The stated purpose of the U.S. National Assessment's Climate Change Impacts on the United States; The Potential Consequences of Climate Variability and Change (USNA) is to "assess the risks and opportunities for the United States...associated with increased climate change." But the USNA has turned out to be one of the most misleading publicly funded reports on climate change this nation has ever produced. The two climate models on which it is primarily based—one developed by the Canadian Climate Centre and the other by the Hadley Centre in the United Kingdom—cannot correctly reproduce observed climate. What's more, the two models often produce markedly different forecasts of future climate. • In addition to large-scale inaccuracies, the models' spatial resolution is too coarse to include most small-scale processes—the type of processes responsible for local weather patterns. Yet the USNA breaks the country into eight regions and within each region depicts local ecosystem changes as a result of their predicted climate trends during the next 100 years. In this continuing series, we examine in detail not only each of those regions, comparing past observations with forecasts, but also other issues relating to the USNA and its influence.
 

Hiding Facts with Figures

The U.S. National Assessment and the latest report from the Intergovernmental Panel on Climate Change both forecast that extremely high daily temperatures will occur more frequently. Fortunately, trends in historical observations just don't seem to support the idea.

The distribution of daily temperatures for a particular location, when examined in aggregate, tends to take the shape of a bell curve—also known as the "normal distribution." The temperature on the majority of the days turns out to be near the average (or mean), and the number of days with temperatures above or below average drops off as the size of the departure from average grows. The normal distribution is characterized by both its average value, and its "variance"; that is, how close most of the observations are to average. A large variance means that a lot of observations are quite different from the mean, while a small variance means that most temperatures are very close to the mean (Figure 1).

Figure 1. These two illustrations have the same mean, but the variance is larger for the bottom figure, thus there are more occurrences of extremely high and low temperatures.

If temperatures change in the future, they may do so in a number of ways; the average temperature may increase or decrease and/or the variance may increase of decrease. The impact of any temperature change is critically dependent on the combination of those changes, since the probability of extreme temperature occurrence can vary greatly depending on the exact nature of the changes in the distribution. A rise in the mean accompanied by an increase in the variance leads to an enormous rise in the frequency of extremely hot days. Conversely, a rise in the mean temperature that is accompanied by a decrease in the variance can actually result in fewer numbers of extremely hot days (examples of such cases are illustrated in Figure 2). Obviously, a key to assessing future impacts is understanding the relationship between annual mean temperature and daily temperature variance.

Figure 2. Top: A rise in temperature accompanied by a decrease in variance can lead to a reduction in the number of extreme temperatures. Bottom: A rise in temperature accompanied by an increase in variance will lead to a far greater number of extremely hot days.

This relationship across the United States was the subject of an investigation by Indiana University climatologist Scott Robeson. Robeson carefully analyzed the joint behavior of the average temperature and the temperature variance for 1,062 stations distributed across the United States. He found that the vast majority of the stations (covering more than 90 percent of the country) either exhibited no relationship between average annual temperature and daily temperature variance or a negative relationship (that is, as the average temperature increased the average variance decreased) (Figure 3).

Figure 3. The percentage of the United States where warmer temperatures are associated with declining variance (inverted triangles) is much greater than the portion where rising temperatures are associated with increasing variance (triangles).

It is somewhat curious that in the latest IPCC release, there is a section dealing with the relationship between temperature and temperature variance that includes a flashy illustration that shows only the possibility for bad things to happen (more record hot weather) while the accompanying text, now supported by Robeson's new research results, describes actual observations that tell a different story (one closer to Figure 2a, above).

There can be only one reason: The IPCC is much more interested in pushing extreme scenarios and impacts than they are in pushing a much more benign truth.

References:

Robeson, S. M., 2002. Relationships between mean and standard deviation of air temperature: implications for global warming. Climate Research, 22, 205–213.

IPCC, 2001. Climate change 2001: The scientific basis. Contribution of Working Group I to the Third Assessment Report of the IPCC. Cambridge:Cambridge University Press, 881 pp.