Abstract/Summary

Satellite observations have revealed that the Arctic is undergoing rapid climate change. Climate model projections unanimously simulate that year-round reductions in Arctic sea ice will continue through the 21st century. The primary goal of this thesis is to investigate the implications of these changes for trans-Arctic shipping. Arctic routes offer a substantial distance saving over conventional routes, with potentially significant global economic implications. Shipping in Arctic waters is a hazardous endeavour and increases in shipping traffic heighten the need for robust projections of future shipping accessibility to assess the risks involved. However, all global climate model (GCM) simulations contain intrinsic biases in their simulation of sea ice. This thesis has produced a calibration technique to constrain and reduce these biases. Applying this approach to a suite of state of the art GCMs reveals that the Arctic may become “ice-free” in the 2050s, a decade earlier than without the calibration technique. Projections of Arctic shipping are also made using data from these calibrated climate models, likely adding to their robustness. Using the calibrated multimodel ensemble reveals that, by mid-century, Arctic transit potential doubles for standard ‘open water’ vessels; most years become navigable for some period, irrespective of future emissions scenario, with the currently inaccessible Trans-polar Sea Route across the central Arctic becoming accessible for the first time. European routes to East Asia become 10 days faster on average than alternatives by mid-century, and 13 days faster by late-century, while North American routes become 4 days faster. Future greenhouse-gas emissions play a significant role by late-century; the shipping season reaching 8 months in RCP8.5, double that of RCP2.6 which exhibits substantial interannual variability. Moderately ice-strengthened vessels would enable fast and reliable trans-Arctic shipping, essentially year round, from mid-century. Climate model projections reveal that sea ice will be present throughout the 21st century winter, regardless of future greenhouse gas emissions. This implies that Arctic sea routes will continue to open and close annually. This, combined with increased shipping in the region, highlights the need for improved seasonal predictions of conditions on the Arctic sea routes. The upper lead-time limit of predictability of the opening of Arctic sea routes is explored using ensembles of simulations in a ‘perfect model’ approach. Initial results indicate the skill of forecasts drops dramatically before May, indicating the presence of a predictability barrier.