A comprehensive new study extending the U.S. Historical Climatology Network (USHCN) record back to 1899 finds that both hot and cold temperature extremes across the contiguous United States have declined over the past 127 years. The research, performed by Dr. John R. Christy, Alabama State Climatologist (retired) and professor of atmospheric and Earth science at The University of Alabama in Huntsville, analyzed more than 40 million daily temperature observations to provide the most complete long-term view to date of U.S. extreme heat and cold. The paper is published in Theoretical and Applied Climatology. Excerpts below with my bolds and some added images.
Abstract
Knowledge of temperature extremes, and their potential changes within a climate system of increasing greenhouse gases, is of vital interest for humans and the infrastructure which supports them. To produce a better understanding of how daily extreme temperatures have changed over time in the conterminous US (CONUS), the United States Historical Climate Network (USHCN) database was extended back to 1899 and forward to 2025. The original 1,218 stations, selected in the 1980s by NOAA as capable of addressing climate concerns, have since been neglected – almost half of the stations have closed since 2000. Incomplete station records were supplemented with nearby stations with high correlation and removeable biases to provide time series for 1,211 of the stations with at least 92% of data present. Extreme temperature metrics for summer daily maximum temperatures and winter daily minimum temperatures were calculated. The general result is that metrics for extreme summer heat, e.g., hottest values, number of heatwave days, etc., show modest negative trends since 1899. Extreme cold temperature metrics also indicate a decline in their occurrences especially since the 1990s. In sum, instances of both hot and cold extreme metrics have declined since 1899. To demonstrate an application of this dataset we examined the claims of one source regarding changing temperature extremes, The National Climate Assessment 5.
This metric determines for each day of the season the particular year in which the hottest (coldest) TMax (TMin) occurred. There are 153 (122) days in the May-Sep (Dec-Mar, leap year) for which a daily record will be achieved. The number of extremes occurring in each year is calculated per station then geographically interpolated as discussed above. This metric is more robust than the single All-Time metric above as each station contributes 153 (122) values to the time series rather than just one. This also provides an indication of the incidence of multiple hot and cold records to help identify periods of excess heat (cold).
The expected value for a purely random process for the number of daily TMax (TMin) records would be 1.20 (0.96) in a given year per station for a 127-year record (i.e., 1.20 = 153/127 and 0.96 = 122/127). The results (Fig. 4) indicate again that 1936 contributed the most daily hot records for the CONUS at 6.7 per station but followed more closely by other years, with 1934 (5.3), 1931 (3.4) and 1911 and 1925 (3.3) completing the top five.
The number of coldest records occurred in 1899 (3.7) in association with February 1899 event. The following years experienced extreme cold as well, 1917 (3.3), 1989 (2.9), 1924 (2.4) and 1936 (2.4). Thus, 1936 was a year with many extremes on both ends of the thermometer.
Comparing the two metrics in Figs. 10 and 11 produces Fig. 12 which displays the sum and the difference, year-by-year of the 15-yr running means. The sum of days in extreme heat/cold declined from over 120 in the 1930s to about 75 since 1960. The conclusion here would be that the CONUS has experienced a decline of around 30% of these durative extreme events in the past 100 years. Along with this decline has been an increase in heatwave days vs. cold wave days since the 1970s, mainly due to the increase in heatwave days in the West (Fig. 10) and the decline in cold wave days overall.
Discussion
Overall, our project indicates that extremes in summer heat-related metrics for the CONUS as defined in the four questions above do not show increasing trends, but rather modest negative trends, and thus appear to be substantially affected by other forcings such as natural variability in addition to GHGs. There are positive TMax metric trends in western regions which are countered by larger negative trends elsewhere.
The number of cold extreme events has declined in the past 30 years too and this is likely, in part, related to the development of infrastructure around the stations which disturbs the nocturnal boundary layer, inhibiting the formation of the cold, shallow layer in which TMin is observed. Additionally, this result may be an early sign of atmospheric warming of the coldest air masses by the added GHGs (e.g., Krayenhoff et al. 2018), though this hypothesis has not been confirmed as a direct result of GHGs (e.g., Huang et al. 2023). Observations of the deep atmospheric temperature in the polar region north of + 60° latitude indicate a warming trend of + 0.47 °F (+ 0.26 C) decade− 1 since 1979 compared with a global trend of + 0.27 °F (+ 0.15 C) decade− 1 (Spencer et al. 2017). This would suggest Arctic air intrusions into the CONUS may be slightly warmer now than in the past century or so (for whatever reason) and thus consistent with the results shown here for a lessening of the magnitude of cold events in recent decades. However, we note the same area in the southern hemisphere shows virtually no warming (+ 0.05 °F (+ 0.03 C) decade− 1).
Conclusions
In the field of climate change, attention has been drawn to extreme metrics occurring in the last several years as evidence for human influences through increasing GHGs (e.g., USGCRP 2017; Seneviratne et al. 2021; Jay et al. 2023). Examining this aspect of climate and weather is appropriate since human thriveability is often constrained by the magnitude of the extremes that we experience. We describe here a dataset constructed to examine the occurrence through time of extreme temperature metrics in the CONUS for the coldest winter and hottest summer days since Dec 1898. The dataset is based on the 1,218 USHCN stations 1,211 of which have been supplemented to be “complete”, i.e., each station having at least 92% of days available for analysis.
The results indicate that extremes in heat-related metrics for daily TMax in the summer have not increased and in fact often show modest declines since 1899, due mostly to the early heat events during 1925–1954. This is consistent with Seneviratne et al. 2021 (IPCC AR6, their Fig. 11). Cold-related extreme events based on winter TMin show evidence of decreasing occurrences, two causes of which were proposed, (1) increasing human development around weather stations, and (2) an early response to increasing GHGs as they warm the coldest air first. When taken together, the occurrences of heat and cold extremes have declined over the past 127 years in the CONUS, i.e., the climate over the CONUS has become less impacted by temperature extremes to this point.
Relating this reduction of extreme events to increasing GHGs would be difficult
as the magnitude of the regional natural variability of weather and climate
is considerable in comparison to a small GHG-induced temperature rise.
Once the shifts were accommodated, the time series (Fig. 14) for Fresno 12-month running anomalies indicate very different results between TMax and TMin, which is a clear indication of the NCI of urbanization. The TMax time series indicates no trend through 2012 (slightly negative) but contains a relatively sudden rise in 2013 which is consistent with the entire western CONUS as seen in Figs. 5 and 10. The overall TMax trend is + 0.03 °F (+ 0.02 C) decade− 1. The trend in TMin is + 0.43 °F (+ 0.24 C) decade− 1.
The impact of Non-Climatic Influences (NCI) was considered in the temperature evolution of one USHCN station, Fresno California, as an example of a clear and large response to forcings unrelated to the increasing GHGs. In this case, the urbanization impact on TMin of 5 °F (~ 3 °C) is clearly apparent, while summer TMax (with urbanization) indicates a trend not significantly different from zero. Voluminous research has and will be performed on this aspect of surface temperature records as these types of influences need to be identified and removed so that changes in the background climate due to GHGs may be estimated with more confidence. We also demonstrated that one must be cautious when interpreting official statements about extreme weather events for the CONUS.
Science Matters 