Non-Gaussian Cold Side Temperature Distribution Tails: Meteorological Mechanisms and Impact on Extreme Cold Exceedances under Warming

Changes in exceedances of fixed extreme temperature thresholds under warming are manifested in more complex ways when the tails of the underlying distribution are non-Gaussian compared to if the distribution were normal. Here we investigate the special case of non-Gaussian cold side tails. Under the simplest prototype of a uniform warm shift across the temperature distribution, a location with a short cold side tail would see a more rapid decrease in extreme cold threshold exceedances than if the tail was Gaussian or longer. Similarly, a long cold tail would result in highly anomalous cold extremes to decrease in frequency more slowly than for a Gaussian. Short and long cold tails are shown to occur in spatially coherent regions globally, with northern hemisphere winter cold tails showing the strongest departures from Gaussianity. Climate models generally reproduce these short tails reasonably well and mostly project a more or less rapid than Gaussian decrease in extreme cold temperature exceedances by mid-century for regions of short and long tails respectively. This implies that some regions, such as the northeastern portions of North America and Eurasia, will experience a rapid decrease in extreme cold in the near future. Meanwhile, other regions, such as western North America may continue to experience rare extreme cold events due to the underlying longer-than-Gaussian tails. We further investigate the meteorological mechanisms responsible for the presence of non-Gaussian tails. Low-level flow patterns over a spatially non-uniform horizontal temperature gradient help explain the occurrence of departures from Gaussianity at the tails in many cases. Results carry implications for many impacts that would occur in the absence of exceeding certain cold thresholds, such as pest infestations. In other cases, a reduction in extreme cold due to short tails could make certain regions more favorable for human and economic activity.