Arctic shipping is better, even accounting for sea fog.

5 Re-Designing Shipping Routes by Minimizing the Impacts of Sea Fog

Can we reduce the delays by detouring from some sea fog regions? We re-optimize the Route-I using a cost function that considers the change of the cruising speeds under different sea-fog and sea-ice conditions. The cost function is minimized with respect to the sailing time. Thus, the new route, referred to as Route-F, has the shortest sailing time (GEF-UNDP-IMO GloMEEP Project and members of the GIA, 2020) (Text S4 and Figure S7 in Supporting Information S1). Based on the CMIP5 multi-model ensemble means of sea ice and the SFF, we compute the path and the sailing time of the Route-F every 10 years in 21st century under both RCPs.

Along the NWP, the Route-F can gain back 0.3–1 days of sailing time compared to the Route-I (Figure 4b), albeit it is longer. The Route-F through the Canadian Arctic Archipelago is similar to the Route-I because of the island locations; the major detours are the northward shifts of the Route-F relative to the Route-I along the Alaska coastline to avoid high-SFF regions (Figure 3). The shift along the Alaska coast segment (6,000–7,000 km) becomes gradually obvious due to the retreat of sea ice over the Beaufort Sea before the midcentury. After the midcentury, the impacts of the SFF becomes dominant. By the end of this century, the SFF over this segment will decrease by 10%–13% (Figure 5d). Although the length of the Route-F is longer than the Route-I, the decrease in sea fog along the Route-F helps for saving the sailing time up to 1 day. In RCP4.5, with the high-SFF region along the NWP, the routes between St. John's and Bering Strait switch the transit via the NSR during 2026–2035 and 2056–2065 (Figure S8 in Supporting Information S1). Although the routes are about 1,000 km longer than the NWP through the Canadian Arctic Archipelago, the high-SFF region is reduced by 1,260 km (Figure S9 in Supporting Information S1).

In contrast, along the NSR, the Route-F can only gain back 0.1–0.4 days of sailing time compared to the Route-I (Figure 4). The Route-F tends to avoid high-SFF region along the Eurasian coast to navigate at higher latitudes during the first half of the century. After the midcentury, the Route-F tends to avoid the northeast coast of Greenland and shifts eastward to the Fram Strait. The route no longer passes through the high-SFF region by the end of this century. The detour of Route-F is similar under RCP4.5 (Figures S8 and S9 in Supporting Information S1).

We further compare the Route-F with the traditional routes such as the Suez Canal and the Panama Canal (Table S5 in Supporting Information S1). These routes start from Shanghai and terminate at St. John's and Rotterdam. The result shows that the Route-F of NSR or NWP can still save 30%–33% of the distance and 20%–30% of the time of traditional routes on average in the 21st century, suggesting great economic benefits of the Route-F.

6 Conclusions

Previously designed shipping routes based on the shortest distance are not necessarily the safest routes because the presence of sea fog along the routes has not been considered. For safety, the captain may reduce the cruising speed when sailing in sea fog. By parameterizing the captain's decision as a reduction of the cruising speed in sea fog, we found that the sailing time of the shortest routes should have been 5%–25%, or 1–4 days, more than previously estimated. Our finding is particularly important along newly formed sea-ice margins, not only because a safe voyage is more difficult along the margins, but also because sea fog here is projected to be increasing under global warming. Our result highlights the importance of Arctic sea fog when designing the trans-Arctic shipping routes in the future when Arctic sea ice continues its decline but still covers the NSR and the NWP to some extent.

The re-designed trans-Arctic shipping routes based on both sea fog and sea ice conditions over a decade window are helpful for strategic decision-making for ship operations and capacity building. Although our analysis is based on monthly data sets, and the results may not be applied directly on the operational decision-making, applying the method suggested in this analysis to real-time data can provide short-term navigation services for ship owners.