Delft3D FM Alaska Tide and Storm Surge Model Output - CMIP6 climate downscaling (historical 1979-2014 and future RCP 8.5 2020-2050)

README – Delft3D FM Alaska model output – CMIP6-forced historical and future runsDate: 2026-08-15Prepared by: Kees Nederhoff ([e-mail]), Deltares USA Description This archive contains raw Delft3D Flexible Mesh (DFM) tide and storm surge output for the Alaska coastline, forced directly by atmospheric fields from five high-resolution CMIP6 global climate models. Each model is provided for a historical period (1979–2014) and a future period under SSP5-8.5 forcing (2020–2050). The model, domain, grid and output locations are identical to the ERA5-forced hindcast release (doi.org/10.5281/zenodo.15807568). The only difference is the meteorological forcing: where the hindcast is driven by ERA5 reanalysis, these runs are driven by GCM wind and sea level pressure fields without bias correction. Model setup and forcing Wind and sea level pressure from five global climate models participating in the High Resolution Model Intercomparison Project (HighResMIP; Haarsma et al., 2016) within CMIP6 were used to force the hydrodynamic model directly. No bias correction or delta-change transformation was applied to the forcing. The models are: CMCC-CM2-VHR4 – CMCC, Italy (Cherchi et al., 2019) CNRM-CM6-1-HR – CNRM-CERFACS, France (Voldoire et al., 2019) EC-Earth3P-HR – EC-Earth Consortium (Haarsma et al., 2020) GFDL-CM4C192 – NOAA-GFDL, USA (Held et al., 2019) HadGEM3-GC31-HM – Met Office Hadley Centre, UK; prescribed-SST (highresSST) configuration (Roberts et al., 2019) Each model was run over both periods with identical model settings, so that the difference between a model's future and historical run isolates the climate change signal for that model, free of any difference in model configuration. SSP5-8.5 (Kriegler et al., 2014) is used as an upper bound on plausible forcing. By recent accounts it is less plausible in the long term (van Vuuren et al., 2026), but that matters little here: the emissions trajectories of SSP5-8.5, SSP3-7.0 and SSP2-4.5 stay relatively similar through mid-century and diverge mainly after 2070, and these simulations stop at 2050. What this means when using the data Read these four points before analysing the files. Absolute water levels carry GCM bias. The forcing is raw GCM output. Each model has its own biases in storm track, storm frequency and wind speed, and those propagate directly into the simulated water levels. Absolute magnitudes and return levels will differ systematically from observations and from the ERA5 hindcast, and should not be used as design values without bias correction. Use intra-model differences, not absolute levels. The intended quantity is the difference between a model's future run and its own historical run. Comparing one GCM against another, or against a tide gauge record event by event, mixes model bias into the signal. Use the ERA5 hindcast as the observed-climate reference. The companion ERA5-forced release uses the identical grid, settings and output locations, so it is the appropriate baseline for assessing each model's historical bias before interpreting its change signal. Extreme value results above roughly the 30-year recurrence level assume stationarity. Each window is about 31–36 years, so return levels beyond about that horizon are extrapolations of a stationary fit rather than a projected trend. A small number of output points are not hydraulically connected to the open coast and carry no meaningful signal (a permanently dry cell, or a water level pinned to the bed level). In the ERA5 hindcast this affected 82 of 941 coastal points, whose standard deviation is below 2 cm. The same points are affected here. Check the variability of a point before using it. Contents The release consists of ten ZIP archives, one per model and period: CMIP6_CMCC_historical_1979-2014.zip and CMIP6_CMCC_projection_2020-2050.zip CMIP6_CNRM_historical_1979-2014.zip and CMIP6_CNRM_projection_2020-2050.zip CMIP6_ECEARTH_historical_1979-2014.zip and CMIP6_ECEARTH_projection_2020-2050.zip CMIP6_GFDL_historical_1979-2014.zip and CMIP6_GFDL_projection_2020-2050.zip CMIP6_HadGEM3_SST_historical_1979-2014.zip and CMIP6_HadGEM3_SST_projection_2020-2050.zip Each archive contains: _overview.kml – station/output locations for Google Earth _overview.xls – station names, coordinates and metadata his*.nc – NetCDF time series, one file per output location Data format All model output is provided in NetCDF format and includes time series of water levels together with metadata describing variable units and attributes. The files are compatible with common analysis software such as Python (xarray, netCDF4), MATLAB and R. References Cherchi, A. et al. (2019). Global mean climate and main patterns of variability in the CMCC-CM2 coupled model. J. Adv. Model. Earth Syst., 11, 185–209. doi.org/10.1029/2018MS001369 Deltares (2023). Delft3D Flexible Mesh Suite. Delft, The Netherlands. deltares.nl Haarsma, R. J. et al. (2016). High Resolution Model Intercomparison Project (HighResMIP v1.0) for CMIP6. Geosci. Model Dev., 9, 4185–4208. doi.org/10.5194/gmd-9-4185-2016 Haarsma, R. et al. (2020). HighResMIP versions of EC-Earth: EC-Earth3P and EC-Earth3P-HR – description, model computational performance and basic validation. Geosci. Model Dev., 13, 3507–3527. doi.org/10.5194/gmd-13-3507-2020 Held, I. M. et al. (2019). Structure and performance of GFDL's CM4.0 climate model. J. Adv. Model. Earth Syst., 11, 3691–3727. doi.org/10.1029/2019MS001829 Hersbach, H. et al. (2020). The ERA5 global reanalysis. Q. J. R. Meteorol. Soc., 146(730), 1999–2049. doi.org/10.1002/qj.3803 Kriegler, E. et al. (2014). A new scenario framework for climate change research: the concept of shared climate policy assumptions. Clim. Change, 122, 401–414. doi.org/10.1007/s10584-013-0971-5 Nederhoff, K., Laan, S., Erikson, L., Jagers, B., & Maio, C. (in preparation). Ice-modified wind drag and its contribution to extreme water levels along the Alaskan coast. Roberts, M. J. et al. (2019). Description of the resolution hierarchy of the global coupled HadGEM3-GC3.1 model as used in CMIP6 HighResMIP experiments. Geosci. Model Dev., 12, 4999–5028. doi.org/10.5194/gmd-12-4999-2019 van Vuuren, D. P. et al. (2026). The Scenario Model Intercomparison Project for CMIP7 (ScenarioMIP-CMIP7). Geosci. Model Dev., 19, 2627–2656. doi.org/10.5194/gmd-19-2627-2026 Voldoire, A. et al. (2019). Evaluation of CMIP6 DECK experiments with CNRM-CM6-1. J. Adv. Model. Earth Syst., 11, 2177–2213. doi.org/10.1029/2019MS001683