Wave Projections for United States Mainland Coasts
Wave Projections for United States Mainland Coasts
By Li H. Erikson, Christie E. Hegermiller, Patrick L. Barnard, and Curt D. Storlazzi
Executive Summary
Spatially- and temporally-varying waves dominate coastal morphology and ecosystem structure, impact coastal infrastructure, natural and cultural resources, and coastal-related economic activities along much of the United States (U.S.) coastline. Yet the influence of climate change on future wave conditions along the mainland U.S. shores are not well understood. Wave heights, periods, and directions were forecast through the year 2100 using wind outputs from four separate atmosphere-ocean coupled global climate models of the Coupled Model Inter-Comparison Project, Phase 5 (BCC-CSM1.1, INM-CM4, MIROC5,and GFDL-ESM2M). Two climate scenarios wree modeled, the Representative Concentration Pathway (RCP) 4.5 and RCP 8.5, each of which corresponds to moderately mititigated and unmitigated greenhouse gas emissions, respectively. Wind fields from the global climate models were used to to drive the WAVEWATCH III numerical wave model on a near global grid (1.00º x 1.25º resolution) and on a one-way nested grid (0.25º x 0.25º resolution) encompassing the Eastern North Pacific that included the southern Alaska and U.S. West Coasts. Hourly time-series wave data were saved at 59 locations surrounding the continental United States (15 sites along the West and East, and 2 along the Southern coasts) and Alaskan (25 sites) coasts for a historical (1976-2005) time-period and two future time-periods, the mid-century (2026-2045) and end-of-century (2081-2100).
All time series data are available for download at http://dx.doi.org/10.5066/F72B8W3T. The data are provided as-is for comparison of historical simulations to past observations and projected future wave conditions. It is anticipated that the data may be used in numerous research and application studies such as detailed evaluations of underlying mechanisms that might impart changes in future wave conditions and as oundary conditions for coastal impact studies. Several such studies have been conducted or are on-going (for example, Storlazzi and others 2015; Erikson and others 2015a,b). This methods summary presents the modeling approach and provides an overview of projected changes in the wave climate for the mid- and end-of-century time-periods compared to the historical time-period.
Overall, model results indicate that significant wave heights, Hs, and peak wave periods, Tp, will decrease along much of the mainland U.S. West, East and Gulf Coasts and increase along the southeast, west, and Arctic Alaskan coasts. Projected changes under the influence of RCP 8.5 climate scenario are spatially similar to those of RCP 4.5, but larger in magnitude. The greatest projected changes in Hs are along the Arctic Alaskan coast under RCP 8.5 during the end-of-century, for which extreme Hs are projected to increase by more than 60 cm during the open water season (typically May/July through October/November). Projected changes in Hs are even greater (greater than 75 cm) during December-February, which suggests changes in storm intensity, rather than changes in sea ice extents, since the presence of sea ice was not included in the wave simulations.
Mean Hs along the Alaskan Arctic coast are projected to increase by more than 10-30 cm under both RCP 4.5 and RCP 8.5 during the end-of-century, with the latter scenario consistently exhibiting greater increases. Large increases in Hs variability, σ2Hs, are projected for the central Arctic Alaskan coast and show a spatial northeasterly transgression between the mid- to end-of-century time periods. Increases in Tp are also projected for the Arctic coast, which combined with Hs, have implications for wave steepness and wave runup, factors that are important to sediment transport across the continental shelf, boating safety, inter-tidal habitats, coastal erosion, and flooding.
Extreme conditions, defined here as the mean of the top 5 percent highest waves and the mean values of Tp and most frequent Dp associated with those events, are projected to increase significantly along the entire Arctic Alaskan coast during all seasons and for both RCP 4.5 and RCP 8.5. Consistent with projected changes of mean conditions, projected increases of Hs are greater under RCP 8.5. Spatially, the Chukchi Sea (western Arctic Alaskan coast) is expected to endure greater increases compared to the Beaufort Sea (eastern Arctic Alaskan coast). Projected increases between March and November have potentially negative implications for coastal erosion, as sea ice cover is less during these months, leaving the coast exposed and vulnerable to wave attack.
The western Bering Sea coast of Alaska is projected to experience increases in mean Hs during boreal winter months December through January (DJF) under RCP 4.5 and for all seasons under RCP 8.5 by the end-of-century. Results indicate increasing extreme Hs under RCP 4.5 during DJF mid-century, whereas for RCP 8.5 extreme Hs is shown to increase for all seasons along this stretch of coast. By the end of the century, extreme Hs are projected to increase substantially along the entire west coast of Alaska for both climate scenarios.
The southern coast of Alaska is expected to experience both increasing and decreasing mean H_s through the mid- and end-century; increases are projected for the eastern portion of the southern coast, whereas decreases are projected near the western portion of the Aleutian Islands. A similar spatial trend is projected for extreme Hs, particularly during DJF.
The mainland west coast of the U.S. is projected to experience an overall decrease in extreme Hs and an increase in associated Tp. Little to no changes are projected for mean Hs, but mean Tp are projected to increase by 0.25-0.50 s during the Northern Hemisphere summer and fall seasons along most of the coast. Extreme Hs are projected to decrease off the coast of southern Oregon and southward toward southern California during all seasons under both RCP 4.5 and RCP 8.5 during the mid-century. Decreases of 10-30 cm in Hs are projected for nearly the entire coast during DJF, with the greatest decreases extending from southern Oregon to central California. Extreme Hs and σ2Hs are projected to decrease substantially (by more than 30 cm and 40 cm, respectively) along nearly the entire region by the end-of-century and all seasons, but most notably during DJF.
The mainland U.S. East and Gulf Coasts are projected to experience decreases in mean Hs during the mid-century of more than 20-30 cm during DJF, and to undergo little to no change during the remaining seasons. By the end-of-century, mean Hs are projected to decrease for all seasons under RCP 8.5. Extreme Hs are projected to decrease by mid-century during DJF under both RCP 4.5 and RCP 8.5, but to increase in the vicinity of 40°N for the remainder of the season under RCP 4.5. By the end-of-century, extreme Hs are shown to decrease by more than 30 cm during DJF and by 10-30 cm during the remaining seasons along the entire East and Gulf coasts. Mean Tp associated with extreme Hs are also projected to decrease during DJF and to undergo little to no change during the remaining seasons.
The results of this modeling effort and brief evaluation of changes in the wave climate elucidate the complexity of the link between changing global atmospheric patterns and consequent variations in wave climate. The regional data summaries presented herein exhibit vast differences with respect to spatial patterns and magnitudes; for example, mostly increasing Hs along the Alaskan coast and decreasing conditions at the lower latitudes. Although the patterns of change are complex, there appears to be a consistency in that the patterns of change within each region are largely coherent between the RCP 4.5 and RCP 8.5 scenarios, but with more pronounced changes for the latter scenario. Evaluations provided by these data are indicative of potential future trends in the coastal wave climate, but there is still much uncertainty in projections of how wind patterns will evolve. Additionally, the along-coast wave climate modeled as part of this effort may be vastly different from nearshore wave conditions in shallower water, particularly in areas with complicated geography, bathymetry, or where broad continental shelves allow for refraction and energy dissipation. For these situations, regional downscaling of the models is required.
Chapter 1. Introduction
Coastal managers and ocean engineers rely heavily on projected mean and extreme wave conditions for planning and design purposes, but when working on a local or regional scale, they are often faced with much uncertainty in the projections as changes in the global climate imparts spatially-varying trends in wave and wind conditions . Long-term (greater than 30 years) wave climatologies, in terms of means, extremes and variability, are sparse along the United States’ coasts. Even in regions where climatologies are well defined, wave conditions over the past several decades may not be indicative of the future wave climate since waves are the result of winds which are affected by climate change. Winds are driven by atmospheric variability and the complex interaction between the Earth’s atmosphere and ocean systems, making it difficult to project future wind and consequent wave conditions. Fortunately, the relatively recent availability of large-scale databases from atmosphere-ocean coupled general circulation models (AOGCM) or global climate models (GCM) now allows for modeling of future wave conditions in response to changes in this complex system of atmosphere/ocean exchange.
Two common approaches for estimating waves from GCM-simulated winds use statistical or dynamical downscaling techniques. Statistical downscaling relies on relationships between historical wind and wave conditions in combination with future wind conditions to project future wave conditions. Dynamical downscaling involves the use of numerical models and historical and future winds to compute wave growth and propagation on ocean scales. In this work, dynamical downscaling was used to project 21st-century wave conditions along United States’ mainland coasts and evaluate differences compared to a historical time period. The historical GCM simulations were initiated from arbitrary times of quasi-equilibrium control runs, and thus the timing of past observations are not expected to coincide precisely with results from so-called GCM ‘historical’ model runs (Taylor 2012). It should instead be considered that GCMs aim to represent the overall seasonal and longer term variability as well as mean and extreme conditions.
Near-surface winds simulated by four separate GCMs and two climate-change scenarios (RCP 4.5 and RCP 8.5) were used as forcing to a global numerical wave model. One historical time period (1976-2005) and two future time-periods [mid- (2026-2045) and end-of-the 21st century (2081-2099/2100)] were simulated. Comparisons of significant wave heights (Hs), peak periods (Tp), and peak wave directions (Dp) were made between the historical and future time periods. The results are presented graphically and with cursory descriptions in this methods summary.
The results presented herein are derived from time-series of modeled Hs, Tp, Dp, wind speed (Ua), and wind direction (Uθ) at 59 distinct points along the U.S. mainland coasts that were extracted from the wave and GCM models; the data are available for download at http://dx.doi.org/10.5066/F72B8W3T. Wind data are provided at the native GCM 3-hourly time-resolution whereas wave data are provided at hourly time-steps as computed with the numerical wave model (see Data and Methods section) and linear temporal interpolations of winds. The distinct model output points are spaced approximately 300 km apart in the along-coast direction or co-located with observation buoys and are limited to grid points where water depths were greater than approximately 300 m and distant from the coast. These limitations on selection of model output points were imposed to avoid the influence of shallower water depths and topographic interactions that are not well captured with the relatively coarse wave model resolution (ranging from 0.25° to 1.25°). Furthermore, in areas closer to land, wind fields are affected by finer-scale land-sea interactions, such as orographic and katabatic effects, that are not well represented with the coarse resolution GCM wind fields (ranging from 1.4° to 2.8°).
The time-series data provides information on trends and variability of geophysical variables that are expected to respond to changes in global-scale forcing. It is anticipated that the data will be used for evaluation of trends and variability in offshore conditions and as boundary conditions for regional and local scale coastal hazard models. Because winds and waves are the key processes driving extreme water levels and inundation, the data are expected to be crucial for projecting future transient sea level extremes on coasts and for defining areas that might be vulnerable to changing wind and wave conditions.
Figure 1. Overview of model output locations along U.S. coasts and U.S. and U.S. affliated Pacific Islands. Data summaries of the Alaskan, U.S. West and U.S. East and Gulf Coasts are presented in this methods summary. Assessments of changing wave conditions near U.S. and U.S. Affiliated Pacific Islands are presented in a separate report (Storlazzi and others, 2015). [Larger version]
Chapter 2. Study Area
Figure 2. Regional overview of model output locations. North American West Coast (NAWC) and North American East Coast (NAEC) stations are located approximately 300 km apart and identified with blue text and circles. Filled squares refer to model output locations that are co-located with observation buoys. [Larger version]
Projected changes in deep-water (greater than approximately 300 m water depth) wave conditions along the U.S. mainland coasts are grouped into three regions (Fig. 1): Alaska Coast, West Coast, and East and Gulf Coasts. Assessments of changing wave conditions in the U.S. and U.S. Affiliated Pacific Islands are presented in a separate report (Storlazzi and others, 2015; see highlighted region in Fig. 1).
The Alaskan region includes a total of 25 model output points (Fig. 2A, Table 1). Six output points surround the Arctic coast, eight surround the Aleutian Islands, four are within the shallow region of the Bering Sea, and the remaining seven are within the Gulf of Alaska.
The U.S. West Coast region stretches from the U.S.- Mexico border to the U.S.- Canada border and includes open coast areas of California, Oregon, and Washington. The West Coast region includes fifteen model output points (Fig. 2B, Table 1). Eight model output points are co-located with observation buoys and are identified by National Oceanic and Atmospheric Administration National Data Buoy Center (NDBC, http://www.ndbc.noaa.gov/) station numbers (N46229, N46213, N46214, N46042, N46028, N46069, N46219, N46047).
The U.S. East and Gulf Coasts encompass fifteen coastal states stretching from the Gulf Coast States and Florida in the south to the U.S.-Canada border north of Maine. The region includes seventeen model output points (Fig. 2C, Table 1); seven are co-located with NDBC observation buoys (N44011, N44014, N41001, N41002, N41010, N42001, N42055).
Data summaries for the U.S. East and Gulf Coast regions are provided from the 1.25° x 1.00° global (NWW3) wave model grid (described in Data and Methods section below). Data summaries for the U.S. West Coast region are available from the NWW3 grid and from the finer resolution 0.25° x 0.25° Eastern North Pacific (ENP) grid nested within the NWW3 grid. Data summaries for the southern coast of Alaska are also available from the ENP grid. In cases where model data exist for both the NWW3 and ENP grids (Table 1), both sets of data are available for download (http://dx.doi.org/10.5066/F72B8W3T), but only results from the ENP grid are used in the data summaries presented here. Accuracy of results is generally better with the ENP grid, but for evaluation of patterns and relative change of future versus historical climatologies, the more coarse NWW3 grid is sufficient for many purposes (Erikson and others 2015a).
Table 1. Model output locations along U.S. mainland coasts consisting of hourly wave data.
| West Coast Region | East Coast Region | Alaska Coast Region | ||||||
| Station ID | Latitude (°N) | Longitude (°W) | Station ID | Latitude (°N) | Longitude (°W) | Station ID | Latitude (°N) | Longitude (°W) |
| NAWC28E | 47.86 | -126.9282 | NAEC2 | 43.7931 | -58.0184 | NAWC1 | 70.9931 | -138.2133 |
| NAWC29 E | 45.2319 | -126.0056 | NAEC3 | 42.5699 | -61.441 | NAWC2 | 71.6267 | -146.732 |
| NAWC30 E | 42.2034 | -126.1622 | NAEC4 | 41.2702 | -65.2718 | NAWC3 | 72.2725 | -153.1403 |
| NAWC31 E | 39.2777 | -125.5579 | NAEC5 | 39.8917 | -68.0866 | NAWC4 | 71.9163 | -161.8808 |
| NAWC32 E | 36.6892 | -123.9792 | NAEC6 | 39.0268 | -71.6807 | NAWC5 | 69.9575 | -167.5118 |
| NAWC33 E | 34.0787 | -121.9753 | NAEC7 | 37.0694 | -74.094 | NAWC6 | 68.0386 | -170.4998 |
| NAWC34 E | 32.1948 | -119.7264 | NAEC8 | 34.414 | -75.1802 | NAWC7 | 64.3881 | -169.7887 |
| N46229 E | 43.769 | -124.551 | NAEC9 | 32.7315 | -77.9523 | NAWC8 | 61.5759 | -169.0895 |
| N46213 E | 40.294 | -124.74 | NAEC10 | 30.3311 | -79.6556 | NAWC9 | 59.0748 | -166.2795 |
| N46214 E | 37.945 | -123.47 | NAEC11 | 27.736 | -79.022 | NAWC10 | 57.1052 | -162.6716 |
| N46042 E | 36.789 | -122.404 | N44011 | 41.105 | -66.6 | NAWC11 | 55.3438 | -168.3401 |
| N46028 E | 35.741 | -121.884 | N44014 | 36.611 | -74.842 | NAWC12 | 53.4152 | -175.9537 |
| N46069 E | 33.67 | -120.2 | N41001 | 34.675 | -72.698 | NAWC13 | 49.7663 | -176.7464 |
| N46219 E | 33.221 | -119.882 | N41002 | 32.309 | -75.483 | NAWC14 | 50.9305 | -173.0363 |
| N46047 E | 32.403 | -119.536 | N41010 | 28.906 | -78.471 | NAWC15 E | 51.7254 | -167.7831 |
| N42001 | 25.888 | -89.658 | NAWC16 E | 52.9285 | -163.5232 | |||
| N42055 | 22.203 | -94 | NAWC17 E | 54.0654 | -158.3292 | |||
| NAWC18 E | 55.431 | -154.3737 | ||||||
| NAWC19 E | 56.5148 | -151.0793 | ||||||
| NAWC20 E | 57.5747 | -148.1874 | ||||||
| NAWC21 E | 58.956 | -145.753 | ||||||
| NAWC22 E | 58.5627 | -142.3421 | ||||||
| NAWC23 E | 57.5634 | -139.8463 | ||||||
| NAWC24 E | 55.8667 | -137.5199 | ||||||
| NAWC25 E | 53.7141 | -135.338 | ||||||
|
E Stations denoted with “E” include results from both the near-global (NWW3) and Eastern North Pacific (ENP) grids; only results from the NWW3 grid are available for all other stations. Summary plots presented in this methods summary include ENP grid data where available. Station name convention: NAWC: North American West Coast; NAEC: North American East Coast; N#: NDBC buoy identification number (buoys are operated and maintained by either NDBC or Scripps Institution of Oceanography Coastal Data Information Program, CDIP). |
||||||||
Chapter 3. Data and Methods
Global Climate Model Wind Data
Datasets of near-surface (10-m height) winds simulated by four separate GCMs were used in this study (Table 2):
- Beijing Climate Center, Meteorological Administration, China (BCC-CSM1.1),
- Institute for Numerical Mathematics, Russia (INM-CM4),
- Model for Interdisciplinary Research on Climate, Japan (MIROC5), and
- Geophysical Fluid Dynamics Laboratory, National Oceanic and Atmospheric Administration, U.S.A. (GFDL-ESM2M).
Criteria for selection of the four GCMs was based on availability of projected near-surface winds to the year 2099/2100, output frequency (3-hourly non-averaged synoptic winds), and completed GCM simulations at the onset of this study. All global simulations follow the Fifth Phase of the Coupled Model Intercomparison Project (CMIP5) protocol for long-term simulations (Taylor, 2012). Only outputs from the first ensemble GCM simulations (r1) were used when multiple ensemble (differing initial conditions) runs were available.
The middle (2026-2045) and end (2081-2100) of 21st century time-periods were simulated under two climate scenarios. Representative concentration pathway (RCP) 4.5 and RCP 8.5 represent estimates of the average global radiative forcing by the year 2100 relative to the 1850 pre-industrial period (van Vuuren and others, 2011). RCP 4.5 represents a future with relatively ambitious emissions reductions so that global radiative forcing is stabilized shortly after year 2100 (Thomson and others, 2011). RCP 8.5 represents a future with no policy changes to reduce emissions (Riahi and others, 2011). RCP 4.5 and RCP 8.5 scenarios are roughly equivalent to the B1 and A2 emission scenarios (Moss and others 2010) of the IPCC Special Report on Emission Scenarios (SRES; Meinshausen and others 2011), respectively. Historical (1976-2005) simulations were used to assess model skill (Erikson and others 2015b) and determine magnitude of change in future scenarios.
Table 2. List of global climate models and resolutions of wind fields used for modeling waves.
| Modeling Center | model | model resolution |
|---|---|---|
| Beijing Climate Center, Meteorological Administration, China | BCC-CSM1.1 | 2.8° x 2.8° |
| Institute for Numerical Mathematics, Russia | INM-CM4 | 2.0° x 1.5° |
| Model for Interdisciplinary Research on Climate, Japan | MIROC5 | 1.4° x 1.4° |
| NOAA Geophysical Fluid Dynamics Laboratory, USA | GFDL-ESM2M | 2.5° x 1.5° |
Wave Model
The third-generation, spectral wave model WAVEWATCH III (WW3; version 3.14; Tolman and others, 2002) was forced by historical and projected GCM-simulated winds. The model was applied over a near-global grid (NWW3; latitude 80°S–80°N) with 1.00° × 1.25° spatial resolution, and a nested ENP grid with 0.25° × 0.25° spatial resolution (approximately 27 km at latitude 37°N). Bathymetry and shoreline positions were populated with the 2-min Naval Research Laboratory Digital Bathymetry Data Base (DBDB2) v3.0 and National Geophysical Data Center Global Self-Consistent Hierarchical High-Resolution Shoreline (GSHHS; V1.7). Wave spectra were computed with 15° directional resolution and 25 frequency bands ranging non-linearly from 0.04 to 0.5 Hz. Wind-wave growth and whitecapping were modeled with the Tolman and Chalikov (1996) source term package and nonlinear quadruplet wave interactions were computed with the Hasselmann and others (1985) formulation. Parameterizations of physical processes (source terms) include wave growth and decay due to the actions of wind, nonlinear resonant interactions, dissipation, bottom friction, depth-induced breaking, and scattering due to wave-bottom interactions. Spatial maps of daily average wave heights, periods, and directions and wind speeds and directions were saved on a daily basis; wind and wave parameters at deep-water output locations (Fig. 1, Table 1) were saved at hourly increments.
Sea Ice
Variations in global-scale sea ice coverage were excluded from all wave simulations and thus, historical wave data should be evaluated in consideration of this limitation, and projected wave data can be considered to be from an ocean free of seasonal sea-ice. The absence of sea ice affects the results in that waves are present throughout the year along the Arctic Alaskan coast and within the Bering Sea, for both the future and historical time-periods even though it is a known fact that waves have not been historically present in these regions for the winter months of the simulated historical time-period. Additionally, Hs and Tp are likely slightly over-estimated during the start (May-July, historically) and end (September-November, historically) of the open water season as the lack of sea ice contributed to greater fetch over areas that are typically covered with sea ice. During other times of the year and for the future time-periods, this effect is likely negligible since storm waves appear to be duration-limited rather than fetch-limited (unpublished results). In lower latitudes, model results may also exhibit somewhat higher Hs and Tp because swell wave growth in the Southern Ocean and North Sea was not limited by sea ice.
Model Output Data
Hourly time series of modeled output data include significant wave height (Hs), mean wave period (Tm), mean wave direction (Dm), peak wave period (Tp), peak wave direction (Dp), mean wind speed (Ua) and wind direction (Uθ) (Table 3). These data are available for download in both ASCII and NetCDF formats at http://dx.doi.org/10.5066/F72B8W3T for the sites listed in Table 1.
Table 3. Definitions of model output parameters.
| Parameter | Definition | Units |
| Significant wave height, Hs |
Hs = 4√E, and E = ∫ 2π o∫ ∞ o F ( f , θ ) d f d θ where F ( f , θ ) is the wave spectrum across all frequencies ( f ) and directions ( θ ). | Meters (m) |
| Mean wave period, Tm | Tm = 2 πσ–1, where σ = 2πf. | Seconds (s) |
| Mean wave direction, Dm | Dm = atan(b / a), where a = ∫ 2π o∫ ∞ o cos( θ ) F ( σ , θ ) d σ d θ and b = ∫ 2π o∫ ∞ o sin( θ ) F ( σ , θ ) d σ d θ. |
Direction waves are traveling from in degrees (°) clockwise from North |
| Peak wave period, Tp | Inverse of the frequency associated with the maximum energy of the one-dimensional frequency spectrum using a parabolic fit. | Seconds (s) |
| Peak wave direction, Dp | Defined like Dm, using the frequency bin containing the spectrum that contains the peak frequency only. | Direction waves are traveling from in degrees (°) clockwise from North |
| Mean wind speed, Ua | Average wind speed interpolated from 3-hourly GCM data. | Meters per second (m/s) |
| Wind direction, Uθ | Incident wind direction associated with Ua. | Direction winds blow from in degrees (°) clockwise from North |
|
Units and definition of directions (nautical, coming from) are consistent with formats of downloadable data. |
||
Wave Parameter Summary Statistics
Regional changes in Hs, Tp, and Dp are summarized in this methods summary. Means, extremes, and variances (or standard deviations) of seasonal time series data were calculated for the historical, mid-century, and end-of-century time-periods under RCP 4.5 and RCP 8.5. Seasonal changes in bulk parameters were calculated by subtracting historical values from projected values with seasons defined as December-February (DJF), March-May (MAM), June-August (JJA), and September-November (SON).
Seasonal means (X) and variances (σ2) of Hs and Tp were calculated as,
(1)
(2)
where xi is the parameter value and N is the total number of data points. Summary statistics for Dp were calculated differently because of the circular nature of direction and strongly bimodal conditions, particularly in the Alaska region. As an example, the arithmetic mean (Eq. (1) of northwest and northeast incident waves would indicate northerly incident waves along the Arctic Alaskan coast, where this wave condition rarely occurs. Therefore, Dp summary statistics were calculated as the most commonly occurring incident wave direction for a given season. Variations of Dp were calculated as the circular standard deviation,
θDp = √—2lnR(3)
where R is the vector length of circular data.
Extreme Hs summary statistics were calculated as the mean (Eq. 1) and variance (Eq. 2) of all seasonal data that exceeded the seasonal 95th percentile. Extreme Tp and Dp were calculated as described using data associated with extreme Hs. Means, variances, and standard deviations of all wave data were computed from the combination of the four GCM-driven WW3 model simulations to generate multi-model ensembles, which are expected to better represent predictions than any one model alone (Erikson and others, 2015b; Donat and others, 2010).
Chapter 4. Regional Summaries
Alaska Coast [Figs. 3 through 50]
Mid-Century: 2026-2045
RCP 4.5: Mean
Changes in mean Hs and Tp (denoted as δHs and δTp) are projected to be within ±10 cm and ±0.25 s, respectively for all seasons and around the entire Alaska coast (Figs. 2A-9A). An exception is an increase of 10-20 cm in mean Hs in the southern Chukchi Sea during DJF. Most common incident wave directions are projected to rotate slightly clockwise (CW) along the southern coast of Alaska during DJF, but otherwise are not projected to change.
RCP 4.5: Top 5 Percent
Extreme mean Hs are projected to increase by 10-30 cm along the Arctic Alaskan coast for all seasons; smallest increases are projected to occur during the winter months DJF when sea ice will likely be present protecting the shoreline from direct wave attack (Figs. 2B-5B). The realtively greater increases during the remaining seasons has the potential to increase coastal erosion as the ice pack will be far from the coast during many of these months. Projected increases in DJF extreme Hs extend to nearly the entire western coast and parts of the southern coast. Projected changes in mean Tp associated with extreme Hs are negligible. Most frequent Dp associated with extreme Hs are projected to rotate clockwise (CW) along the western Arctic Alaskan coast and counter-clockwise (CCW) along the eastern Arctic Alaskan coast. Projected CW and CCW rotations along the west and east Arctic Alaskan coasts likely indicate poleward migration of storm tracks. Along the southern coast, Dp are projected to rotate CCW, whereas within the Bering Sea Dp are projected to rotate CW (Figs. 10B-13B). These projections also suggest migrations of storm centers.
RCP 8.5: Mean
Similar to projections under RCP 4.5, results indicate little to no change in magnitude or variability of mean Hs and Tp, with the exception of the southern Chukchi Sea where increases in Hs of 10-20 cm are projected during DJF (Fig. 14A). Most frequent incident Dp are projected to rotate slightly CW along the south Alaskan coast during DJF, but otherwise appear unchanging.
RCP 8.5: Top 5 Percent
Extreme Hs are projected to increase by more than 10-30 cm along the Arctic Alaska coast, with the greatest projected increases during DJF in the southern Chukchi Sea. In contrast to projections under RCP 4.5, which indicate little to no change in Hs along the west coast except during DJF, projections under RCP 8.5 indicate increases during all seasons along this stretch of coast (Figs. 14B-17B). Decreases in Hs of 10-20 cm are projected for areas of the southern exposed coast of the Aleutian Islands during MAM, JJA, and SON (Figs. 15B-17B). Projected changes in the variance of Hs (σ2Hs) are negligible, except along the southern coast of the eastern Aleutian Islands during DJF. Mean Tp associated with extreme Hs are projected to increase by 0.25-0.50 s for all seasons within the Chukchi Sea (Figs. 18B-21B). Since both historical and future simulations were conducted without sea ice, projected increases in mean Tp are not the result of increased fetch, but rather might be an indication of increasing storm intensity. Projected changes in direction are similar to those under RCP 4.5.
End-Century: 2081-2100
RCP 4.5: Mean
Mean Hs are projected to increase along the entire Arctic Alaska coast by at least 10-20 cm during all seasons, and up to 30 cm in the Chukchi Sea during DJF. Hs variances are projected to increase in this region by more than 40 cm. Similar increases in mean Hs are projected along the entire western Alaska coast during DJF (Fig. 32A). Remaining areas of the coast are not projected to experience any appreciable changes in mean Hs during any season. Mean Tp are expected to increase in the Chukchi Sea and decrease along parts of the western coast (Fig. 30A). Projected changes in Dp are similar to those under RCP 4.5 during the mid-century.
RCP 4.5: Top 5 Percent
Extreme Hs are projected to increase significantly during all seasons along the western and Arctic Alaska coasts (Figs. 26B-29B). Along the southern coast, changes in Hs are projected for the eastern section with increases of 10-20 cm, and western section with decreases of 10-20 cm during DJF (Fig. 26B). Mean Tp associated with extreme Hs are projected to increase by 0.25-0.5 s along the Arctic Alaskan coast; no appreciable changes are projected elsewhere. Most frequent Dp associated with extreme Hs are projected to experience similar but greater rotations as under RCP 4.5 and RCP 8.5 during the mid-century.
RCP 8.5: Mean
Mean Hs and Tp are projected to increase during all seasons along the Arctic Alaskan coast (Figs. 38A-45A). The largest increases in Hs (> 30 cm) and σ2Hs (> 40 cm) are projected during DJF, but other seasons are also expected to experience significant increases (20-30 cm). Large increases in σ2Hs are projected for the central Arctic Alaska coast, a spatial northeasterly transgression compared to projections under RCP 8.5 during the mid-century. Mean Tp are projected to increase by 0.25- 0.5 s along the Arctic Alaskan coast.
Hs are projected to increase along most of the western Alaskan coast during all seasons. Within the same area, mean Tp are projected to decrease during DJF. Negligible changes are projected for the remaining areas of the coast except during DJF when increases of 10-20 cm are projected along the eastern portion of the southern Alaska coast and decreases of 10-20 cm are projected in the vicinity of the western section of the Aleutian Island chain (Fig. 42A).
RCP 8.5: Top 5 Percent
Extreme Hs and mean Tp associated with extreme Hs are projected to increase substantially along the Arctic Alaskan coast during all seasons (Figs. 38B-45B). Most notable are the extreme δHs which consistently exceed more than 30 cm. Tp are projected to increase between 0.50-0.75 s, except in the Chukchi Sea where δTp is projected to increase more than 0.75 s during DJF.
Extreme Hs are projected to increase along the western coast and within most regions of the Bering Sea as well as along the eastern portion of the southern Alaskan coast during all seasons. Decreases of 20-30 cm are projected in the vicinity of the western section of the Aleutian Island chain during DJF. Surrounding this area, σ2Hs are projected to increase during all seasons.
Mean Tp associated with extreme Hs are projected to increase by 0.25-0.50 s along the western coast (Bering Sea) during MAM, JJA, and SON, whereas along the southern coast, δTp are negligible. Projected changes in Dp are similar to those under RCP 4.5 during the mid- and end-of-century with respect to direction: CCW rotations along eastern Arctic coast, CW rotations along western Arctic coast, and mostly CW rotations along the southern Alaskan coast. Projected changes in magnitude are similar to those under RCP 8.5 during the mid-century.
Figures 3 through 50 show maps of projected seasonal changes in mid- and end-of-century magnitude and variance of the mean and top 5 percent significant wave heights, peak periods, and peak directions compared to the historical time-period. Seasons are abbreviated as DJF for December through February, MAM for March through May, JJA for June through August, and SON for September through November.
West Coast [Figs. 51 through 98]
Mid-Century: 2026-2045
RCP 4.5: Mean
Mean Hs and Tp are projected undergo little to no change during all seasons throughout the region (Figs. 50A-57A). σ2Hs values are projected to decrease during DJF in the vicinity of the Oregon / California border. Most frequent Dp are projected to rotate only slightly in both CW and CCW directions.
RCP 4.5: Top 5 Percent
Extreme Hs are projected to decrease off the coast of southern Oregon and southward toward southern California during all seasons (Figs. 50B-53B). Decreases of 10-30 cm in Hs are projected for nearly the entire coast during DJF, with the greatest decreases extending from southern Oregon to central California (Fig. 50B). The region of greatest decrease is also expected to experience σ2Hs decreases by more than 40 cm. Mean Tp associated with extreme Hs are not projected to change appreciably (Figs. 54B-57B). Most frequent Dp associated with extreme Hs are projected to rotate CCW across much of the coast.
RCP 8.5: Mean
Similar to projected changes under RCP 4.5, little to no change in Hs and Tp are projected (Figs. 62A-69A). Most frequent Dp are projected to rotate only slightly in both CW and CCW directions.
RCP 8.5: Top 5 Percent
Projected extreme Hs patterns are similar to those under RCP 4.5 during the mid-century, but more pronounced. The California coast in particular is projected to experience spatially-continuous decreases in Hs and increasing variability. Mean Tp associated with extreme Hs are projected to have little to no change. Most frequent incident Dp are projected to rotate slightly CW, but σDp are projected to decrease between 5-10° in southern California (Fig. 70).
End-Century: 2081-2100
RCP 4.5: Mean
Similar to projected changes under RCP 4.5 during the mid-century, little to no change in Hs and Tp are projected (Figs. 74A-81A). Most frequent Dp are projected to rotate only slightly in both CW and CCW directions.
RCP 4.5: Top 5 Percent
Extreme Hs are projected to decrease along most of the California coast for all seasons, and to undergo little to no change over the remaining region. An exception is the projected increase of 10-20 cm in Hs at the very north end of the region (near the U.S. - Canada border). Mean Tp associated with extreme Hs are projected to increase by 0.25-0.50 s along most of the California coast during DJF. Most frequent Dp associated with extreme Hs are projected to rotate CCW for most of the region.
Extreme Hs are projected to decrease along most of the California coast, but by less than that projected during the mid-century. Additionally, end-of-century projected decreases are limited to the California coast, while mid-century decreases are projected for nearly the entire coast.
RCP 8.5: Mean
Little to no changes in mean Hs are projected for all seasons, but mean Tp are projected to increase by 0.25-0.50 s during MAM, JJA, and SON at nearly all locations within the region. No changes in mean Tp are projected during DJF.
RCP 8.5: Top 5 Percent
Extreme Hs and σ2Hs are projected to decrease substantially (by more than 30 cm and 40 cm, respectively) along nearly the entire region during all seasons, but most notably during DJF (Figs. 86B-89B). Mean Tp associated with extreme Hs are not projected to change appreciably. Most frequent Dp associated with extreme Hs are projected to rotate CW for much of the California coast during DJF and CCW offshore of Oregon during all seasons.
Figures 51 through 98 show maps of projected seasonal changes in mid- and end-of-century magnitude and variance of the mean and top 5 percent significant wave heights, peak periods, and peak directions compared to the historical time-period. Seasons are abbreviated as DJF for December through February, MAM for March through May, JJA for June through August, and SON for September through November.
East and Gulf Coasts [Figs. 99 through 146]
Mid-Century: 2026-2045
RCP 4.5: Mean
Mean Hs are projected to decrease by more than 20-30 cm along the entire East and Gulf coasts during DJF (Fig. 98A) and undergo little to no change during the remaining seasons. Mean Tp are projected to decrease along the northern section of the East Coast and at one location associated with the Gulf Coast during DJF (Fig. 102A) and to undergo little to no change elsewhere and during the other seasons.
RCP 4.5: Top 5 Percent
Extreme Hs are projected to decrease by 20-30 cm along the entire East and Gulf coasts during DJF (Fig. 98B) and to increase by 10-20 cm near 40°N during MAM, JJA, and SON (Figs. 99B-101B). Tp are projected to decrease along the northern half of the East Coast and within the Gulf Coast region during DJF and to undergo little to no change elsewhere and during the remaining seasons.
RCP 8.5: Mean
Similar to projections under RCP 4.5 during the mid-century, mean Hs are projected to decrease along the entire East and Gulf coasts during DJF (Fig. 110A) and to undergo little to no change during the remaining seasons. Compared to projections under RCP 4.5, however, the magnitudes of change are substantially greater with projected decreases greater than 30 cm. Tp are projected to decrease along the northern section of the East Coast and at one location associated with the Gulf Coast during DJF (Fig. 102A) and to undergo little to no change elsewhere and during the other seasons. The change in Tp (δTp) during DJF exhibits a spatial trend with greater decreases to the north and smaller decreases to the south.
RCP 8.5: Top 5 Percent
Extreme Hs are projected to decrease by more than 30 cm along the entire East and Gulf coasts during DJF (Fig. 110B) and to undergo little to no change during the remaining seasons. Mean Tp associated with extreme Hs are projected to decrease during DJF (Fig. 114B), with a trend of greater decreases to the north, and to undergo little to no change during the remaining seasons.
End-Century: 2081-2100
RCP 4.5: Mean
Mean Hs are projected to decrease by more than 20-30 cm, with most locations experiencing decreases greater than 30 cm, along the entire East and Gulf coasts during DJF (Fig. 122A) and undergo little to no change during the remaining seasons. Mean Tp are projected to decrease along the northern section of the East Coast and at one location associated with the Gulf Coast during DJF (Fig. 126A) and to undergo little to no change elsewhere and during the other seasons. A spatial trend of greater decreases to the north is evident along the New England coast.
RCP 4.5: Top 5 Percent
Extreme Hs are projected to decrease by more than 30 cm along the entire East and Gulf coasts during DJF (Fig. 122B) and to decrease by 10-20 cm along the mid and northern sections of the East Coast during MAM, JJA, and SON (Figs. 123B-125B). Tp are projected to decrease along the northern half of the East Coast and within the Gulf Coast region during DJF (Fig. 126B) and to undergo little to no change elsewhere and during the remaining seasons. During DJF, a spatial trend of greater decreases in δTp to the north is evident along the New England coast.
RCP 8.5: Mean
Mean Hs are projected to decrease by 20 cm to more than 30 cm along the entire East and Gulf coasts during DJF (Fig. 134A) and up to 20 cm along the northern part of the East Coast during the remaining seasons (Figs. 135A-137A). During DJF, projected decreases in σ2Hs are greatest (>40 cm) along the northern half of the East Coast and at the south end of the Gulf Coast. Tp are projected to decrease along the northern half of the East Coast and within the Gulf Coast region during DJF (Fig. 138A) and to undergo little to no change elsewhere and during the remaining seasons. During DJF, a spatial trend of greater decreases in δTp to the north is evident along the New England coast.
RCP 8.5: Top 5 Percent
Extreme Hs are projected to decrease by more than 30 cm during DJF (Fig. 134B) and by 10-30 cm during the remaining seasons along the entire East and Gulf coasts (Figs. 135B-137B). Mean Tp associated with extreme Hs are projected to decrease during DJF (Fig. 114B), with a trend of greater decreases to the north, and to undergo little to no change during the remaining seasons. Tp associated with extreme Hs are projected to decrease along the East Coast except the very southern portion, and within the Gulf Coast region during DJF (Fig. 138B) and to undergo little to no change elsewhere and during the remaining seasons. For the DJF season, a spatial trend of greater decreases in δTp to the north is evident along the outer open East Coast.
Figures 99 through 146 show maps of projected seasonal changes in mid- and end-of-century magnitude and variance of the mean and top 5 percent significant wave heights, peak periods, and peak directions compared to the historical time-period. Seasons are abbreviated as DJF for December through February, MAM for March through May, JJA for June through August, and SON for September through November.
Acknowledgments
This work was funded by the USGS Coastal and Marine Geology Program and carried out under the USGS Climate Change Impacts to the U.S. Pacific and Arctic Coasts Project. Gratitude is extended to Liv Herdman (USGS) and Kurt Rosenberger (USGS) for support in finalizing data tables and Karin Ohman (now with Michael Baker International and formerly with USGS) for maintaining runs during the initial stages of the work, Nadine Golden (USGS) for heading up the web posting, design, and delivery, James Shope (USGS and U.C. Santa Cruz) for assistance on various items, and Ann Gibbs for reviewing the document. We acknowledge the World Climate Research Programme's Working Group on Coupled Modeling, which is responsible for CMIP5, and we thank the climate modeling groups (listed in Table 2) for producing and making available their model output.
References Cited
Donat, M.G., Leckebusch, G.C., Wild, S., and Ulbrich, U., 2010. Benefits and limitations of regional multi-model ensembles for storm loss estimations: Climate Research, v. 44, p. 211–225.
Erikson, L.H., Hegermiller, C.A., Barnard, P.L., Ruggiero, P. andvan Ormondt, M., 2015b. Projected wave conditions in the Eastern North Pacific under the influence of two CMIP5 climate scenarios. Journal of Ocean Modelling, (0000)1-15, doi:10.1016/j.ocemod.2015.07.004
Erikson, L.H., Hemer, M.A., Lionello, P., Mendez, F.J., Mori, N., Semedo, A., Wang, X.L., Wolf, 2015a. Projection of wave conditions in response to climate change: A community approach to global and regional wave downscaling. Proceedings Coastal Sediments, 2015.
Hasselmann, S., Hasselmann, K., Allender, J. H., Barnett, T. P. , 1985. Computations and parameterizations of the nonlinear energy transfer in a gravity-wave spectrum. Part II: parameterizations of the nonlinear energy transfer for application in wave models, J. Physical Oceanography, 15, 1378–1391.
Meinshausen, M., and Coauthors, 2011. The RCP greenhouse gas concentrations and their extensions from 1765 to 2300. Climate Change, 109, 213-241.
Moss, R. H., and Coauthors, 2010: The next generation of scenarios for climate change research and assessment. Nature, 463, 747-756.
Riahi, K., Krey, V., Rao, S., Chirkov, V., Fischer, G., Kolp, P., Kindermann, G., Nakicenovic, N., and Rafai, P., 2011. RCP 8.5: Exploring the consequence of high emission trajectories. Climatic Change, 109(1-2), 33-57.
Storlazzi, C.D., Shope, J.B., Erikson, L.H., Hegermiller, C.A., and Barnard, P.L., 2015, Future wave and wind projections for United States and United States-affiliated Pacific Islands: U.S. Geological Survey Open-File Report 2015–1001, 426 p., doi:10.3133/ofr20151001.
Taylor, K.E., Stouffer,R.J., Meehl,G.A., 2012. An overview of CMIP5 and the experiment design. Bull. Am. Meteorol. Soc. 93, 485–498.
Thomson, A.M., Calvin, K.V., Smith, S.J., Kyle, G.P., Volke, A., Patel, P., Delgado-Arias, S., Bond-Lamberty, B., Wise, M.A., Clarke, L.E., Edmonds, J.A., 2011. RCP4.5: A Pathway for Stabilization of Radiative Forcing by 2100. Climatic Change, 109(1-2), 77-94.
Tolman, H. L., and D. Chalikov, 1996. Source Terms in a Third-Generation Wind Wave Model. Journal of Physical Oceanography, 26, 2497-2518
Tolman, H. L., Balasubramaniyan, B., Burroughs, L. D., Chalikov, D. V., Chao, Y. Y., Chen H. S., Gerald, V.M., 2002. Development and implementation of wind generated ocean surface wave models at NCEP. Weather and Forecasting, 17(2), 311-333.
van Vuuren, D.P., Edmonds, J.A., Kainuma, M., Riahi, K., Thomson, A.M., Hibbard, K., Hurtt, G.C., Kram, T., Krey, V., Lamarque, J-F., Masui, T., Meinshausen, M., Nakicenovic, N., Smith, S.J., and Rose. S, 2011. The representative concentration pathways: an overview. Climatic Change, 109, 5-31.
Conversion Factors
SI to Inch/Pound
| Multiply | By | To obtain |
|---|---|---|
| Length | ||
| centimeter (cm) | 0.3937 | inch (in.) |
| millimeter (mm) | 0.03937 | inch (in.) |
| meter (m) | 3.281 | foot (ft) |
| kilometer (km) | 0.6214 | mile (mi) |
| kilometer (km) | 0.54 | mile, nautical (nmi) |
| meter (m) | 1.094 | yard (yd) |
Abbreviations
Suggested citation: Erikson, L.H., Hegermiller, C.E., Barnard, P.L., and Storlazzi, C., 2016, Wave projections for United States mainland coasts: U.S. Geological Survey methods summary to accompany data release, https://doi.org/10.5066/F7D798GR.
