Recent studies have identified a number of changes in the hydroclimatology of north-east USA over the last three or four decades. Many of these are concerned with reduced snow and ice storage during winter and earlier release of meltwater into the regional river network in spring, thereby altering the timing of the annual spring meltwater flood event and distribution of streamflow through the season. A number of investigators have proposed that the coincidental tendency of the North Atlantic Oscillation (NAO) towards a more positive state is linked to these changes in meltwater generation, but little process-based investigation of this hypothesised link has been undertaken. A more complete understanding of the nature of the chain of causality linking climate and streamflow will both aid explanation of the causes of these hydroclimatic variations and help establish the extent to which the coincidental trend in the NAO is linked to New England streamflow. Such investigation will also be necessary to assess how predicted climatic change may impact water resources in this region, and in the development of predictive relationships for forecasting of high and low flow periods.
This paper addresses this research gap by identifying the climatic processes controlling high and low monthly streamflow across New England, and thus whether conditions associated with high and low streamflow can be linked with variation in the NAO. This is achieved through a composite analysis of a range of climatic variables under high and low streamflow conditions between 1958-2001, using data from the ERA-40 reanalysis and the USGS Hydro-Climatic Data Network. It is thought that such an approach will yield greater information on large-scale climatic forcing of streamflow compared to a correlation-based study. Statistically significant and physically consistent differences are found between high and low streamflow and a number of climatic variables. Notably, surface temperature and atmospheric thickness are shown to be positively associated with streamflow (mainly in winter), as are geopotential height and humidity. Wind speed is negatively associated with streamflow. The differences are consistent with those expected from NAO forcing of climate over the study region. Study results are summarised in the form of a conceptual model that may lay the basis for future investigations of the cascade of processes linking large-scale climate and streamflow.