Interdecadal variability of streamflow driven by atmospheric circulation in the Hudson Bay lowlands watersheds

Abstract:

The Hudson Bay, located in subarctic Canada, is one of the major collectors of the freshwater entering the Arctic Ocean. The rivers in the Hudson Bay lowlands region are also a source of major floods, threatening the local communities. In the recent decades, this region is affected by major hydroclimatic changes mostly attributed to climate change. Previous studies have also highlighted high interannual and interdecadal variability due to internal variability of climate. A period of low streamflow in the 1980s and 1990s have been associated to the Arctic oscillation (AO) and an increase in cold and dry spells into the region. The period 1990s-2000s have seen an increase in streamflow due to the advection of wetter air masses and the degradation of permafrost associated to the warming. The main driver of the streamflow change is still unclear, and the most recent years have not been yet investigated. In this study we use ERA5 reanalyses, hydrometric observations, and the hydrological model MESH set up in two watersheds to investigate the impact of Atmospheric circulation on the inter-decadal variability of streamflow between 1979 and 2018 in the Hudson Bay lowlands. The daily see level pressure (SLP) from ERA5 reanalyses have been discretized in eight weather regimes using the k-means algorithm applied to a large region of North America. The results show a large occurrence of low-pressure regimes in the 1996-2010 period that was associated with larger precipitation amounts and high streamflow in the west part of the region. In the eastern watersheds, precipitation and streamflow have been less affected by this shift in atmospheric circulation. In the most recent period, despite lower precipitation amounts in the western watersheds, streamflow remains high. This recent trend is not properly simulated by MESH, suggesting a non-climatic driver such as the degradation of permafrost. This study improves our understanding in the inter-decadal variability of streamflow in this sensitive area of the Arctic and will help to reduce the uncertainties in the future projections of streamflow.