Spatial and Temporal Variability of Extreme Weather in The United States

Abstract

As the global climate is influenced by increased warming, the frequency of climate extremes will likely become more variable. The United States Historical Climate Network (USHCN) station data from 1900-2011 is used to quantify trends in daily extreme heat events, daily extreme cold events, and extreme daily precipitation within the contiguous United States. Climate data was spatially aggregated into respective Koeppen-Geiger climate zones where the 3 main zones are; arid, warm temperate and snow. Results show a gain or loss of 20 extreme temperature events and a gain or loss of 4 extreme precipitation events. The arid zone exhibited a loss of extreme minimum temperature a gain of extreme maximum temperature and a mixed result for extreme precipitation. The warm temperate zone indicated that the eastern region follows the exact same temperature trends as the arid zone but exhibits an overall loss of extreme precipitation. The western portion of the warm temperate zone exhibits a loss in extreme temperature events, a gain in extreme minimum temperature events and a gain in extreme precipitation. The trends in the snow zone reveal a mixed signal for extreme maximum temperature, a decline in extreme minimum temperature and a gain in extreme precipitation. Paired t-test results indicate statistically significant shifts in the magnitude of extreme weather events in each of the 3 main climate zones (arid, warm temperate and snow). This spatiotemporal analysis highlights how daily trends in extreme heat, extreme cold, and extreme precipitation have changed in the last 100 years in the context of specific climate zones. In order to understand the impact regional climate has on extreme events, trends in teleconnection patterns were examined in conjunction to daily weather. Teleconnection patterns that directly impact US weather are El Niño Southern Oscillation (ENSO) and Pacific Decadal Oscillation (PDO). The use of wavelet analysis, including the continuous wavelet transform, the wavelet cross wavelet transform and the wavelet transform coherence provide insight into the timescales of influence from ENSO and PDO on extreme weather events. The power spectra results from each wavelet analysis have been averaged across Koeppen-Geiger zone. Results indicate that extreme precipitation events have significantly different power spectras than normal events across all timescales. Specific patterns at the annual scale and shorter are found the arid zone for extreme maximum temperature, where results for minimum temperature trends vary.