The interplay between coastal flooding, shoreline infrastructure, transportation networks, and governance institutions creates a complex environment. To provide insight on how urbanized coastal communities can be better prepared to make effective decisions about shoreline infrastructure planning and traffic operations, it is necessary to understand how traffic flow patterns react to inundation scenarios, depending on the extent of sea level rise and the configuration of shoreline infrastructure. Using the results of hydrodynamic simulations, we can predict inundated regions as a function of shoreline protection infrastructure and sea level rise. This allows us to simulate the resulting disruptions to transportation networks, such as closure of certain links or inaccessible nodes. These disruptions result in changes in travel routes and demands, i.e. they produce new traffic flow patterns, which can be predicted using various traffic assignment models.
Fig 1. Effect of sea level rise and shoreline protection on traffic conditions in the San Francisco Bay Area during the morning commute. The panel on the left shows congestion with no shoreline protection, while the panel on the right shows congestion after protecting the entire shoreline from flooding.
We have also developed a methodological framework to understand how decision makers can find the best shoreline protection configuration to minimize traffic disruptions due to sea level rise. We have formulated and solved a wide range of game theoretical problems in terms of the number of decision makers involved in the games. First, a centralized decision maker’s problem is formulated and solved with the objective to minimize the additional vehicle-hours-traveled due to coastal flooding, where the objective is to determine the optimal levee configuration along the entire shoreline of the San Francisco Bay under a given budget.
Fig 2. Pareto frontier showing optimized combinations of shoreline protection at the county level.
Fig 3. Comparisons of vehicle hours traveled for competitive and cooperative games. Results demonstrate that all counties benefit from cooperation.
Second, we have formulated and solved multiple decision maker problems, for both competitive versus collaborative cases and static versus hierarchical scenarios. In the San Francisco Bay Area, decision makers can be counties, sub-counties, or cities. In this framework, hydrologic and traffic dynamics must be modeled accurately because protection of a certain shoreline can result in more severe flooding in a nearby unprotected area, and inundated highway links in a county can cause significant congestion in other counties.