Daily temperature variability patterns during ENSO and NAO events in United States winters

For decades climatologists have researched teleconnection patterns to understand the impacts of these phenomena on seasonal climate variability in the United States. This research expands upon many previous works that have primarily addressed climate variability in terms of average departures from monthly and seasonal mean temperature and precipitation conditions. Here, two prominent teleconnections with strong winter signatures in the United States, the El Niño/ Southern Oscillation (ENSO) and North Atlantic Oscillation (NAO), are examined to thoroughly describe intra-seasonal climate variability during each teleconnection phase over the last 55 winters. January – March daily maximum and mimimum temperature variability is assessed with Varimax rotated principal components analysis for each month and teleconnection phase, including neutral events. The spatial variability of results obtained across the country are assessed in GIS and the magnitude of phase variability is examined. Prior to statistical analyses, the global climate change signal is removed from temperature records since the mid 1970s using a hinge-fit linear trend line technique. In addition, an apparent trend in the NAO index, that biases the record prior to 1979 toward negative values and the period following 1979 toward positive values, is also removed to create a new NAO index for use in long-term climate variability assessments.

Significant intra-seasonal climate variability is identified in all ENSO and NAO phases. Results indicate that two dominant modes and one subordinate mode exhibit a pattern of large and regionally consistent daily temperature variability in all phases, frequently over the Southwest, Northeast and North Central United States. In most ENSO and NAO events, the standard deviation of phase variability is approximately 6– 8°C. Variability patterns of Neutral ENSO and NAO phases are often different each month, indicating that these events are not simply times of “near normal” conditions and are associated with unique regional temperature variability. Further, since the spatial patterns of daily maximum temperature variability in a given phase are frequently found to be dissimilar to that of minimum temperature in the same phase, the use of average temperatures may be misrepresentative of the true range of temperature variation in a given region. The results of this research add information to climate forecasts and should aid confidence statistics issued alongside regional climate forecasts of phase average temperature departures.