Velocity hazard projections at Elim, Alaska

Individual surfbeat and non-hydrostatic 2D XBeach models were stood up and tested against known past storm events (models can be found in this release).

A set of extreme water level and wave conditions was identified at the approximate open boundaries of each 2D XBeach model. These extreme conditions were derived using wind, atmospheric pressure, and sea ice fields from high-resolution global climate models (GCMs) as forcing inputs to wave and hydrodynamic models. The dynamic downscaling process is described in Erikson and others (2022) and the manuscript of Engelstad and others (2024). Extreme value statistics for 1-, 10-, 20-, 50-, and 100-year return period coastal storms were calculated from both historical (1979–2015) and future projection (2020–2050) GCM runs. The percent change between historical and projected extremes was then computed and applied to similar statistics derived from model runs forced by ERA5 reanalysis wind, pressure, and sea ice data (1979-2022). Percent changes were calculated from the ensemble mean of four GCM simulations: CMCC-CM2-VHR4-r1i1p1f1_gn, CNRM-CM6-1-HR-r1i1p1f2, EC-Earth3P-HR- r1i1p2f1_gr, and HadGEM3-GC31-HM_highresSST-future_r1i1p1f1_gn.

2DXBeach production runs: All scenario simulations were done in a two-staged approach whereby XBeach was first run in the surfbeat mode (XB-SB), bed level changes interpolated onto the non-hydrostatic (XB-NH) model grid, and the latter model run. This two-stage approach leverages the strengths of both XBeach modes for a comprehensive simulation of wave dynamics, particularly in the context of swash zone (active beach area) processes and wave runup. The SB mode is computationally efficient and well-suited for capturing long-wave (infragravity) variations and their interactions with the shoreline, including morphodynamic change (erosion and sedimentation). The NH mode provides a detailed, wave-resolved simulation that includes individual wave overtopping. Ran each of the 2DXBeach SB models 30 times for all combinations of 6 SLR scenarios and the 1-, 10-, 20-, 50-, and 100-year return period water level and wave extremes. Interpolated the final bed elevations onto the 2DXBeach NH model grids and ran each of the 2DXBeach NH models 36 times for all combinations of SLR and storms plus the no storm condition.

Post-process and finalize data products: All simulation data were downloaded from the USGS Advanced Research Computing Center’s Hovenweep Supercomputer and post-processed to extract maximum water levels at each grid point over the entire simulation and subsequently interpolated onto the finer resolution (1m x 1m) digital elevation mesh. Interpolations were done using The Mathworks MATLAB linear scatteredinterpolant function which in turn uses a Delaunay triangulation algorithm of the sample points. Significant wave height, erosion and sedimentation, and the velocity-depth product were similarly processed. Maximum water elevation surfaces were depth-differenced to the DEM to separate low-lying vulnerable areas from contiguous coastal flooding. Other output data (water elevation, maximum depth x velocity, sedimentation/erosion, and maximum wave heights) were clipped to the same final flood hazard extent. Additionally, all data below the mean higher high-water line (MHHW) were removed from flood depth projections. Flood potentials (maximum and minimum flood uncertainty) were derived by adding/subtracting estimated model error, DEM uncertainty, 95th percent confidence intervals of the storm event water level, and projected vertical land motion (VLM) for a given SLR to the respective water elevation surface. See O'Neill and others (2018) for a similar approach. Velocity hazard GeoTIFFs were created for all combinations of six SLRs (0, 0.5, 1.0, 1.5, 2.0 and 3.0 m) and 5 storms (1-year, 10-year, 20-year, 50-year, and 100-year return period coastal events), for a total of 30 scenarios. Final GeoTIFFs were separated by community (Velocity_hazard_projections-Elim.zip). Data are organized by storm scenario ('RP') and SLR amount.