Analyzing and Predicting Sea Ice Conditions in the Northwest Passage

The Northwest Passage (NWP), a sea route that runs through the Canadian Archipelago, connects Baffin Bay in the East to the Beaufort Sea in the West. A common misconception of the NWP is that it is one route, but it is a conglomerate of many sub-routes. Many of these sub-routes are filled with sea ice year-round, and until 2007, ships not rated for breaking sea ice were unable to traverse the passage without the assistance of an icebreaker. In 2007, satellite observations found that there was a sequence of sub-routes that could have been utilized for ships of lower ice-classes to make it through the NWP due to anomalously low sea ice concentration (SIC) in those areas. Again in 2008, the NWP had so little sea ice that a non-icebreaking ship could make it through the NWP. The Crystal Serenity, a cruise ship carrying roughly 1,200 passengers traversed the NWP from West to East in the year 2016, making it one of the more famous voyages by a vessel in the NWP. Conducting an analysis of the environmental conditions that allow for the NWP to be "open" (passable by ships without icebreaker aid) or "closed" reveals that a combination of thermodynamic and dynamic factors lead to the NWP being passable. To observe the NWP, satellite observations and reanalysis are combined to measure surface temperature, ice concentration, freezing degree days (FDD), geopotential height fields, and anomalous wind patterns. By measuring the thermodynamic and dynamic properties that lead to the open and closed phases of the NWP similar patterns begin to emerge. Anomalously warm surface temperatures coupled with above normal geopotential height fields show a thermodynamic cause for the opening of the NWP. Anomalous winds assist in the transport of sea ice. When these winds push sea ice orthogonal, or out of the NWP, there is a greater chance that the NWP will be open. With a continued loss of sea ice more than likely in the region, the NWP is likely to continue to open and to be open for longer in the late Summer. If this is indeed the case, the NWP may become a viable shipping lane for tourism and trade in the near future, making viable sea ice prediction an increasing necessity for safe passage. Using environmental reanalysis combined with satellite remote sensing data, machine learning models are created for different grids within the NWP's waters to predict SICs in advance. By training these models on historical data, loss is minimized leading to more accurate predictions.