U.S. Cold Snap in Context
The recent cold snap may be the most severe the eastern United States has experienced in more than 30 years. Do a few weeks of cold temperatures in one region of the world mean that global warming has stopped?
There are a few variables to consider – the behavior of the Arctic Oscillation, the ratio of record highs to record lows and long-term trends.
The Arctic Oscillation
While the eastern United States and central Asia have been in a cold snap, the Arctic has been warmer than normal. This is not coincidental. Weather is controlled by a variety of cycles – daily cycles, seasonal cycles, 11-year solar cycles and even several thousand year cycles that affect Earth’s orbit. One cycle – the Arctic Oscillation – reflects shifts in the difference in atmospheric pressure between the Arctic and the mid-latitude regions of the Northern Hemisphere. This pressure difference affects the strength of the upper-atmospheric westerly winds that drive much of the Northern Hemisphere’s weather, especially during winter. When the westerly winds are strong, it is more difficult for frigid air at the poles to “invade” the mid-latitudes, including the eastern United States. The frigid air is trapped at the pole, keeping the Arctic colder than normal and the mid-latitudes warmer than normal. When the westerly winds are weak, it is easier for Arctic air to invade the mid-latitudes. This leaves the North Pole less frigid than normal, but sends the mid-latitudes into a cold snap like the one experienced over the past few weeks. The state of the Arctic Oscillation is a good predictor of how extreme winter temperatures will be. In Chicago, for example, there are on average three times as many days each year when the temperature drops below zero degrees Fahrenheit during negative phases (weak westerly winds), versus positive phases (strong westerly winds). See the map of recent temperature anomaly distributions below, which is typical of negative phases of the Arctic Oscillation.
BOTTOM LINE: The recent cold snap reflects a change in the distribution of the heat Earth holds – there is more cold air in the mid-latitudes and less cold air in the Arctic. There is no evidence to indicate there is either more or less heat in the system than there was before the cold snap occurred.
Record Highs vs. Record Lows
Even during periods when average temperatures rise, variables may come together to cause a given locality to experience a record low temperature for a given day in the calendar year. The same could be said for occurrences of record high temperatures during periods of cooling. During the warmest decade on record (the last ten years), lots of record lows were set – 142,420 record lows between Jan. 1, 2000 and Sept. 30, 2009. There are close to 5,000 quality-controlled weather stations across the United States providing daily temperature data to the National Climatic Data Center. Yet, it is not the existence of record highs or record lows that indicates whether a warming or cooling trend is occurring. Instead, it is the proportion of record highs to record lows that tells you whether things are getting warmer or cooler. All other things being equal – meaning that there is no increase or decrease in average surface temperature – the ratio of record highs to record lows should be around 1:1. But, from January 2000 to September 2009, 291,237 record high temperatures were set, giving a high to low ratio of more than 2:1. The disparity between record highs and record lows reflects the above normal temperatures experienced over the last decade. It will take many more winter cold snaps like this to even out the record high to low ratio.
BOTTOM LINE: A given locality will experience some record warm temperatures during periods of cooling and some record cold temperatures during periods of warming. But, one record event doesn’t create a trend. If you want to know whether things are getting warmer or cooler over the years, you have to look at the ratio of record warm events to record cold events. In the United States over this last decade, the ratio has been around 2:1.
More Cold Snaps to Come?
So, should we be expecting more of these cold snaps in the future? Analyzing past trends might help to answer this question. Patterns of periodic warming and cooling over the North Atlantic in the past – linked to periodic strengthening and weakening of the circulation that brings warm waters into the Atlantic basin from the south – suggest that the Atlantic may cool slightly over the next decade. As this happens, average surface temperatures in North America and Europe may stop their rising temperature trends or even cool slightly. Looking at long-term data (50+ years), which includes periods of both warm and cool North Atlantic temperatures as well as warm and cool periods of other major natural oscillations that help drive our weather, suggests that the extreme cold experienced over the past few weeks is becoming less common for the United States as a whole. Adding up all the cold winter days (when temperatures fall in the lowest ten percent of daily mean surface temperature distribution) and warm winter days (when temperatures fall in the top 10 percent of daily mean surface temperature distribution) over the past 50 years shows that despite significant regional differences, there has been an overall decreasing trend in the number of cold winter days and an increasing trend in the number of warm winter days. In the West, there has been a downward trend in the number of cold winter days, while most of the rest of the country has shown no trend. On the other hand, the West, along with the northern Great Plains and upper Midwest, has witnessed a rise in the number of warm winter days. The only region where warm winter days have become less common is the southeast, particularly the Gulf Coast area. South Florida has experienced a decline in the number of cold days and a rise in the number of warm days.
BOTTOM LINE: Despite significant regional differences, 50-plus year trends in U.S. temperatures show fewer cold and more warm winter days. These trends account for both warm and cold periods of the major natural oscillations that determine year-to-year variability in the severity of America’s winters.
Top image: 50-year trend in the number of “warm” winter days (days falling in the warmest ten percent of daily mean surface temperature distribution). Bottom image: 50-year trend in the number of “cold” winter days (days falling in the coldest ten percent of daily mean surface temperature distribution).
(Sources: Higgins, RW et al. “Relationship between Climate Variability and Winter Temperature Extremes in the United States.” Journal of Climate 15 (2002) : 1555-1572 and Meehl, GA et al. “Relative increase of record high maximum temperatures compared to record low minimum temperatures in the U.S.” Geophysical Research Letters 36 (2009): L23701 and Thompson, David W.J. “Regional Climate Impacts of the Northern Hemisphere Annular Mode.” Science 293 (2001): 85-89 and Keenlyside, NS et al. “Advancing decadal-scale climate prediction in the North Atlantic Sector.” Nature 453 (2008): 84-88.)