Sea-Level Rise Hazards and Decision Support

Science Center Objects

The Sea-Level Rise Hazards and Decision-Support project assesses present and future coastal vulnerability to provide actionable information for management of our Nation’s coasts.  Through multidisciplinary research and collaborative partnerships with decision-makers, physical, biological, and social factors that describe landscape and habitat changes are incorporated in a probabilistic modeling framework to explore the future likelihood of a variety of impacts and outcomes.  Scenario-based products and tools can be applied to inform adaptation strategies, evaluate tradeoffs, and examine mitigation options.


Although the general nature of the changes that can occur on ocean coasts in response to sea-level rise (SLR) is widely recognized, it is difficult to predict exactly what changes may occur, or when they may occur. The ability to predict the extent of these changes is limited by uncertainties in both currently available data that describe the coastal environment, as well as gaps in understanding of some of the driving processes that contribute to coastal change (e.g., rate and magnitude of sea level rise, changes in storminess).  Additionally, the cumulative impacts of physical and biological change on the quantity and quality of coastal habitats are not well understood, and potential societal responses to SLR are uncertain. Nonetheless, coastal managers need actionable information to make decisions that account for future hazards, including SLR.  

This project brings together scientists from the disciplines of geology, hydrology, geography, biology, and ecology to synthesize information on coastal environments to address the effects of SLR on our Nation’s coasts. The approach uses a probabilistic framework, which allows researchers to incorporate observations and account for uncertainties, to evaluate the likelihood of a variety of SLR impacts, including:

Decision makers depend on the future coastal environment having certain characteristics. For example, homeowners desire a home that is at low risk of loss due to coastal erosion. Local planners and managers also need to be able to identify infrastructure that could be at risk to make effective long-term adaptation or mitigation decisions.  Land managers may target parcels for acquisition that provide critical habitat for threatened and endangered species. Flora and fauna require specific habitat attributes to survive and flourish.  To proactively plan for an uncertain future, decision makers need the ability to consider alternative response measures and assess the benefits and costs of options.  Consequently, there is a need to develop decision frameworks that combine detailed and sometimes complicated scientific information in a way that improves the ability to translate it into decision making scenarios.

Conceptual diagram demonstrating how Bayesian networks used in this project incorporate data and knowledge

Conceptual diagram demonstrating how Bayesian networks used in this project incorporate data and knowledge to provide predictions with decision-support applications.  Learn more

(Credit: Erika Lentz, Woods Hole Coastal and Marine Science Center. Public domain.)

Probabilistic Framing

The Bayesian statistical framework is ideal for using data sets derived from historical or modern observations such as long-term shoreline change or wetland accretion/elevation trends. This information can be combined with model simulations and used to define the relationships between key variables in coastal environments. A Bayesian network provides a means of integrating these data to evaluate competing hypotheses regarding the relationships between forcing factors (e.g., rate of SLR, suspended sediment concentration, elevation change) and responses (e.g., shoreline change, wetland vertical accretion, water table change). This framework allows scientists to make probabilistic predictions of the future state of coastal environments for outcomes such as shoreline change, wetland survival, and changes in the depth to groundwater. The predictions also have estimates of outcome uncertainty that can be expressed as both numbers (e.g., 90%) and words (e.g., very likely). The ability to communicate SLR impacts in terms of a probabilistic prediction can improve scientists’ ability to support decision making and evaluate specific management questions about alternatives for addressing SLR.