Geographic Distribution of Trends and Cycles in Eastern North American Instrumental Climate Data

Abstract

The potential environmental impact of a changing climate has been an increasingly prominent focal point of scientific and popular literature. Much of this literature has focused on global or continental trends, despite there being differences in long-term trends and medium-term variability on smaller geographic scales. This dissertation considers historical temperature and precipitation records in eastern North America, analyzing long-term changes and the influence of oscillatory climatic drivers.Three broad themes are addressed in this study, with a common methodological core. An assessment of extreme weather trends throughout the eastern North America study region is presented first. Next, two case studies are done to more thoroughly examine localized phenomena, specifically, an examination of historical climate variability in Ottawa, Ontario, Canada, and an examination of climatic drivers affecting lake ice phenology in southwest New Brunswick and adjacent eastern Maine.The methodological basis for this study is a combination of time-series analysis techniques, applied to time-dependent instrumental or proxy data. The results indicate that regional extreme weather time series, as well as local weather or weather-related time series, are best characterized in terms of long-term periodic trends due to known climatic drivers.The most significant of these drivers is the 11-year Schwabe Solar Cycle (SSC), which affects most if not all of the weather time series studied. Otherwise, cyclic patterns can be divided into interannual, interdecadal, and multidecadal scales. The most influential drivers of interannual climate patterns were the Quasi-Biennial Oscillation and the El Niño Southern ii Oscillation. At the interdecadal level, in addition to the SSC, the Pacific Decadal Oscillation was a common driver of extreme and non-extreme weather phenomena in the region.The combination of a larger-scale study with two case studies made it possible to discern spatial patterns in regional weather. For extreme weather, the region of interest can be subdivided into three climatically distinct subregions by applying hierarchical clustering to extreme weather counts. Subregional spatial coherence in eastern North America was further demonstrated by localized coherence between seasonal temperature records in eastern Ontario and southwestern Quebec, and similar localized coherence in ice phenology in southern New Brunswick and eastern Maine.