Probabilistic SLR flood modeling using ROMS reanalysis

Abstract:

Hydrostatic modeling, also called “bathtub modeling”, is the most commonly used method for depicting the impacts of sea-level rise (SLR). It is less common, though still widely appreciated by the scientific community, that the digital elevation models (DEMs) used for hydrostatic modeling carry a vertical accuracy that is related to the original source data as well as choices in data processing methodology. Examples of where vertical accuracy is applied in modeling are found in Cooper et al. (2013) and Gesch et al. (2020), as well as the NOAA SLR Viewer. However, these do not take into account the physical variability of the ocean surface on intra- to inter-annual timescales. We analyze data from a 10-year Regional Ocean Modelling System (ROMS) reanalysis for the region surrounding the main Hawaiian Islands. The reanalysis simulates sea-level variability induced by atmospheric and tidal forces. By analyzing this data, we have been able to calculate a mean higher high water (MHHW) value of 0.28 ± 0.02 m above mean sea-level which closely matches that of the Oahu tide gauge (0.33 m) and produce cumulative distribution functions of ocean surface variability for locations surrounding Oahu. We incorporate the probability of flood depths above MHHW in various SLR scenarios. Additionally, we use an urbanized DEM to allow viewing of SLR impacts as a rotating 3D image. Preliminary results illustrate an increase in the reach and the total area flooded by direct marine flow compared to that of pure hydrostatic modeling. We show model results in the form of maps that use color coding to identify levels of likelihood (per IPCC).