data-driven validation and variance-based sensitivity analysis workflows are proposed and performed using the historical data from Hurricane Ian in Fort Myers, Florida. This paper examines the wind, flood, and wave-induced damage models in literature and software tools, i.e., Regional Resilience Determination Tool (R2D), IN-CORE, and HAZUS-MH. The project highlighted the importance of understanding the regional characteristics of the target region and specific damage mechanisms. This is achieved through the parametric study of the post-hazard reconnaissance data. The important variables identified through sensitivity analysis were first-floor elevation (FFE) and wave height. Such findings can eventually lead to more precise models and simulations, enabling engineers and emergency managers to have a better estimate of damage from future hurricanes.
Graduate Student Researcher
Oklahoma State University
Spatiotemporal Analysis of Rainfall Erosivity in Oklahoma
Co-Authors: Deb Mishra (Oklahoma State University) and Jaime Schussler (Oklahoma State University)
Abstract: The Universal Soil Loss Equation and its based models have long been used to predict soil loss and guide erosion mitigation, relying on precipitation data predating 1957 to develop the Rainfall Erosivity (R-factor). Despite climatic changes and improved precipitation data, the outdated isoerodent map from AH703 remains widely used to estimate soil loss. This study addresses the need for updated estimations by conducting a spatiotemporal analysis in GIS, utilizing high-resolution, 5-minute interval rainfall data from 111 Oklahoma Mesonet sites with an average of 28 years of historical data recorded between 1994 and 2024. New R-factor values were calculated monthly and annually for two geographical data sets: NOAA-defined state climate divisions and EPA-defined Level III ecoregions. Results showed an increase in rainfall erosivity from northwest to southeast Oklahoma, with April to October contributing 86% of the annual R-factor. The study also produced an updated isoerodent map for Oklahoma, revealing significant differences from the original AH703 map, with R-factor changes ranging from -20% to 112%. Most of Oklahoma experienced a 20-60% increase in R-factor. These findings suggest that modern precipitation data should be used to update the R-factor, enhancing soil loss estimates and conservation efforts. Future R-factors were projected under various greenhouse gas emission scenarios using the Coupled Model Intercomparison Project (CMIP) Phase 6 General Circulation Models (GCMs). The historical R-factor data from 1994 to 2024 served as a benchmark for assessing the impacts of climate change on R-factor estimates, providing insights for soil conservation and guiding future mitigation strategies.
Postdoctoral Fellow
University of Memphis
Towards a generalized, scalable framework for effective green stormwater infrastructure implementation based on hydrologic connectivity
Co-Authors: Claudio I. Meier (University of Memphis)
Abstract: In the context of climate change and urbanization growth, nature-based flood mitigation approaches (collectively referred to as “green infrastructure” - GI), are gaining increasing popularity as more effective and sustainable stormwater management strategies, compared to traditional infrastructure. Their successful implementation presents several challenges, including the adaptation to the local hydrologic and land-development conditions, and their effective spatial placement. Hydrologic modeling is key for tackling these issues, by simulating flood dynamics under a range of precipitation, land-use, and GI scenarios. However, correctly representing these distributed strategies into models is not straightforward, since GI solutions alter the hydrology of a basin through localized modifications in infiltrability and interception, and by introducing sinks. Furthermore, many models do not explicitly account for the presence of GI, necessitating localized re-calibrations of their parameters to correctly capture the induced changes in hydrologic processes, increasing the risks for biases and overfitting calibration observations. One way to address these limitations is to validate hydrologic simulations using more than one model, to test alternative modeling assumptions. We propose that a methodology using connectivity metrics represents an alternative, parsimonious framework for evaluating the hydrologic disconnection effected by alternative GI strategies, hence providing a benchmark for traditional hydrologic simulations. Since it uses geospatial raster data available worldwide, the proposed connectivity-based approach can be easily implemented for any geographic location, with only minor limitations concerning the size of the case-study region, allowing for evaluating GI scenarios also at locations with limited calibration data, where it is difficult to perform traditional hydrologic simulations.
Graduate Student Researcher
University of Memphis
Comparing methodologies to detect trends in extreme rainfall statistics in the Southeastern US
Co-Authors: Claudio I. Meier (University of Memphis)
Abstract: Climate change is driving an increase in the frequency and intensity of extreme rainfall worldwide, resulting in significant damage to infrastructure due to urban pluvial flooding. In the Southeastern U.S., short-duration convective storms are a main cause of urban flooding, making it essential to examine trends in short-duration extreme rainfall to update urban drainage system design criteria.
However, there are few studies on trends in short-duration extreme rainfall in the US, especially for sub-hourly durations. Most existing methodologies focus on trends in frequency rather than magnitude and often use statistics like daily or monthly rainfall or the number of events exceeding certain thresholds, which are less relevant for urban flooding and civil engineering. In contrast, only a few studies have explored trends in depth-duration-frequency (DDF) values for shorter durations, with variable results that depend on the chosen statistics, methods, and assumptions.
In this study, we analyze rainfall data from 473 stations in the Southeastern US, with temporal resolutions of either 15 minutes (326 stations) or 1 minute (147 gauges). We conduct various trend tests on various rainfall statistics including rainfall maxima for short durations, ranging from 5 minutes to 6 hours, to assess the robustness of different methodologies in detecting trends. We also fit the maxima series to extreme value models for non-stationary sequences and perform spatial trend tests to identify any patterns. These analyses will provide valuable insights into rainfall trends, highlighting the advantages and disadvantages of different methods.
Adjunct Professor
University of Alberta
Resilience Metric for Post-Flood Road Network
Co-Author: Vibhu Vemana (University of Calgary)
Abstract: Recovery of functionality of road network is one of the critical elements of urban resilience in a post-flood event. Multiple factors affect the recovery such as intensity of the flood event, vulnerability of the critical infrastructure components, social-economic factors affecting the urban center, and other systemic administrative and political issues. To quantify the effect of these factors in the resilience, a metric of road network resilience is the first requirement. If the road network connectivity prior to the flood event along with the connectivity status at different time intervals after the event can be characterized then a resilience metric for connectivity can be developed. A challenging aspect of such quantification metric is the extraction of the road network connectivity in a post-flood scenario.
In recent years there are multiple deep learning tools to extract the roads from the high-resolution satellite imagery. Furthermore, given the recent advances in the digital twinning of the roadways, it is possible to construct pre-flood event connectivity status based on a combination of maps, and aerial imagery. In the proposed work, BRAILS capability to extract road network elements is utilized to identify the connectivity status at one- and five-year interval after a flood event for fifty urban centers. This dataset would enable development of a resilience metric for post-flood road network connectivity.
Graduate Student Researcher
Texas A&M University
Assessing Transportation System Resilience for Coastal Testbeds under Hurricane-Induced Storm Surge
Co-Authors: Maria Koliou (Texas A&M University)
Abstract: The resilience of transportation networks is critically shaped by the performance of bridges during extreme weather events such as hurricanes. This research establishes a comprehensive framework for resilience estimation, specifically for hurricane scenarios. Two areas along the southern coast of Texas, near Bay City and Victoria, were chosen as testbeds for this analysis. Using hazard data from 446 synthetic storms, the distribution of storm intensity metrics was calculated for each bridge within the transportation network. Performance-based assessments were conducted to estimate the probability of bridge damage states through Monte Carlo simulations and fragility analysis.
Following this, a transportation traffic network simulation model was developed using the SUMO software to evaluate the impact of bridge failures on overall network performance. The traffic flow for the two testbeds was estimated using AADT data provided by TxDOT and analyzed through the simulation model. Key performance indicators—including travel time, economic losses, and loss of accessibility—were employed to construct a resilience index. Unlike previous studies, this research introduces a multi-dimensional resilience index encompassing robustness, rapidity, redundancy, and resourcefulness, each functioning as an individual sub-index within the broader resilience framework.
Mitigation strategies were optimized and analyzed for the two testbeds, and the resulting resilience indices were compared. The findings demonstrate the efficiency, accuracy, and comprehensiveness of the proposed assessment framework. This approach offers decision-makers and engineers a valuable tool for evaluating transportation network vulnerabilities and developing more effective mitigation strategies to reduce the risks posed by hurricanes induced- storm surges.