Surface temperature departures from the long-term average for the week of July 20-27, 2011, the height of the heat wave in much of the southern, midwestern, and eastern U.S. Temperature measurements were collected by NASA’s Terra satellite. Reds indicate warmer than average, and blues below average. More info. (NASA)

In total, more than 9,000 warm-temperature records were either tied or set during the month of July, and more continue to fall in August. Last week, Dallas, Texas, saw this summer’s streak of consecutive 100-degree days end at 40, two short of the city’s all-time record. On Monday, Houston hit 100 degrees for the 15th day in a row, setting a record.

The severity and vast reach of this summer’s extreme heat has led to questions about whether global climate change is already causing the emergence of a new summer temperature regime, with more frequent and intense heat waves, as has been projected by numerous studies.

To get an expert perspective on this issue, I turned to Noah Diffenbaugh, an assistant professor of environmental Earth systems science and a center fellow in the Woods Institute for the Environment at Stanford University. Diffenbaugh has published several studies on how extreme heat events may change in a warming climate. Here is an edited transcript of our conversation, which took place on Aug. 10.

Andrew Freedman: What do you say to people who are experiencing the ridiculously hot weather this summer and wondering what it means?

Noah Diffenbaugh: Well we can expect just from a statistical perspective that ... what was the hot extreme in the cooler conditions will likely become more frequent with mean warming. There has been a lot of work on observations of the occurrence of hot days or the number of record high temperatures in the U.S. ... we’ve known for a while that the expectation of increasing hot extremes with increasing global mean temperature has been emerging. That can occur simply from being a result of the statistics.

In terms of the atmospheric conditions, that’s a more complicated question, because what we experience in terms of heat extremes, those phenomena are not just a result of statistics. They’re a result of how the real physics of the real world operate. We do understand the atmosphere and surface conditions that are associated with extreme heat quite well, and in fact we even understand some of the oceanic conditions that have teleconnections through the atmosphere to create temperature extremes in some locations. A lot of the work that I’ve been focused on over the last couple of years is focused on extreme heat and trying to understand how we expect those processes, those atmosphere and ocean and soil moisture processes, to respond to global warming.

Our work suggests both over the near-term period, over the next few decades, both in the U.S. and elsewhere the occurrence of what has been extremely hot conditions is likely to become far more normal than extreme. And in terms of the atmospheric conditions we see not only that statistical shift that we’d expect with global warming but we also see ... for the U.S. ... a change in the atmosphere and soil conditions that’s very consistent with changes that have occurred during extreme hot events.

We find to my surprise that looking at a large suite of climate model experiments that large areas of the globe are likely to see emergence of that unprecedented heat regime over the next few decades, including large areas of the United States. And when we look at ... observations of temperature from around the globe, we see that many regions of the globe - land regions where people live - have been experiencing this emergence over the last four decades, so that what was the hottest season earlier in the 20th century has become more and more common over the late 20th century and into the 21st century.

AF: Where in the U.S. do your model simulations suggest a regime shift to hotter summer temperatures will have the greatest impact?

ND: In general our results suggest that the Western U.S . shows the most robust intensification, and then the second area of intensification would be along the Eastern Seaboard, particularly in the Northeast. It’s the Central US that is the noisiest. This is partly due to the fact that the real climate dynamics are noisy in terms of the warm season - there is a lot of warm season variability. Whereas in the Southwest it’s already hot and dry.... There’s lower baseline variability, so the same amount of warming causes greater occurrence of the hot extreme ... than if the variability is high.

AF: It would seem to me that that the Eastern Seaboard would be pretty variable as well, even if not as much as the Central U.S.?

ND: One of the differences, at least in the climate model world, is that there is more of a direct influence of the ocean over the Eastern Seaboard than there is over the Central U.S. the ocean warms, then those land temperatures warm in a more constrained way than over the Central U.S., where the atmospheric variability plays a larger role - in the model world

We see this pattern ... with less robust intensification over the Central U.S., both in our high-resolution climate model experiments and when we look at the large suite of global climate models that have been developed around the world.

AF: The ironic thing is the way this summer has played out is kind of opposite of the general pattern shown by the climate models - we’ve had an ocean-related cool regime in the West, ridiculous drought and heat in the South Central states, heat plus flooding in the North Central states, and then just repeated unbearable heat in the mid-Atlantic and Northeast. So it’s a little different from what may be coming down the pike?

ND: I think there’s a pretty vocal interest in comparing the current conditions with expectations for elevated greenhouse forcing, and basing judgment about the credibility of those expectations on the current conditions. And there’s really no scientific basis for that judgment.

As humans we’re very used to considering very long-term patterns and long-term trends within very noisy series of information about a system, and parsing our conclusions about the long-term patterns and the long-term trends within the context of that noisy series.

A couple of real-world examples: If you look at the box score from last night’s baseball game, there’s a lot of risk in concluding which player is the best hitter on the team based on what you find in that one box score. We’re very used to seeing, say, player performance vary from game to game and making conclusions about the most productive player based on a large sample of observations, likewise with the stock market. We know that the stock market varies not only day-to-day, but hour-to-hour and minute-to-minute, and we can watch that variation online almost in real-time….

There is a lot of risk in drawing conclusions about what the Dow Jones industrial average index value will be 10 years from now based on what happened yesterday… So again, the current state, or the most recent departure, is not a rational basis for making conclusions about the long-term trend or the long-term state of a system that’s noisy.

A short way to put it is, I’m looking at the NCDC July 2011 divisional ranks for temperature right now and I don’t think that the paper that we published in Climatic Change about severe heat globally is any more correct because many divisions in the Eastern Seaboard are much above normal for July and that looks like the Eastern Seaboard in that paper.

July 2011 climate divisional ranks. (National Climatic Data Center)

I think that often our understanding of climate change gets held to a different standard. It gets associated with the current conditions and agreement or disagreement with the current conditions, whereas I think people - as the examples that I gave with the box score and with the Dow Jones - people are used to experiencing the noise in a series even if that series is changing. So I think people would have a basis for making decisions about what the change in the stock market or a change in a particular stock over the next 10 or 20 or 30 years is likely to be based on their understanding of the conditions that influence the stock market, and that’s a different problem than understanding the fluctuations that occurred today or yesterday or over the last month.

I’m not putting any more stock in the current conditions that agree - that look the same as what we conclude for the future - and less in those that disagree. I’m simply saying that one snapshot is not a basis in making any conclusion about our understanding of the response of the climate system to a transient change in radiant forcing.

Diffenbaugh added:

… For all of the uncertainty in all of these extremes and trying to get to local scales and trying to quantify the time of emergence [of heat extremes] in certain locations, I don’t think that any of that uncertainty really calls into question ... that global mean temperatures are increasing and ... that human emissions of greenhouse gases are the primary driver of that increase in global mean temperature, and ... that in general, all things being equal, we would expect that extreme hot events would increase with an increase in global mean temperature.

…I’m not calling any of that into question.