An Evaluation Of Seasonality Through Four Delineation Methods PDF Download

Structured into three, interrelated projects, this dissertation is organized under the research themes of climate change and human mortality responses associated with heat and cold events. Using diverse, methodologies which incorporate spatiotemporal variability into seasonal definitions, consistent results were found pertaining to seasonal climate change. Climatological data for 60 U.S. locations was chosen for a lengthy period of record (1948-2012), and despite the various delineation methods, results generally indicate earlier starts to summer while winters have become shorter. Using percentile thresholds of daily mean apparent temperature, heat, extreme heat, cold, and extreme cold events (anomalous temperature events or ATEs) were defined and analyzed. In addition to increased frequency and duration, heat and extreme heat events have started earlier over the course of the year since 1948. Conversely, cold and extreme cold have shown the opposite with less frequent, shorter, and later occurring events taking place. While not directly explored, a relationship with ATEs and large-scale teleconnection patterns such as El Niño seems to exist. In bioclimatological research, human health responses have been associated with duration, intensity, and seasonal timing. When comparing mortality responses, a relatively unexplored method in climatology was introduced. More common in epidemiological research, a distributed lag non-linear model (DLNM) assessed the delayed, non-linear responses associated with environmental outcomes. In regards to both heat and cold events, greater risk was associated with early season events compared to later occurring ones. Elevated thresholds of extreme heat and extreme cold showed higher risks when compared to less extreme events. Geographic variability was also shown with higher risks for heat-related mortality in more northern locations while cold-related risks were higher in locations less accustomed to heat. Comparing mortality responses following ATEs, the consistent relationships were found, irrespective of seasonal and ATE definitions. Understanding the relationship between human health outcomes, changing seasons, and anomalous temperature events will remain an emphasis of future climate-health research. With projected changes to the climate system, future research may further examine differential responses which may be dependent on age, sex, cause of death, and race.