| Previous Advisory | Return to Advisories List | Download PDF |

Free Acrobat Reader needed

Advisory 01/07:

THE HOT GET HOTTER?
URBAN WARMING AND AIR QUALITY

This Advisory was prompted by a trip through our Nation's Capital and National Airport on August 8, where these subjective eyes perceived air quality about as bad as ever seen in the region. Temperatures in the upper 90s were common in the metropolitan area, along with dewpoints in the soppy upper 70s. (My home weather monitoring equipment in Augusta County hit a ridiculous 80° dewpoint late in the afternoon).

Thanks in no small part to the Nitrogen Oxides emanating from the massive numbers of Sport-Utes idling in unison on I-395 and I-66, and with additional help from a few other more exotic anthropogenic compounds, visibility tanked as ozone levels climbed under a relentless sun. The lack of any appreciable breeze made sure that the chemical cauldron continued to brew.

The fact is that this type of sleaziness increases exponentially with temperature, everything else being equal (a condition that rarely obtains unless you are flipping a perfect coin). So this prompted the usual question: are the hottest temperatures getting hotter?

TAKE IT FROM THE TOP...

Let's start at the beginning, which is the statewide average annual temperature history, beginning in 1896 (Figure 1).This record comes from the National Climatic Data Center in Asheville, North Carolina and is generally thought to be one of the most reliable records in existence.

Figure 1. Statewide average temperatures in Virginia show no obvious warming trend.

Despite what you read every day in the newspaper, it is pretty apparent that there is simply no warming trend in the context of this very long record. Temperatures in recent decades look an awful lot like they did in the 1920–60 period, and it's apparent that the 1960s and 70s (which is when everyone, for some reason, seems to start their climate "memory") were quite cool. One can fashion a warming trend from 1896 to about 1950, or before there was much of a change in the carbon dioxide greenhouse effect. Then there's a sharp drop which has yet to be completely recovered.

The next time a story comes along about how Virginia is warming up etc., let's hope (in vain, for sure) that some enterprising reporter actually prints this graphic.

Most people would look at Figure 1 and say something like, "The more things change, the more they remain the same." One logical conclusion would be that extreme temperatures are probably behaving just like they have in the past. In other words, the number of daily high and low temperature records that are set should remain fairly constant.

Let's start this discussion with the best computer-accessible rural record in Virginia, which is Woodstock, a teeny burgh in the middle of the Shenandoah Valley. Its temperature history starts in 1930 (though much more data is available in hard copy). As is apparent from our Figure 2, which plots out temperatures from four locations, Woodstock mirrors the Virginia average in showing no overall net temperature change.

Figure 2. Temperatures in recent decades for Woodstock, Washington National, Charlottesville, and Richmond Byrd. The warming of the three urban stations is evident.

(A peculiarity of data management is that most digital records begin in 1948, even though we often have hard-copy records of much greater length. So our comparative graphs in this article start in that year, even though the records we examine, described in the text, are much longer).

We then examined the record, beginning on January 1, 1990, to see if there is any unusual recent behavior in record temperatures. This is 15% of the total record, and, everything else being equal, it should contain roughly 15% of the record temperatures, or 56 record highs and as many record lows.

In reality, this period shows 48 record highs and 37 record lows—a bit shy of the expected value for each, but not alarmingly so. The difference between the number of high and low records isn't beyond what we would expect, given that record data often occurs in runs of several days. The dominance of the 1930 records on July and August temperatures (which we have written about in previous Advisories) is very evident in the Woodstock data, for example.

Now, let's amble about 30 miles up I-81, turn right onto I-66 at Strasburg, and make our way to DC National Airport (actually in Northern Virginia). About 20 miles west of there, horse farms and pastures begin to give way to concrete and houses. On most days, we're likely to join a conga line of panting SUVs, especially as I-66 winds through Falls Church and Arlington. Welcome to the city.

As shown in our comparative graph, National's temperatures tend to run about a degree warmer than Woodstock in the beginning of the record. By the time we get to 1970, the differential is approaching two degrees, and it maxes around 2.3 degrees in recent decades.

There is no magical air conditioner somehow keeping rural Virginia and Woodstock cool, while DC heats up. Instead, this is the well-known urban heat island working full blast, and what an effect it has had on record temperatures, thanks to the twin economic engines of big government and high technology.

Given the length of DC's record, we should see 31 daily high and low temperatures since January 1, 1990. Beginning in 1990, there have been 57 days which set record highs, or nearly twice the expected value, and only three record lows (!), or about one-tenth what we would expect in the absence of either local or regional climate change. What is remarkable about this is that DC appears to have saturated its urban warming effect about 25 years ago.

Now let's drive back out I-66, turn south on US 29 and head to Charlottesville. About eight miles north of town, near CHO Airport, a massive number of new stoplights appears to have sprouted up, along with many strip malls. This is a very recent development. Twenty years ago, the town virtually stopped about two miles north of the University. Charlottesville is experiencing a massive and sudden urbanization.

Charlottesville's Leander McCormick Observatory is one of the best-maintained climate-monitoring sites in the nation. Its record begins in 1896, so since January 1, 1990 there should be around 38 record highs and lows.

Note the sudden appearance of urbanization in Charlottesville's record. In Figure 2, there is a sudden split between the Washington National and Charlottesville temperatures beginning in the early 1970s, and concurrent with the first major developments in Northern Virginia. Charlottesville just experienced this change, and by the late 1990s, temperatures are suddenly back to Washington's values. In fact, this corresponds exactly to the period when McCormick Observatory became surrounded by development on all sides.

So, we should expect an imbalance of high temperature records, though not so massive as in DC. Charlottesville shows 58 days with record highs (about 50% more than expected) and 39 record lows, which is pretty much what should have occurred.

Figure 3. Bar charts showing the number of record highs and lows, along with expected values, for our four stations. The imbalance at National Airport is truly amazing.

RICHMOND

Finally, Richmond. Since 1937, the official record is from Byrd Airport. The overall record we used extends back to 1895 and is supplemented with downtown data prior to 1937. Although most people might think this is an urban record, the region around Byrd has been fairly slow to develop a case of exponential growth.

As is the case for Charlottesville, we would expect 38 record highs and lows. Instead, again there's a bias towards the highs, with 49, compared to only 22 lows, evidence for substantial urbanization, but not as severe as around National Airport.

SEASONALITY

The next obvious question pertains to seasonality. Forget about it. There's none. We looked at cold/warm half-year samples and the split is almost exactly even. Unfortunately, DC's propensity to set a record high temperature in the summer is the same as it is in the winter. For what it's worth, since 1990, the chance of setting an individual daily record in the warm half of the year on a given day has been running at 1 in 70. It "should" have been 1 in 125 during this period.

Similarly, the chance of setting a daily record low in our Nation's Capital during this period was a stunningly low 1 in 675.

SUMMARY

1. There's no warming trend in the 106-year statewide Virginia temperature record.

2. There's no warming trend at most rural Virginia stations (which is no surprise, given the average behavior of the record).

3. A disproportionate number of high temperature records are being set in recent years, and this is consistent with the amount of urbanization surrounding each station.

4. In DC, 54 days have set record high temperatures since January 1, 1990, while only three have shown record lows. This stunning disproportionateness, along with the belching of a jillion idling vehicles, including lots of pretty emissive SUVs, is creating increasingly lousy air quality.

This is not a pretty picture. It is very clear that even without any contribution from "global warming," we are warming up the urban cores of our cities at a prodigious rate. Record high summer temperatures almost invariably associate with very low air quality, and we are making more and more of these. As locations such as Charlottesville and Richmond become increasingly "DC-like" in their pavement and emissions, expect local air quality to plummet as summer high temperatures increase.

There are some ways around this, but that's the subject of another Advisory.