Project Title

Transportation Infrastructure Flooding: Sensing Water Levels and Clearing and Rerouting Traffic out of Danger

Collaborating Universities

University of Virginia
351 McCormick Dr.
P.O. Box 400742
Charlottesville, VA 22904-4742

Virginia Tech
1424 S Main St.
Blacksburg, VA 24061

Principal Investigator(s)

Pamela Murray-Tuite (VT)

Kevin Heaslip (VT)

Venkataramana Sridhar (VT)

Jonathan Goodall (UVA)

Funding Source(s) and Amounts Provided (by each agency or organization)

USDOT: $199,106 (Federal)
Virginia Tech: $159,179 (Match)
Virginia Beach: $40,000 (Match)

Total Project Costs

$199,106 (Federal) / $199,179 (Match)

Start Date


Completion Date



Flooding in urban areas, driven by both precipitation and high tide events, can have a devastating effect on a region’s transportation system and economic viability. In the City of Virginia Beach, the problem is acute as nuisance flooding in heavily populated areas impacts both communities and transportation infrastructure. The critical needs to identify the magnitude of floods are to measure and model precipitation intensity with a short lead time and relate to high tide events to plan proper protective measures for and diversion from problem areas. This study adopts a multi-disciplinary approach (hydrology, regional climate and precipitation forecasting, and transportation engineering) to predict roadway flooding and mitigate travelers’ danger from the flood and delays. We will study two flood-prone locations in Virginia Beach.

From the hydrology/precipitation perspective, the research addresses flooding due to a complex relationship between tide levels and rainfall events. We hypothesize that a data-driven approach whereby patterns of tidal levels and rainfall intensities and durations that cause flooding can be identified. Then forecasted rainfalls and tide levels can be used to forecast periods when roadways may be flooded.

From the transportation perspective, we are concerned about two types of drivers: those who are on the road as the flood occurs and those who have not yet entered that particular road and must be re-routed. For the first group, warning and road closures must be provided in time to remove these drivers from the impact area. The amount of time required to clear the link depends on network traffic conditions and potentially other flooded areas. The second group must be re-routed so as not to enter the affected link(s) and place the drivers in danger from flooding.

The research project consists of 7 main tasks. In Task 1, rainfall and tidal gauge data will be obtained from the City of Virginia Beach and other organizations and then analyzed using standard data mining approaches to identify relationships and patterns. In Task 2, this data and identified influential factors will be used in conjunction with weather forecasts in independent simulation using the Weather Research and Forecasting (WRF) Model to develop rainfall hyetograph forecasts. This will then be used within the models developed through Task 1 to project if the roadway will flood for the forecasted rainfall event. Task 3 takes the outputs from Tasks 1 and 2 and provides a protocol for communicating predicted flooding events and a decision support tool for the traffic management center (TMC) to put out advisories through variable message signs (VMS) and 511 systems for immediate deployment. In future deployment, when connected vehicles are more common, these advisories will also be sent to travelers, as in Task 6. In Task 4, we develop a small to medium sized network around the two study locations that is used in Tasks 5 and 6. In Task 5, we conduct microscopic traffic simulations under a variety of scenarios based on the conditions and timing related to key factors identified in Tasks 1 and 2, weather conditions, seasons (including tourism and tidal effects), times of day, and other incidents that would involve a Fire/EMS response. These simulations will provide the clearance time(s) of the soon-to-be-flooded link(s); a distribution of the clearance times will be developed for comparison with the flood warnings from Task 2. Task 6 involves developing routing recommendations for drivers who are en-route but have not yet entered the flooded or soon-to-be-flooded links through a hyperpath generating algorithm while considering load balancing. Finally, in Task 7, we address uncertainty concerns by evaluating the trade-offs between providing a warning and road closure unnecessarily and failing to issue a warning/road closure when one is needed. Costs associated with this task include property damage and rescue, among others, for failing to issue a warning when it is needed and delay costs (Tasks 5 and 6) when issuing a warning that is not needed.


The research team has been in contact Gregory Johnson (stormwater) and Steve McLaughlin (transportation), both with the City of Virginia Beach, who recommended the selected study areas. This research is in-line with the City of Virginia Beach’s 15-20 year plan, which is intended to have ways to automatically close roads due to flooding.

The City of Virginia Beach has approximately 14 rain gauges deployed and is in the process of deploying a network of approximately 20 tidal gauges. If necessary, the team will deploy its own rain gauges that can be turned over to the City. The roadway flooding prediction will be directly useful to the City of Virginia Beach and the methodology will be useful to other vulnerable coastal areas with similar concerns. The project involves developing a communication protocol for use in the local traffic management center so that advisories can be provided to the public through variable message signs and 511 systems in the present day, even without communication through connected vehicle technologies, which are expected in the near future.

The research team also anticipates holding or participating in a meeting of stormwater management professionals with concerns about water issues in the Virginia Beach and Hampton Roads area (e.g., from the Hampton Roads Planning District Commission, Hampton Roads Transportation Planning Organization, and Virginia Department of Transportation). The attendees will be informed about the study, its methodology, and its findings. Presentations will be made to these stakeholder groups at the study’s conclusion.

The material may also be presented through a short webinar series, where each webinar focuses on a different aspect of the study. Linkages across the aspects will be highlighted. Such a series will be interdisciplinary. Those desiring a bigger picture can attend the entire series while those interested in a specific aspect can attend the associated webinar. The interdisciplinary nature is especially important for students who are increasingly encouraged to pursue interdisciplinary studies and to increase their perspectives. This will help them address real world problems in the future.


Virginia Beach’s 15-20 year plan is to have the ability to automatically close roads due to flooding. Nuisance flooding is already an issue now (Baltic and 21st flooded four times this past summer), and it is expected that nuisance flooding will become a larger issue in the future. While the first approach for addressing this issue would be to prevent flooding, this may not be practical because such solutions can be expensive and may not be practical for all roadways that experience periods of flooding in the future. Our research is timely because it can help the city toward reaching their longer-term goals.

Flooding of roadways not only impacts drivers, but also emergency services such as fire, police, and ambulances. The Shore Drive location that will be used as a case study has a fire department that is sometimes trapped due to flooding. Having predictive capability for this location could allow firemen at this station to move to a different location if flooding is projected to occur due to a forecasted rainfall event.

Since the city cannot afford to rebuild all vulnerable infrastructure, this project will help mitigate the impact of flooding from a traveler perspective as well as help mitigate the danger it imposes to travelers and delays. The predictive capability will allow better allocation of limited resources during critical periods.

Web Links to Reports and to the Project website

Final Report (Oct. 2017): Transportation Infrastructure Flooding: Sensing Water Levels and Clearing and Rerouting Traffic out of Danger