Arctic open-water periods are projected to lengthen dramatically by 2100

The dramatic decline in Arctic sea ice extent since the late 20th century serves as a stark indicator of anthropogenic climate change. While the pan-Arctic September sea ice extent provides a valuable long-term climate indicator, regional variations are substantial. For stakeholders in the Arctic, the length of the seasonal open-water period is arguably more relevant, impacting various aspects of life and industry. This includes phytoplankton productivity, coastal erosion, hunting and fishing practices, marine shipping, and tourism. The timing of sea ice retreat and advance is also critical; delayed ice advance, for example, intensifies the effects of ocean swell, particularly during the intense Arctic storm season (November to February). Since 1979, nearly every region of the Arctic Ocean has experienced an increase in the open-water period, attributed to both earlier retreat and later advance. However, the Pacific-side Arctic shows a more dominant influence of later advance, linked to increased summer ocean heat uptake due to earlier open water formation. Previous studies using CMIP5 models primarily under a high-emissions scenario (RCP8.5) projected significant open-water period lengthening. However, the impacts of lower warming levels, crucial given the Paris Agreement's goals, remained largely unaddressed. This study uses CMIP6 model output under low, medium, and high emission scenarios (SSP126, SSP245, and SSP585) to project open-water periods for 15 Arctic regions and key shipping routes (Northern Sea Route and Transpolar Sea Route). The projections are assessed in terms of both time and global temperature anomalies, providing regionally specific predictions crucial for stakeholders.

Numerous studies have documented the rapid decline of Arctic sea ice extent and thickness, with projections indicating frequent ice-free Septembers with 2°C of warming above pre-industrial levels. The focus on pan-Arctic September sea ice extent, while important, overlooks significant regional variability. Several studies have analyzed future projections of open-water periods using CMIP5 data, but mostly focusing on high-emission scenarios. These simulations highlighted the dominance of later ice advance in lengthening the pan-Arctic open-water period. However, assessments of lower warming levels, such as those aligning with the Paris Agreement's goals of limiting warming to well below 2°C, were lacking. This necessitates a more comprehensive assessment using a newer generation of climate models and scenarios.

This study utilizes sea ice concentration (SIC) data from CMIP6 models (historical, SSP126, SSP245, and SSP585) and compares these projections with observations from multiple satellite datasets (Bootstrap, NASA-Team, and OSI SAF). Monthly surface air temperature data from four observational sources (Berkeley Earth, GISTemp v4, HadCRUT4.6.0.0, and NOAAGlobalTemp v5) are also employed. Sea ice retreat day is defined as the last day SIC falls below 15% before reaching the annual minimum, while advance day is the first day SIC rises above 15% after the annual minimum. The period between these days constitutes the open-water period. A 5-day moving average is applied to daily SIC to mitigate short-term fluctuations. The sea ice year is dynamically defined based on the median of days with maximum SIC in January-April. Model bias assessment follows the SIMIP Community methodology; multi-model means are computed from the first simulation of each model, and internal variability is determined by the average standard deviation across model ensembles. Observational uncertainty is calculated as the range across different datasets. The study includes 15 Arctic regions and the Northern Sea Route and Transpolar Sea Route. Analyses compare historical model simulations against observational data (1979-2013). Future projections assess open-water period length, sensitivity to temperature anomalies, percentage of regional area with retreat/advance before/after specific dates for 15% and 80% SIC thresholds. Analysis of trends and sensitivity to temperature anomalies are then carried out for the open-water period, retreat and advance days.

CMIP6 models show a generally good match with observations for the pan-Arctic average open-water period, although regional biases exist. For example, models overestimate the open-water period in Hudson Bay and underestimate it in the Barents Sea. The temperature sensitivity of the open-water period is generally higher in observations than in the multi-model mean, but this difference can often be attributed to internal variability. However, for temperature sensitivity specifically, there are significantly more cases of models falling below the uncertainty range than above, suggesting that the open-water period in some CMIP6 models may not be sensitive enough to warming. Projections reveal a steady increase in the open-water period across all emission scenarios until the 2040s, after which the rate of increase decreases in low-emission scenarios (SSP126), remains consistent in medium scenarios (SSP245), and accelerates in high-emission scenarios (SSP585). Under SSP585, several regions become ice-free year-round by 2100. Significant changes are projected for the Kara, Laptev, East Siberian, Chukchi, and Beaufort Seas, with open-water periods extending dramatically under SSP585. The central Arctic Ocean, historically characterized by perennial ice cover, is also projected to experience substantial changes, with open-water periods reaching 3 months (SSP126) or nearly 8 months (SSP585) by 2100. With 2°C warming, the open-water period increases on average by 2 months (63 days), with the most substantial changes occurring in the Barents, Chukchi, Kara, and East Siberian Seas. The increase stems from both earlier sea ice retreat and later advance, with the change in advance exceeding that of retreat, consistent with the ice-albedo feedback. Maps showing the year or temperature threshold at which the open-water period exceeds 90, 180, or 270 days highlight the increased accessibility of shipping routes with continued warming. The Northern Sea Route choke points open for over 90 days with 3°C warming and almost the entire Arctic Ocean (including the Transpolar Sea Route) shows a 90-day open-water period with 3.5°C warming. The timing of this consistent exceedance depends heavily on the emissions scenario.

The findings of this study provide critical updates to projections of Arctic open-water periods, leveraging the CMIP6 model data. The study highlights both the overall trend towards longer open-water periods and the significant regional variability. CMIP6 models generally reproduce key characteristics of the historical open-water period and its response to warming, including the disproportionate influence of later ice advance on the lengthening of the open-water period. However, some biases remain, particularly in temperature sensitivity, suggesting that model projections might be conservative. The projected changes have substantial implications for various stakeholders and ecosystems. While the increased accessibility of shipping routes presents potential economic benefits, it must be considered in the context of potential negative consequences, including increased coastal erosion and disruptions to traditional hunting and fishing practices. The findings emphasize the need for adaptation strategies to mitigate the impacts of these dramatic changes, particularly in regions such as Hudson Bay, the Chukchi Sea, and the Beaufort Sea, which are projected to undergo significant transformations even with a 2°C warming.

This study utilizes CMIP6 models to refine projections of Arctic open-water periods, incorporating regional variations and sensitivity to global temperature anomalies. The results reveal a dramatic lengthening of open-water periods across most of the Arctic by 2100, particularly under high-emission scenarios. While the pan-Arctic average shows good agreement with observations, regional biases exist, particularly concerning the temperature sensitivity of sea ice advance. The findings highlight the significant implications for shipping, ecosystems, and Indigenous communities, emphasizing the need for adaptation and mitigation strategies. Future research could focus on improving the representation of regional processes in climate models, examining specific ecosystem impacts, and developing more precise regional forecasts for stakeholders.

The study's reliance on CMIP6 models introduces inherent uncertainties associated with model limitations and internal variability. Regional biases, particularly in the temperature sensitivity of sea ice advance, suggest the possibility of conservative estimates. Furthermore, the study primarily focuses on the open-water period based on a 15% SIC threshold; other thresholds may reveal different patterns. Lastly, the assessment of impacts on ecosystems and human activities is limited to a general overview; more detailed studies focusing on specific regions and sectors are needed.