Skip to main content
NCBI home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2024 Jul 8;121(29):e2400355121. doi: 10.1073/pnas.2400355121

Evolving Arctic maritime hazards: Declining sea ice and rising waves in the Northwest Passage

Martin T Henke a,1, Tyler Miesse a, Andre de Souza de Lima a, Celso Ferreira a, Thomas Ravens b, Ralph Pundt c
PMCID: PMC11260157  PMID: 38976732

Significance

This study analyzes two interconnected hazards which threaten maritime operations in the Arctic: sea ice and ocean waves. Here, climate models are used to project future sea ice and to drive an Arctic wave model. From this, we derive a 50 y ensemble projection of sea ice risk and waves to analyze how both hazards evolve along the Northwest Passage maritime route. From this study, we find that as sea ice risk declines in the region there is a significant rise in waves. This phenomenon differs between months, as summer and winter months of July and November see similar trends in diminishing sea ice risk, the latter month experiences rapid increases in the wave climate compared to the former month.

Keywords: climate change, Arctic, maritime development, sea ice, ocean waves

Abstract

The ongoing and projected retreat of Arctic sea ice has garnered international interest toward the utilization of Arctic maritime corridors for shipping, tourism, and development. Yet, with potential for increasing traffic in Arctic regions, it’s important to consider additional environmental variables affected by climate change which may threaten maritime operations. Here, we use four climate model projections to produce ocean wave simulations and investigate the future magnitude and seasonality of sea ice risk coupled with wave hazards. Analyzing the potential 5 mo shipping season spanning July to November along the Northwest Passage maritime route between 2020 and 2070, our results show a substantial decline in sea ice risk over the analysis time period, resulting in near open-water conditions along the route for a 5 mo period by 2070. However, as seasonal ice coverage retreats, there is a significant upward trend in wave heights along the route during July and November, with the timing of the greatest wave height shifting away from September toward later in the season. This result is pertinent as the possibility of seasonally unprecedented extreme waves coupled with subfreezing late fall temperatures makes for an especially hazardous environment, thus emphasizing the importance of considering the interaction between evolving sea ice and interdependent hazards when predicting the risks and challenges faced by Arctic maritime operations.

The Arctic region has historically been largely inaccessible to maritime operations, as sea ice poses a substantial threat to all but the most capable vessels. Yet, as the observed retreat of Arctic sea ice (1) is projected to continue into the next century and eventually reach near ice-free conditions by midcentury (2–4), there is now considerable interest in commercial activities such as transpolar shipping, resource extraction, and tourism within this region (5–8), as evidenced by increasing maritime traffic over recent decades (9). Arctic maritime routes such as the Northwest Passage (NWP) or Northern Sea Route used for transpolar shipping offer shorter voyage lengths in comparison to alternative routes, thereby yielding reduced travel times and fuel consumption (10, 11). However, increased development in the region also brings greater exposure to the distinct environmental hazards of the Arctic. Considering that Arctic ecosystems are highly sensitive to the effects of maritime accidents (12) and the high risk of rescue operations within the region (13), research evaluating future environmental hazards is crucial for informing future Arctic maritime activities.

Sea ice is the critical environmental variable that has historically occluded Arctic maritime passage and is currently present within the Arctic year-round, cycling between the annual maximum and minimum extents in March and September, respectively (14). Maritime operations are endangered by sea ice as vessel–ice collision may cause hull breaches and thick ice-pack can immobilize and disable the vessel. Moreover, sea ice represents an economic disincentive as freight passage in proximity to sea ice demands slower speeds, alternative routes, and costly ice-breaker assistance and escort fees (15–17). Yet, over the recent past decades, the prevalence of sea ice in the Arctic has diminished substantially, with a significant downward trend in seasonal ice coverage and a pan-Arctic thinning of ice as dense multiyear ice gives way to single-year ice (18, 19). These trends are projected to continue into future decades even under moderate climate emission scenarios (2, 20–22) and have sparked interest in investigating navigational conditions along Arctic maritime routes dependent on future sea ice coverage (23–30); with studies generally agreeing that even non-ice-strengthened vessels may be able to navigate the Arctic for multiple months by midcentury. These analyses of Arctic maritime operations have been extended further beyond navigability, examining additional factors affecting feasibility such as economic viability, geopolitical implications, environmental ramifications, and logistical constraints (5, 15, 31–38). Within the transportation system, operations, and governance research literature, there is frequent acknowledgment that there are multiple unique environmental hazards that threaten Arctic maritime transit (39); yet, in most geophysical research, projection, and assessment of Arctic environmental hazards is typically limited solely to sea ice. Given that many environmental variables will change in-tandem with climate change and sea ice loss, an analysis of future projections of these additional hazards is needed.

One such environmental variable regulated by sea ice is ocean wave heights, as ice prevents wave growth and dissipates incident waves (40–42). With the decline in sea ice, the regional wave climate of the Arctic has increased substantially over recent decades (43–48). Two factors primarily drive this trend: i) increasing open-water area allows for extended fetch and thus larger waves and the development of swells and ii) the lengthened open-water season extends further into fall and winter when storm climatology is more severe. Through improved modeling techniques, recent studies have projected the future state of the Arctic wave climate (48–55). These analyses agree that as sea ice recedes, there will be significant future increases in the wave climate of Arctic regions and this growth is already associated with increasing North American Arctic coastal wave hazards. These trends are projected to continue over coming decades with further ice loss exacerbating the threat of inundation and coastal erosion (55–58). From the perspective of Arctic maritime traffic, this connection between declining sea ice and wave growth has not been rigorously assessed—yet is needed, as ocean waves hinder maritime activity by slowing transit speed, affect safety, and potentially lead to accidents or pollutant emissions (59, 60). Moreover, waves pose an additional threat in freezing climates, as sea-spray generated by vessel-wave interaction freezes and accretes onto vessels in a phenomenon known as vessel icing—potentially disabling or even capsizing the vessel (61–63). At northern latitudes, this phenomenon has been responsible for numerous lost vessels and resulting loss of life (63, 64). Thus, understanding future wave conditions within the Arctic is especially vital for informing safe maritime vessel design loads, standards, and practices.

The objective of this study is to compare the evolution of two major interdependent environmental hazards that threaten maritime operations in the Arctic: sea ice and waves. In this effort, four climate models (listed in Methods) belonging to the future climate projection shared socioeconomic pathway (SSP) 5 to 8.5 are used to create a multimodel ensemble (MME) projection of both variables which are analyzed through the lens of risks which threaten maritime shipping. To focus analysis on regions of interest in future development, this study is limited to the seas that make up the northern portion of the NWP route. Projections of sea ice concentration and thickness are sourced from the MME, and the polar operational limit assessment risk index system (POLARIS) (65) sea ice risk assessment framework is applied to derive future risk levels specific to a non-ice-strengthened maritime vessel. To assess wave hazards, the MME provides wind forcing and sea ice fields as inputs into an Pan-Arctic implementation of the wave model WaveWatchIII (WW3) (66) to project future waves; these wave simulations are then analyzed and trends in the monthly 90th percentile of significant wave height (Hs,90: 90th percentile of the average of the highest third of wave heights) are presented.

This analysis is focused further on three distinct subregions which lie along the NWP: the Beaufort Sea (BS), Canadian Archipeligo (CAA), and Baffin Bay (BB) (67) as shown in Fig. 1. It should be noted that while there are mutiple routes that serve as passage through the NWP, this dimension of analysis is not included in this study as the entire region is sampled evenly. The analysis is also divided by months, analyzing the sea ice seasonal low month of September and the surrounding “shoulder” months of July and November.

Fig. 1.

Fig. 1.

Open in a new tab

The subregions of analysis making up the Northwestern Passage route analyzed within this study. Subregions were delineated to aid in the discussion of spatial differences in wave hazards and ice risk and to provide sampling areas for narrowing results according to regions of interest in the following figures.

Results

Sea Ice Decline.

First, analyzing future sea ice area (SIA—defined as the sum of area multiplied by respective sea ice concentration) within the three subregions, Fig. 2 presents monthly time series for the four climate models and the ensemble average. While SIA does not directly correspond to sea ice risk, it communicates the trend and timeline of sea ice coverage along the route and indicates growing open-water conditions—favoring reduced sea ice risk and a growing wave climate. Within each of the three subregions for July and November, Fig. 2 shows that at least three out of four ensemble members agree average regional SIA has a significant decreasing trend detected at the 95% confidence level using a modified Mann–Kendall (MK) method (see SI Appendix, SI Materials and Methods section Trend Analysis: for further details) over the 50 y time series.

Fig. 2.

Fig. 2.

Open in a new tab

Regional monthly (A to I) time series of SIA for the ensemble average (black) and individual climate models (colors). Symbols in the top right corner denote the detection of a significant decreasing trend for individual models. The averaged trend value is only provided for regions and months where at least three ensemble members detect a significant decreasing trend and is derived as the mean of those ensemble member’s Sen’s slopes. Due to the varying resolution in climate model grids, the maximum SIA (SIA assuming all grid cells within the region have 100% SIC) varies between climate models analyzed. However, averaging over the ensemble, the maximum SIA for BS, BB, and CAA is 0.42, 0.43, and 0.53 × 106 km2, respectively, and are plotted as a dashed horizontal line within each subplot.

For September (the annual SIA minimum), SIA is already well below the regional maximum for all three regions at the beginning of the time series in 2020 and a significant downward trend is only detected in two of the ensemble members for BS and BB (Fig. 2 B and E). Nonetheless, the ensemble average shows September SIA for these regions and CAA to decline to near-zero values by 2050 with only one ensemble member intermittently showing sea ice presence within the region. Comparing the shoulder months, July and November, the former shows reliably less SIA beginning in 2020 for all three regions in comparison to the latter. Despite this, November shows a more aggressive downward trend in SIA for all regions and reaches similar levels to July by 2070 with the exception of BB. For BB, by 2070, July becomes nearly ice-free, yet November still shows some degree of ice coverage due to the EC-Earth3 model exhibiting high SIA variability within this region as shown in Fig. 2F.

Sea Ice Risk and Wave Hazards.

As seasonal coverage of sea ice declines, the NWP becomes more accessible to maritime traffic and shows a marked increase in high waves. This is shown in Fig. 3, where monthly averages of risk index outcomes (RIO) and Hs,90 are plotted for the NWP and averaged for periods 2020 to 2040 and 2050 to 2070. It is important to note that all RIO figures presented within this study are specific to non-ice-strengthened maritime vessels for which RIO > 0 indicates safe operation and RIO < 0 denotes unsafe operational conditions. Further explanation and the methods employed in deriving RIO values can be found within SI Appendix, SI Materials and Methods section.

Fig. 3.

Fig. 3.

Open in a new tab

Ensemble averaged decadal RIO and Hs,90 fields for 2020 to 2040 and 2050 to 2070 for the months of July, September, and November.

Examining the near period 2020 to 2040, September is the only month to possess positive RIO values through much of the NWP. However, even still during this month, unsafe sea ice conditions persist and threaten passage at the northern border of the BS and throughout the region north of the CAA and the Viscount Melville Sound. During July and November, the NWP is shown to be impassable for UC vessels due to negative RIO throughout the CAA and encroaching on the Alaskan Arctic coast. While these shoulder months show similar sea ice risk during this period, the H s,90 fields for July and November contrast greatly. Where Hs,90 is reliably below 1.5 m for BB and BS in July, November shows Hs,90 to exceed 2 m along the BS coast and in certain regions of BB. Within the CAA, where these months both reliably have high sea ice coverage, Hs,90 is limited to <1 m. September, with the largest open water area, demonstrates the highest Hs,90 values during this period, with the BS, BB, and CAA showing H s,90 ranging up to 2.5 m, 3 m, and 2 m, respectively.

The RIO fields derived for the period 2050 to 2070 show a substantial shift in the associated sea ice risk present along the NWP. For this period, the entirety of the Arctic is shown to be safe for navigation in the month of September. Where sea ice occluded passage through much of the NWP for July and November during the 2020 to 2040 period, 2050 to 2070 shows the route to be largely safe for operation with elevated risk only present in the northern portion of CAA. July, however, does show elevated risk values (RIO < 0) persisting southward from the pole, whereas positive RIO values extend much further northward beyond the CAA for November. To provide further insight into the seasonal opening for the NWP, SI Appendix, Fig. S5 displays the monthly distribution of RIO values within the subregions over the extended period of June through December for different decades. This figure shows that in 2020 to 2030, during only 1 mo (September) all subregions are approximately 90% safe by possessing RIO > 0. By 2040 to 2050, this period is extended to 3 mo (August to October), and by 2060 to 2070, it is extended to 5 mo (July to November). These results demonstrate a considerable decrease in sea ice risk throughout the passage

Examining wave hazards shown in Fig. 3, the 2050 to 2070 period column presents a considerable shift in seasonal wave heights. For this period, it can be seen that November has overtaken September as the month with the highest Hs,90 values in BB and CAA despite September exhibiting greater open-water areas. Comparing November 2020 to 2040 and 2050 to 2070, large increases in Hs,90 can be observed for the BS and CAA where previously Hs,90 was extremely limited and in 2050 to 2070 ranges up to 4 m and 2 m, respectively. Contrasting this considerable rise in the November wave climate, Hs,90 fields for July increase only slightly in similar open-water areas of the BS and CAA, which range at approximately 2 m and 1 m, respectively. Examining all months sequentially, Hs,90 fields can be observed to grow seasonally from July until November. This result demonstrates that in the absence of sea ice, the month of greatest wave hazards in this region will shift toward winter—a result consistent with current literature on the topic (51, 55).

To gain further insight into the regional and seasonal characteristics of Hs,90 growth, the following analysis was undertaken: Monthly, Hs,90 was derived over the analysis period 2020 to 2070 to yield a time series specific to each simulation node, the modified MK test method was employed at the 95% confidence level to test each node’s Hs,90 time series for a significant increasing trend, and the computed nodal trend from Sen’s slope estimator was averaged among ensemble members and plotted in Fig. 4 as the Δ Hs,90/decade contours. Model agreement is mapped through hatching schemes (48, 68) to indicate robust trends and agreement in sign of change for the trend.

Fig. 4.

Fig. 4.

Open in a new tab

The ensemble averaged linear decadal trend in monthly Hs,90 derived for each model node and plotted as contours. To assess robustness, no hatching indicates a significant trend detected by the modified MK method at the 95% confidence level in half or more of the ensemble members and ≥75% of those agree on the sign of change. Cross-hatching indicates that a significant trend was detected in at least half of the ensemble members, but of those ensemble members, ≥75% sign consensus was not reached—denoting conflicting trends for the area. Single hatching indicates that no significant trend was found in any or for more than one member.

In July, a robust increasing trend in the extreme wave climate is detected within the CAA and in the northern part of the BS toward the central Arctic. Yet, for this month, rates of Hs,90 growth widely do not reach more than 20 cm/decade within the NWP. In contrast, November shows rapid growth in the extreme wave climate as all three regions are shown to widely possess significant increasing trends over 30 cm/decade. The greatest growth for this month occurs in Western BS where Hs,90 growth rates exceed 50 cm/decade. These results represent a rapid and widespread increase in extreme wave heights for the region, and agree with existing literature showing historical trends in Arctic extreme wave heights have reached up to 100 cm/decade, with the greatest trends concentrated in fall months (48).

While September exhibited some growth in Hs,90 as seen in Figs. 3 and 4 elucidates that much of this increase took place outside or at the fringes of the BS, BB, and CAA subregions—as a positive significant trend in Hs,90 is detected only at the northern edges of the CAA and BS and intensifies toward the Central Arctic. Despite September showing a significant decline in SIA from 2020 through 2070, the majority of this loss was outside the subregions and had little effect on the subregional Hs,90. This result is expected for regions such as the CAA where fetch is inherently limited by the numerous islands within the region, yet more surprising for areas such as the BS where potential fetch in the northward direction grew substantially as ice receded into the Central Arctic.

To better understand the timeline of regional wave climate growth and the variability in the ensemble wave projection, SI Appendix, Fig. S6 shows the time series of the median monthly Hs,90 of all nodes within each subregion averaged for the ensemble and individual climate models. This figure confirms the widespread growth in the November wave climate, as it far outpaces July (doubling the linear trend in CAA) and eventually reaches a 2060 to 2070 average median Hs,90 for CAA, BS, and BB at approximately 1.5, 2.5, and 2.3 m; as opposed to 0.8, 1.5, and 1.4 m for July; and 1.4, 2.2, and 2.1 m for September. In agreement with results shown for September in Fig. 4 and SI Appendix, Fig. S6 shows no significant trend in the median monthly H s,90 for the three subregions. While change in sea ice cover is the primary variable causing the observed increases in Hs,90 fields, declining sea ice area alone does not explain the stark difference in the wave climate of November over July. To gain insight into the wind climate that drives this seasonal contrast, averages of monthly 90th percentile surface wind speeds for the regions of analysis are plotted in SI Appendix, Fig. S7. This figure confirms that stronger peak wind speeds in November drive the intensity of the increasing trend for this month and eventually lead to the observed wave hazards surpassing July and September once open-water condition persist later into the season.

Discussion

From this analysis, we derive the following key findings:

  • i.

    Sea ice risk diminishes substantially along the NWP in future decades, reaching 3 mo of nearly open navigation by 2050 and 5 mo by 2070.

  • ii.

    As ice declines for all months, there is significant growth in the along-route wave climates for both the shoulder months, July and November. As September sea ice coverage was already low within the BS, BB, and CAA subregions at the beginning of the simulation, further losses in sea ice for this month did not result in significant increases in the subregions’ monthly Hs,90 fields.

  • iii.

    Despite having similar sea ice coverage, July and November diverge significantly in future wave climate due to wind forcings—with November far outpacing July and eventually possessing higher wave values than September by 2070.

Regarding key finding, i) this result agrees with similar studies which have applied POLARIS to study the seasonal navigational window length for the NWP (25, 27, 29). The seasonal length of the shipping navigation is a key component determining the economic viability of Arctic routes (13, 24). Taking this knowledge into account through key finding ii), we find that further attention should be given to “shoulder” months, as the summer and late fall months of July and November may be safer or riskier, respectively, due to the demonstrated contrast in wave hazards between the seasons. This is further emphasized by key finding and iii), where there is a substantial seasonal shift in the peak extreme regional waves away from September and later into the late fall and early winter season. High waves occurring in the early-winter and intersecting with declining temperatures, growing sea ice coverage, and diminishing daylight suggest that this season may be highly hazardous to maritime operations.

Beyond the inherent risk posed by waves to vessels regardless of the location, waves are the dominant source of another unique polar hazard—vessel icing. In Arctic maritime operations, the phenomenon of vessel icing is well known and has caused numerous lost vessels. This process is complex and dependent on multiple variables such as vessel type and speed, wind speed, wave height and direction, and air and sea temperatures (63). Ryerson (61) highlights several cases where shipping vessels with high freeboard experienced significant ice accumulation on the NWP causing navigational difficulties or requiring stops for deicing measures to be taken. While the current level of vessel icing risk along the NWP is unclear, Ryerson (61) mentions that, “few ship or superstructure icing observations have been made in the BS because the area is relatively uninhabited, and historically sea ice has not retreated off shore for large distances.” This statement suggests that the region has the potential for severe icing rates exceeding 2 cm/h during the fall and early winter if open water conditions might intersect with colder temperatures. As our projections show the November wave climate of all three regions to be greater than that of September, this result implies a hazardous outcome, as larger waves (creating more sea spray) combined with colder November air temperatures (well below freezing) may result in ideal conditions for heavy ice accumulation—a risk that would be further exacerbated by increasing exposure due to growing development within the region. This suggested shift in seasonal icing hazards has not been explored in academic literature and a detailed analysis is needed to better account for future environmental hazards encountered in Arctic maritime operations.

Finally, it’s necessary to acknowledge several limitations within the methodology and analysis performed by this study. First, waves have a significant effect on sea ice, causing it to break up and be more susceptible to melting (23, 69), yet these processes are not represented within our modeling framework. Two-way coupling of ice and waves is an important step for more comprehensive future sea ice modeling methodologies—yet such advanced modeling techniques are still in development and were beyond the scope of this study. Furthermore, it should be recognized that climate models may possess considerable bias in simulating the mean spatial distribution of sea ice thickness in the Arctic (70). This limitation imposes uncertainty in the derived RIO values especially in regions where multiyear ice may be present such as the Canadian Archipelago (25). In this analysis, the analyzed UC vessel’s acceptable risk is quickly exceeded by even thin ice, and thus biases in sea ice thickness may have limited effect on the assessment of accessibility. Nonetheless, this outstanding limitation remains an area of future improvement and could have considerable ramifications for any study wishing to project the seasonal window of Arctic navigability for more ice-capable vessels. Another limitation of analysis in this study was that the role of changing wind climatology is only given cursory analysis. Yet, changing wind climatology may be an important consideration for certain regions, as sea ice loss is linked to stronger surface winds and intensified storm climatology (71, 72). For example, Wang et al. (73) found that through the historical record, interdecadal changes in wind speed and direction have affected mean wave heights in BB. Changes in surface wind climate are simulated in this study’s wave simulations’ use of coupled ocean-atmosphere climate models; yet, analysis of monthly 90th percentile surface wind speeds averaged for the regions of analysis shown in SI Appendix, Fig. S7 did not yield a robust significant trend for any month or region. Thus, in this analysis, sea ice is shown to be the primary changing variable resulting in increasing wave heights. Nonetheless, changing mean wind and storm climatology is the topic of significant ongoing research (74) and robust regional projections and analyses are needed for better a understanding of future Arctic wave hazards. Finally, it should be recognized that all climate projections utilized in this study belong to only a single future climate pathway: SSP5-8.5. This climate scenario represents a worst-case emissions scenario and different climate scenarios would undoubtably yield different timelines for declining sea ice risk and the growing wave climate—especially after 2050 when the different climate scenarios begin to show greater divergence in sea ice (75).

Methods

Future Projections.

Four global climate models, CRNM-CM-1-HR, EC-Earth3, MPI-ESM1-2-HR, and MRI-ESM2-0, (SI Appendix, Table S1) from the Climate Model Intercomparison Project’s Sixth Phase on the SSP 5-8.5 future climate scenario were used in projecting future sea ice and providing driving data for wave simulations. All results discussed and figures shown in the manuscript are a result of averaging derived RIO fields or wave simulations driven by individual climate model forcing data. Additional discussion pertaining to the ensemble models is provided in SI Appendix, SI Materials and Methods.

Sea Ice Risk.

To assess sea ice risk, the POLARIS (65) framework as specified by the International Maritime Organization’s Polar Code is used. This framework presents a method to determine whether an area is safe for maritime operation accounting for sea ice concentration, thickness, and the assigned ice-class of a maritime vessel. In this analysis, all figures shown are specific to one vessel class: Unclassified (UC) (also referred to as open-water or non-ice-strengthened) vessels. Specific equations detailing how this framework was employed and a discussion on different vessel classification are presented in SI Appendix, SI Materials and Methods.

Wave Simulations.

The third-generation wave model WaveWatchIII (66) was utilized to project future wave fields on an unstructured Pan-Arctic grid. The model grid covers the Arctic region from 50°N reaching to 89.75°N with a flexible unstructured mesh of varying resolution. Model inputs include 6-h surface winds and daily sea ice concentration fields. Wave–ice interaction is parameterized through a simple ice-blocking formula, where model grid points which exceed the critical threshold of 50% ice concentration are assumed to have no waves and are deactivated from active simulation until sea ice concentration falls below the threshold again. Further discussion on the use of the simple ice-blocking parameterization is provided within SI Appendix, SI Materials and Methods. The model was validated in a two-step process: i) understand model performance under ideal forcings conditions and ii) gain insight into the potential bias within the climate model ensemble wave projection. Further details for the wave model including grid details, parameterizations, and the results of the validation process are provided in SI Appendix, SI Materials and Methods.

Supplementary Material

Appendix 01 (PDF)

pnas.2400355121.sapp.pdf (2.3MB, pdf)

Acknowledgments

Author contributions

M.T.H., T.M., C.F., T.R., and R.P. designed research; M.T.H. and T.M. performed research; M.T.H. and A.d.S.d.L. analyzed data; and M.T.H., A.d.S.d.L., and R.P. wrote the paper.

Competing interests

The authors declare no competing interest.

Footnotes

This article is a PNAS Direct Submission.

Although PNAS asks authors to adhere to United Nations naming conventions for maps (https://www.un.org/geospatial/mapsgeo), our policy is to publish maps as provided by the authors.

Data, Materials, and Software Availability

Python codes used to process model outputs and generate figures in this article are available at https://doi.org/10.5281/zenodo.12151436 (76). The climate model data used within this study is available through the World Climate Research Programme at https://esgf-node.llnl.gov/projects/cmip6/ (77–80). ERA5 surface winds and sea ice concentration used in validation are available through the climate data store at https://cds.climate.copernicus.eu/ (81). Satellite validation utilized the ESA Remote Sensing Significant Wave Height L3 multisatellite product available at https://archive.ceda.ac.uk/ (82). The simulated wave data generated from the wave model is available at https://doi.org/10.18739/A2ST7DZ74.

Supporting Information

References

  • 1.Stroeve J., Notz D., Changing state of Arctic sea ice across all seasons. Environ. Res. Lett. 13, 103001 (2018). [Google Scholar]
  • 2.Docquier D., Koenigk T., Observation-based selection of climate models projects Arctic ice-free summers around 2035. Commun. Earth Environ. 2, 144 (2021). [Google Scholar]
  • 3.Årthun M., Onarheim I. H., Dörr J., Eldevik T., The seasonal and regional transition to an ice-free Arctic. Geophys. Res. Lett. 48, e2020GL090825 (2021). [Google Scholar]
  • 4.Barnhart K. R., Miller C. R., Overeem I., Kay J. E., Mapping the future expansion of Arctic open water. Nat. Clim. Change 6, 280–285 (2016). [Google Scholar]
  • 5.Bennett M. M., Stephenson S. R., Yang K., Bravo M. T., De Jonghe B., The opening of the Transpolar Sea Route: Logistical, geopolitical, environmental, and socioeconomic impacts. Marine Policy 121, 104178 (2020). [Google Scholar]
  • 6.Boylan B. M., Increased maritime traffic in the Arctic: Implications for governance of Arctic sea routes. Marine Policy 131, 104566 (2021). [Google Scholar]
  • 7.Post E., et al. , The polar regions in a 2°C warmer world. Sci. Adv. 5, eaaw9883 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Zeng Q., Lu T., Lin K.-C., Yuen K. F., Li K. X., The competitiveness of Arctic shipping over Suez Canal and China-Europe railway. Transport Policy 86, 34–43 (2020). [Google Scholar]
  • 9.Dawson J., Pizzolato L., Howell S. E. L., Copland L., Johnston M. E., Temporal and spatial patterns of ship traffic in the Canadian Arctic from 1990 to 2015 + supplementary appendix 1: Figs. S1–S7 (See Article tools). Arctic 71, 15 (2018). [Google Scholar]
  • 10.Stephenson S. R., Smith L. C., Agnew J. A., Divergent long-term trajectories of human access to the Arctic. Nat. Clim. Change 1, 156–160 (2011). [Google Scholar]
  • 11.Somanathan S., Flynn P., Szymanski J., The Northwest Passage: A simulation. Transp. Res. Part A: Policy Practice 43, 127–135 (2009). [Google Scholar]
  • 12.Hauser D. D. W., Laidre K. L., Stern H. L., Vulnerability of Arctic marine mammals to vessel traffic in the increasingly ice-free Northwest Passage and Northern Sea Route. Proc. Natl. Acad. Sci. U.S.A. 115, 7617–7622 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Ellis B., Brigham L., “Arctic council: Arctic marine shipping assessment 2009 report” (Arctic Council, Tromsø, Norway, 2009). [Google Scholar]
  • 14.Fetterer F., Knowles K., Meier W. N., Savoie M., Windnagel A. K., Sea ice index (Version 3, National Snow and Ice Data Center, 2017).
  • 15.Cariou P., Cheaitou A., Faury O., Hamdan S., The feasibility of Arctic container shipping: The economic and environmental impacts of ice thickness. Marit. Econ. Logist. 23, 615–631 (2021). [Google Scholar]
  • 16.Sibul G., Jin J. G., Evaluating the feasibility of combined use of the Northern Sea Route and the Suez Canal Route considering ice parameters. Transp. Res. Part A: Policy Practice 147, 350–369 (2021). [Google Scholar]
  • 17.Zhou X., et al. , Revisiting Trans-Arctic maritime navigability in 2011–2016 from the perspective of sea ice thickness. Remote Sens. 13, 2766 (2021). [Google Scholar]
  • 18.Kwok R., Arctic sea ice thickness, volume, and multiyear ice coverage: Losses and coupled variability (1958–2018). Environ. Res. Lett. 13, 105005 (2018). [Google Scholar]
  • 19.Maslanik J. A., et al. , A younger, thinner Arctic ice cover: Increased potential for rapid, extensive sea-ice loss. Geophys. Res. Lett. 34, L24501 (2007). [Google Scholar]
  • 20.Jahn A., Kay J. E., Holland M. M., Hall D. M., How predictable is the timing of a summer ice-free Arctic? Geophys. Res. Lett. 43, 9113–9120 (2016). [Google Scholar]
  • 21.Wang M., Overland J. E., A sea ice free summer Arctic within 30 years: An update from CMIP5 models: SUMMER ARCTIC SEA ICE. Geophys. Res. Lett. 39, L18501 (2012). [Google Scholar]
  • 22.Sigmond M., Fyfe J. C., Swart N. C., Ice-free Arctic projections under the Paris agreement. Nat. Clim. Change 8, 404–408 (2018). [Google Scholar]
  • 23.Aksenov Y., et al. , On the future navigability of Arctic sea routes: High-resolution projections of the Arctic Ocean and sea ice. Marine Policy 75, 300–317 (2017). [Google Scholar]
  • 24.Li X., et al. , Arctic shipping guidance from the CMIP6 ensemble on operational and infrastructural timescales. Clim. Change 167, 23 (2021). [Google Scholar]
  • 25.Mudryk L. R., et al. , Impact of 1, 2 and 4 °C of global warming on ship navigation in the Canadian Arctic. Nat. Clim. Change 11, 673–679 (2021). [Google Scholar]
  • 26.Smith L. C., Stephenson S. R., New Trans-Arctic shipping routes navigable by midcentury. Proc. Natl. Acad. Sci. U.S.A. 110, E1191–E1195 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Wei T., Yan Q., Qi W., Ding M., Wang C., Projections of Arctic sea ice conditions and shipping routes in the twenty-first century using CMIP6 forcing scenarios. Environ. Res. Lett. 15, 104079 (2020). [Google Scholar]
  • 28.Melia N., Haines K., Hawkins E., Sea ice decline and 21st century trans-Arctic shipping routes: Trans-Arctic shipping in the 21st Century. Geophys. Res. Lett. 43, 9720–9728 (2016). [Google Scholar]
  • 29.Min C., et al. , The emerging Arctic shipping corridors. Geophys. Res. Lett. 49, e2022GL099157 (2022). [Google Scholar]
  • 30.Stephenson S. R., Smith L. C., Influence of climate model variability on projected Arctic shipping futures. Earth’s Future 3, 331–343 (2015). [Google Scholar]
  • 31.Ng A. K. Y., Andrews J., Babb D., Lin Y., Becker A., Implications of climate change for shipping: Opening the Arctic seas. WIREs Clim. Change 9, e507 (2018). [Google Scholar]
  • 32.Didenko N. I., Cherenkov V. I., Economic and geopolitical aspects of developing the Northern Sea Route. IOP Conf. Ser.: Earth Environ. Sci. 180, 012012 (2018). [Google Scholar]
  • 33.Liu C.-Y., Fan H.-M., Dang X., Zhang X., The Arctic policy and port development along the Northern Sea Route: Evidence from Russia’s Arctic strategy. Ocean Coastal Manage. 201, 105422 (2021). [Google Scholar]
  • 34.Theocharis D., Pettit S., Rodrigues V. S., Haider J., Arctic shipping: A systematic literature review of comparative studies. J. Transp. Geogr. 69, 112–128 (2018). [Google Scholar]
  • 35.Zhang Z., Huisingh D., Song M., Exploitation of trans-Arctic maritime transportation. J. Cleaner Prod. 212, 960–973 (2019). [Google Scholar]
  • 36.Yumashev D., Van Hussen K., Gille J., Whiteman G., Towards a balanced view of Arctic shipping: Estimating economic impacts of emissions from increased traffic on the Northern Sea Route. Clim. Change 143, 143–155 (2017). [Google Scholar]
  • 37.Lynch A. H., Norchi C. H., Li X., The interaction of ice and law in Arctic marine accessibility. Proc. Natl. Acad. Sci. U.S.A. 119, e2202720119 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Panahi R., Ng A. K. Y., Afenyo M. K., Haeri F., A novel approach in probabilistic quantification of risks within the context of maritime supply chain: The case of extreme weather events in the Arctic. Accid. Anal. Prev. 144, 105673 (2020). [DOI] [PubMed] [Google Scholar]
  • 39.Lin B., et al. , An overview of scholarly literature on navigation hazards in Arctic shipping routes. Environ. Sci. Pollut. Res., 10.1007/s11356-023-29050-2 (2023). [DOI] [PubMed] [Google Scholar]
  • 40.Squire V. A., Ocean wave interactions with sea ice: A reappraisal. Annu. Rev. Fluid Mech. 52, 37–60 (2020). [Google Scholar]
  • 41.Squire V. A., A fresh look at how ocean waves and sea ice interact. Philos. Trans. A Math. Phys. Eng. Sci. 376, 20170342 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Collins C. O., Rogers W. E., Lund B., An investigation into the dispersion of ocean surface waves in sea ice. Ocean Dyn. 67, 263–280 (2017). [Google Scholar]
  • 43.Thomson J., Rogers W. E., Swell and sea in the emerging Arctic Ocean. Geophys. Res. Lett. 41, 3136–3140 (2014). [Google Scholar]
  • 44.Stopa J. E., Ardhuin F., Girard-Ardhuin F., Wave climate in the Arctic 1992–2014: Seasonality and trends. Cryosphere 10, 1605–1629 (2016). [Google Scholar]
  • 45.Liu Q., Babanin A. V., Zieger S., Young I. R., Guan C., Wind and wave climate in the Arctic Ocean as observed by altimeters. J. Clim. 29, 7957–7975 (2016). [Google Scholar]
  • 46.Thomson J., et al. , Emerging trends in the sea state of the Beaufort and Chukchi seas. Ocean Model. 105, 1–12 (2016). [Google Scholar]
  • 47.Waseda T., et al. , Correlated increase of high ocean waves and winds in the ice-free waters of the Arctic Ocean. Sci. Rep. 8, 4489 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Casas-Prat M., et al. , Wind-wave climate changes and their impacts. Nat. Rev. Earth Environ. 5, 23–42 (2024). [Google Scholar]
  • 49.Morim J., et al. , A global ensemble of ocean wave climate projections from CMIP5-driven models. Sci. Data 7, 105 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Morim J., et al. , Robustness and uncertainties in global multivariate wind-wave climate projections. Nat. Clim. Change 9, 711–718 (2019). [Google Scholar]
  • 51.Casas-Prat M., Wang X. L., Sea ice retreat contributes to projected increases in extreme Arctic Ocean surface waves. Geophys. Res. Lett. 47, e2020GL088100 (2020). [Google Scholar]
  • 52.Khon V. C., et al. , Wave heights in the 21st century Arctic Ocean simulated with a regional climate model: KHON ET AL. Geophys. Res. Lett. 41, 2956–2961 (2014). [Google Scholar]
  • 53.Freitas A., Bernardino M., Guedes Soares C., The influence of the Arctic Oscillation on North Atlantic wind and wave climate by the end of the 21st century. Ocean Eng. 246, 110634 (2022). [Google Scholar]
  • 54.Casas-Prat M., Wang X. L., Swart N., CMIP5-based global wave climate projections including the entire Arctic Ocean. Ocean Model. 123, 66–85 (2018). [Google Scholar]
  • 55.Casas-Prat M., Wang X. L., Projections of extreme ocean waves in the Arctic and potential implications for coastal inundation and erosion. J. Geophys. Res. Oceans 125, e2019JC015745 (2020). [Google Scholar]
  • 56.Barnhart K. R., Overeem I., Anderson R. S., The effect of changing sea ice on the physical vulnerability of Arctic coasts. Cryosphere 8, 1777–1799 (2014). [Google Scholar]
  • 57.Camus P., et al. , Statistical wave climate projections for coastal impact assessments: STATISTICAL WAVE CLIMATE PROJECTIONS. Earth’s Future 5, 918–933 (2017). [Google Scholar]
  • 58.Nederhoff K., Erikson L., Engelstad A., Bieniek P., Kasper J., The effect of changing sea ice on wave climate trends along Alaska’s central Beaufort Sea coast. Cryosphere 16, 1609–1629 (2022). [Google Scholar]
  • 59.Knapp S., Shen J., The effect of changes in wind strength and wave heights on the safety of vessels in shipping. Accid. Anal. Prev. 43, 1252–1266 (2011). [DOI] [PubMed] [Google Scholar]
  • 60.Heij C., Knapp S., Effects of wind strength and wave height on ship incident risk: Regional trends and seasonality. Transp. Res. Part D: Transp. Environ. 37, 29–39 (2015). [Google Scholar]
  • 61.Ryerson C. C., “Icing management for Coast Guard assets” U.S. Army Engineer Research and Development Center. Cold Regions Research and Engineering Laboratory. ERDC/CRREL TR-13-7 (2013). https://apps.dtic.mil/sti/tr/pdf/ADA583017.pdf. Accessed 10 September 2023.
  • 62.Guest P., Vessel Icing (2005). https://www.vos.noaa.gov/MWL/dec_05/ves.shtml. Accessed 10 September 2023.
  • 63.Mintu S. A., “Theoretical and numerical methods for predicting ship-wave impact generated sea spray,” PhD thesis, Memorial University of Newfoundland, St. John’s, Newfoundland, (2021). [Google Scholar]
  • 64.Samuelsen E. M., Ship-icing prediction methods applied in operational weather forecasting. Q. J. Royal Meteor. Soc. 144, 13–33 (2018). [Google Scholar]
  • 65.International Maritime Organization, International code for ships operating in polar waters (polar code). Resolution MEPC 264 (2017). https://www.imo.org/en/ourwork/safety/pages/polar-code.aspx. Accessed 22 September 2023. [Google Scholar]
  • 66.The WaveWatchIII Development Group (WW3DG), User manual and system documentation of WAVEWATCH III. (Version 6.07, NOAA Environmental Modeling Center, 2019).
  • 67.Flanders Marine Institute, IHO Sea Areas (Version 3, 2018). Deposited 10 December 2018.
  • 68.Tebaldi C., Arblaster J. M., Knutti R., Mapping model agreement on future climate projections: MAPPING MODEL AGREEMENT. Geophys. Res. Lett. 38, L23701 (2011). [Google Scholar]
  • 69.Blanchard-Wrigglesworth E., Webster M., Boisvert L., Parker C., Horvat C., Record Arctic Cyclone of January 2022: Characteristics, impacts, and predictability. JGR Atmos. 127, e2022JD037161 (2022). [Google Scholar]
  • 70.Watts M., Maslowski W., Lee Y. J., Kinney J. C., Osinski R., A spatial evaluation of Arctic Sea ice and regional limitations in CMIP6 historical simulations. J. Clim. 34, 6399–6420 (2021). [Google Scholar]
  • 71.Mioduszewski J., Vavrus S., Wang M., Diminishing Arctic Sea ice promotes stronger surface winds. J. Clim. 31, 8101–8119 (2018). [Google Scholar]
  • 72.Rinke A., et al. , Extreme cyclone events in the Arctic: Wintertime variability and trends. Environ. Res. Lett. 12, 094006 (2017). [Google Scholar]
  • 73.Wang X. L., Casas-Prat M., Feng Y., Crosby A., Swail V. R., Historical changes in the Davis Strait Baffin Bay surface winds and waves, 1979–2016. J. Clim. 34, 8879–8896 (2021), 10.1175/JCLI-D-21-0054.1. [DOI] [Google Scholar]
  • 74.Walsh J. E., et al. , Extreme weather and climate events in northern areas: A review. Earth-Sci. Rev. 209, 103324 (2020). [Google Scholar]
  • 75.Notz D., Community S., Arctic sea ice in CMIP6. Geophys. Res. Lett. 47, e2019GL086749 (2020). [Google Scholar]
  • 76.Henke M., Arctic Ice-Wave Hazards Analysis. Zenodo. 10.5281/zenodo.12151436. Deposited 18 June 2024. [DOI]
  • 77.Voldoire A., CNRM-CERFACS CNRM-CM6-1-HR model output prepared for CMIP6 ScenarioMIP ssp585 ssp585. Earth System Grid Federation. 10.22033/ESGF/CMIP6.422. Accessed 8 August 2023. [DOI]
  • 78.Steger C., et al. , DWD MPI-ESM1.2-HR model output prepared for CMIP6 ScenarioMIP ssp585 ssp585. Earth System Grid Federation. 10.22033/ESGF/CMIP6.4479. Accessed 8 September 2023. [DOI]
  • 79.EC-Earth Consortium (EC-Earth), EC-Earth-Consortium EC-Earth3 model output prepared for CMIP6 ScenarioMIP ssp585 ssp585. Earth System Grid Federation. 10.22033/ESGF/CMIP6.4912. Accessed 9 May 2023. [DOI]
  • 80.Yukimoto S., et al. , MRI MRI-ESM2.0 model output prepared for CMIP6 ScenarioMIP ssp585 ssp585. Earth System Grid Federation. 10.22033/ESGF/CMIP6.6929. Accessed 9 May 2023. [DOI]
  • 81.Hersbach H., et al. , The ERA5 global reanalysis. Q. J. R. Meteorol. Soc. 1–51 (2020).
  • 82.Piollé J.-F., Dodet G., Quilfen Y., ESA Sea State Climate Change Initiative (SeaStatecci): Global remote sensing daily merged multi-mission along-track significant wave height from altimetry, L3 product: version 3, NERC EDS Centre for Environmental Data Analysis. 10.5285/e6af67fca81c40b7bb3eddaadde06909. Accessed 2 October 2023. [DOI]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix 01 (PDF)

pnas.2400355121.sapp.pdf (2.3MB, pdf)

Data Availability Statement

Python codes used to process model outputs and generate figures in this article are available at https://doi.org/10.5281/zenodo.12151436 (76). The climate model data used within this study is available through the World Climate Research Programme at https://esgf-node.llnl.gov/projects/cmip6/ (77–80). ERA5 surface winds and sea ice concentration used in validation are available through the climate data store at https://cds.climate.copernicus.eu/ (81). Satellite validation utilized the ESA Remote Sensing Significant Wave Height L3 multisatellite product available at https://archive.ceda.ac.uk/ (82). The simulated wave data generated from the wave model is available at https://doi.org/10.18739/A2ST7DZ74.

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES