I'm a professor at U Michigan and lead a course on climate change problem solving. These articles often come from and contribute to the course.
By: Dr. Ricky Rood , 6:27 AM GMT on November 28, 2013
What Can I Say about Heat Waves? Heat Waves (5)
This blog is about a paper on extreme heat written by my student Evan Oswald and myself. I don’t usually write about my own research, but this paper poses some interesting challenges to think about heat, heat waves, climate change and public health. Or I might say, how do I explain this to my epidemiologist friends?
What we set out to do with this research was to quantify how observations from surface weather stations represent extreme heat events that are threats to human health. We started with the station observations because most often those responsible for heat warnings and those undertaking planning for climate change start with station observations in their locality. There are a lot of reasons for this choice. An important one is that these observations and any of their local peculiarities are usually well known. Hence, there is experience and the knowledge and trust that come from that experience. Once we described the behavior of all of the station observations, we had two planned paths. The first path was to see if the gridded datasets used in climate-change planning had the same behavior as the station observations. The second path was to compare the station observations to a high-resolution network of observations in a city and see how well, for example, the measurements at the airport or weather office represented the details of the city.
Of course, many researchers have looked at the station data and documented trends in heat. To earn a Ph.D., a student has to do original and independent research. There are a number of attributes that distinguish this research. Most notably, we have been working with a team of public health experts (meet Marie O’Neill), and we had a desire to use measures of environmental heat that have been found to be important in public health studies. To a meteorologist, heat might seem simple, but the human health impacts of heat are complicated. For example there is the impact of very high temperatures on those working or training outside. Another example is the threat of persistent heat, day and night, on the chronically ill who might be housebound. There is a link between heat and humidity, with many of us Southerners knowing that “it’s not the heat, but the humidity,” and not thinking about the effects of dehydration that are important in the desert Southwest. For this reason we started by looking only at temperature and not some measure of comfort such as a heat index. Then there’s a sort of obvious one, public health experts are most interested in heat effects during and around summertime, whereas to a climate scientist a “heat wave” in the winter can be as interesting and as important as a summer heat wave. There are many other complications, but I hope I have made my point, there is meaningful research to be done.
In the research reported in “A trend analysis of the1930-2010 extreme heat events in the continental U.S.”, we focused on the warm season, end of spring to the beginning of fall. We also focused on different types and characteristics of heat waves. We defined heat waves for daytime maxima and for nighttime minima. We looked at, for example, duration of heat waves, how many days did they last? Here I am going to only write about the trends that we reported in duration for three different time periods, 1930-1970, 1970-2010, and the combined time period of 1930-2010. Our study area was the continental United States.
Why these three time periods? Lot of reasons, we wanted to include the well known hot times during the 1930s, otherwise we would be accused of cheating. We did not go earlier than 1930, because we felt that the quality of the observations decreased substantially. When I was a student in the late 1970s, I remember getting excited when, say, the data for 1976 was released. Then 1977. I’d write papers about what the future would hold. Now low and behold, I have been fortunate enough to live long enough to have more than my own 30-year period. Thirty years of average temperatures is the traditional definition of “climate.” Hence, splitting things at 1970 we have two equal 40-year records, which allows us to investigate the sensitivity to which 30-year period, which “climate,” is chosen.
Lot of introduction here, so let’s get to a result. In Figure 1, I show the decadal trends at each station in the mean duration of EHEs during the 1930-1970 period. The top map shows heat events based on nighttime minimum temperatures. The bottom map shows heat events based on daytime maximum temperatures. The middle map shows events when both the maximum and minimum were elevated; that is, it did not cool off very much at night.
Figure 1: The decadal trends at each station in the mean duration of EHEs during the 1930-1970 period. The trend significance (alpha=0.10) is indicated by symbol shading. The graduated symbol groupings are based on standard deviations away from the zero value, and are different for each map. The trends in Tmin-based EHEs (top), Tmnx-based EHEs (middle) and Tmax-based EHEs (bottom) are all shown. Tmin is based on nighttime lows, Tmax on daytime highs and Tmnx require both highs and lows to be elevated.
The trends in the minimum temperature are generally positive. The exception is in the northern part of the Great Plains, right in the east-west center of the country. The largest red squares in the figure tell us that for every 2 decades we are seeing about 1 additional day of duration of very warm nighttime temperature. The bottom map for daytime minimum tells a different story. In the West there is, mostly, a warming trend in the daytime maximums. In the center of the country there is a pretty strong cooling trend. Some of my more skeptical readers and friends will go, “see there is no global warming.” In class, I would then make the assignment to describe what this figure does or does not tell us about global warming. Perhaps, I will distract a few people to carry on their arguments in the comments. It would be terribly pedantic for me to make such an assignment here, and pure hubris to expect responses.
In any case, we do see this big area of cooling of daytime maximums in the middle of the country. This was not a surprise to us, because there is growing documentation of the “Midwest Warming Hole.” This does, however, offer a challenge when discussing heat waves with my epidemiologist friends. It also might stand a little in conflict with reports such as “Heat in the Heartland,” a widely used document, for which I provided some review comments. In the next blog, I will breakdown the information in the figure a little bit, and then I will start to buildup a description that might be more usable by the public health planner.
Some earlier Hot Blogs
Russian Heat Wave
Heat Wave Series
The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.
Comments will take a few seconds to appear.