‍RI-CHAMP incorporates numerical modeling to predict the physical hazards associated with coastal storms.  Storm surge modeling is performed using the ADvanced CIRCulation (ADCIRC) hydrodynamic model, which can be used alone or coupled with a wave model for enhance predictions in wave-dominated regions.  Two advanced wave models, Simulating WAves in the Nearshore (SWAN) and WAVEWATCH III (WW3) are utilized to predict wave characteristics such as wave growth, propagation, and dissipation for different frequencies of waves.  As a result, these models provide prediction of wave characteristics such as significant wave height, mean wave direction, and wave period, which vary in space and time.  

Figure 1: ADCIRC predicted maximum water level (left) and corresponding inundation depths over land (right) from a simulation of Hurricane Sandy.

‍ADCIRC is a hydrodynamic model employed to study the complex behaviors of water during coastal storms. It utilizes the finite element numerical method to solve equations related to the continuity of water flow and the conservation of momentum in shallow waters. By doing so, the model can simulate the changes in water levels and currents at specific points called mesh nodes. Additionally, the model includes various physical parameters that are characterized based on results of in-depth studies, such as the work done by URI on parameterizing the air-sea interface during extreme storms and sea states.

One of the key outputs provided by the ADCIRC model is the areas affected by flooding. These results are produced at varying spatial resolutions, ranging from meters to tens of meters, especially in the nearshore regions with complex coastlines. To further enhance the accuracy of the model's predictions, a high-resolution digital elevation model (DEM), which provides information about the terrain's elevation with a resolution of approximately 1 meter, is utilized to produce refined flood maps.

Figure 2: ADCIRC predicted maximum water level (left) and corresponding inundation depths over land (right) from a simulation of Hurricane Sandy.

The model mesh consists of a series of locations (nodes) where various model quantities (e.g., water level)are calculated within the study domain. The mesh must adequately represent both the offshore bathymetry and coastal topography. To account for varying elevations over short distances in nearshore regions, the mesh resolution is adjusted, being coarser offshore and finer in these areas. Moreover, the mesh plays a crucial role in defining the placement of  structures, such as weirs or barriers, like the Fox Point Hurricane Barrier and its associated pumps in Providence.

Model forcing incorporates several parameters that simulate real-life conditions.  These include:

• Tidal and river forcing is included to accurately simulate the spatial and time-varying water levels and currents from tidal forces and river flows.     
• Meteorological forcing (wind and atmospheric pressure fields) is required to drive the storm surge and wave models.
       -  Tropical storm meteorological forcing for hurricanes is derived from a vortex model
           that uses key storm parameters, including maximum wind speed, radius of maximum
          winds, and the storm central pressure. These parameters are based on the forecasted
           storm track and intensity data provided by the National Hurricane Center (NHC).  
        -  Extra-Tropical storms meteorological forcing for extra-tropical storms (nor’easters)
          come from the output of a global meteorological model such as the Global Forecast
           System (GFS).
       -  Wind speed refinement: RI-CHAMP is designed to provide high-resolution wind data
          during storms that cover the entire state of Rhode Island.

Model validation is an important step to determine the accuracy of the model predictions.  It involves comparing the model forcing and model output to observations of winds, water levels, waves, and flooding. This ensures the reliability of the predictions and helps identify areas where improvements may be needed to enhance the model's performance.

Figure 3: (Top) Validation of inundation over land at a location along Poppasquash Road in Bristol, RI during the December 2022 Nor’easter.  The map shows the maximum inundation layer from the RICHAMP dashboard based on the results from the ADCIRC simulation and high-resolution underlying topography; inundation is only plotted at locations 1 foot above mean higher high water. The image presents a photo of flooding taken during the storm at the location indicated by the red arrow. (Bottom) Validation of storm surge along the coast at a location along the sea wall in Narragansett, RI. The image shows a photo of waves overtopping the sea wall at the location circled in white. Photo source: mycoast.org/RI
‍

Figure 3: (Left) Validation of inundation over land at a location along Poppasquash Road in Bristol, RI during the December 2022 Nor’easter.  The map shows the maximum inundation layer from the RICHAMP dashboard based on the results from the ADCIRC simulation and high-resolution underlying topography; inundation is only plotted at locations 1 foot above mean higher high water. The image presents a photo of flooding taken during the storm at the location indicated by the red arrow (Photo source: mycoast.org/RI). (Right) Validation of storm surge along the coast at a location along the sea wall in Narragansett, RI. The image shows a photo of waves overtopping the sea wall at the location circled in white (Photo source: https://turnto10.com/news/local/street-flooding-storm-rain-wind-high-tide-narragansett-westerly-rhode-island-hurricane-barrier-providence-new-bedford-december-23-2022).
‍

Hindcasts
These are simulations of storms that have occurred in the past.  These are often modeled to evaluate how well the model setup works at recreating the observed conditions during the storm.  They can also be useful for exercises to see how different decision making would have unfolded or could be changed to optimize resource allocation or protect assets.

Forecasts
Real-time forecasting is conducted to predict the winds, water levels and waves that may occur as a storm approaches.  These simulations are subsequently utilized to assess the potential risks and impacts on critical assets, enabling emergency managers to make informed decisions regarding resource deployment and/or take actions to mitigate potential damages before the storm makes landfall.  The forecasts are generated using the ADCIRC Surge Guidance System (ASGS), an automated system that runs consecutive forecasts as new storm track predictions or meteorological forcing becomes available.

Hypothetical Storm Events
These are storm events or conditions that did not actually occur but are physically plausible. Hypothetical storm events, such as Hurricanes Ram and Rhody, can be used for various purposes: 1) to gain a better understanding of the range of risks at a specific site, 2) to conduct comparative simulations and assess the effectiveness of different mitigation options, 3) to practice decision making for planning purposes.  

Examples of hypothetical scenarios include:

• Hypothetical hurricane tracks that produceworst-case meteorological conditions and impacts.
• Hypothetical storms with timing of landfall coinciding with the highest tides to simulate more extreme coastal flooding.
• Hypothetical mean sea levels to reflect the projected sea level rise due to climate change.
• Hypothetical shoreline and beach conditions thatanticipate expected future changes.
• Hypothetical mitigation options, such as simulating the closure of a hurricane barrier.