Statistical Projections of Ocean Climate Indices off Newfoundland and Labrador

Abstract

ABSTRACTPresent global climate models (GCMs) are unable to provide reliable projections of physical oceanographic properties on the continental shelf off Newfoundland and Labrador. Here we first establish linear statistical relationships between oceanographic properties and coastal air temperature based on historical observations. We then use these relationships to project future states of oceanographic conditions under different emission scenarios, based on projected coastal air temperatures from global (Canadian Earth System Model, version 2 (CanESM2), Geophysical Fluid Dynamics Laboratory's Earth System Model, version 2M (GFDL-ESM2M)) and regional (Canadian Regional Climate Model (CRCM)) climate models. Estimates based on CanESM2 agree reasonably well with observed trends, but the trends based on two other models result in substantial underestimates. Projected trends are closer to observations under a high emission scenario than under median-level emission scenarios. Over the next 50 years, the increases in projected sea surface temperature off eastern Newfoundland (Station 27) range from 0.4° to 2.2°C. The increases in bottom ocean temperature over the Newfoundland and Labrador Shelves range from 0.4° to 2.1°C. The area of the cold intermediate layer (<0°C) on the Flemish Cap (47°N) section is projected to decrease by 9–35% of the 1981–2010 average. The decline in sea-ice extent off Newfoundland and Labrador ranges from 20 to 77% of the average (0.4–1.5 × 105 km2), and the reduction in the number of icebergs at 48°N off Newfoundland ranges from 30% to nearly 100% of the norm at this latitude. Despite differences among the models and scenarios, statistical projections indicate that conditions in this region will reach or exceed their maxima (sea surface temperature, bottom ocean temperature) and reach or fall below their minima (sea-ice extent, number of icebergs) that were observed during the course of monitoring activities over the past 30–60 years, possibly as early as 2040. We note, however, that the statistical relationships based on historical data may not hold in the future because of the changing influence of input from Arctic waters and because of large uncertainties in projected air temperatures from GCMs.