Study: The Urban Heat Island Penalty Grows, Especially at Night

From the Pacific Northwest National Laboratory (PNNL), more confirmation of a night-time warming bias in climate records due to increased infrastructure.

RICHLAND, Wash.—City living has its perks: Live music, museums, trendy cafés and much more.

But urban living isn’t so cool when it comes to summer weather.

Living in a city translates to an extra two to six hours of uncomfortable weather per day in the summer for people in the eastern and central United States, according to research published Dec. 9 in Geophysical Research Letters. The scientists present this and related work this week at the annual meeting of the American Geophysical Union.

The additional uncomfortable hours occur mainly at night. Cities bake in the summer sun, with concrete, dark pavement and structures soaking in the heat during the day and releasing it at night, raising the nighttime heat index.

Day Night Heat Pulse
Urban areas are hotter than rural areas—and the heat stress penalty in U.S. cities is particularly strong during nighttime hours. (Animation by Sara Levine | Pacific Northwest National Laboratory)

Living in a city boosted the nighttime urban heat stress index anywhere from 1.9 to 4.9 degrees Celsius (about 3.5 to 9 degrees Fahrenheit) compared to nearby rural areas. The effect in the daytime was much less. During daylight hours, some cities actually become slightly more comfortable due to lower humidity.

Researchers at the Department of Energy’s Pacific Northwest National Laboratory also found that when a heat wave hits, the heat penalty of living in a city grows even more. Though everyone in the region is experiencing warmer temperatures, the extra amount of heat that the city residents feel compared to their rural neighbors becomes even larger. The hotter it is, the bigger the penalty—a concern especially in light of future global warming.

Heat Index at Night
​The urban heat index penalty is most marked at night, with an increase of up to about 5 degrees C (about 9 degrees F) in some urban areas. Even rural areas far away from urban centers heat up somewhat. (Graphic courtesy of Yun Qian | Pacific Northwest National Laboratory) ​

Because more than four out of five Americans live in urban areas, the scope of the penalty is likely extensive. Extreme heat is now listed routinely as a leading weather-related cause of death in the U.S., usually ahead of floods, wildfires and hurricanes.

“People have known about the urban heat island effect for more than 180 years, but the fact that the differences between nighttime and daytime and among different cities is so substantial is somewhat surprising,” said PNNL atmospheric scientist Yun Qian, corresponding author of the paper.

“Our findings show that urban living makes nighttime hours in all U.S. cities studied more uncomfortable,” Qian added.

Gap between urban and rural

The researchers found that in the U.S. cities studied, a temperature increase of 1 degree C (about 2 degrees F) translates to an extra 30 minutes of uncomfortable weather in a day for city residents. That means a heat wave, or future warming, that increases the background temperature 6 degrees C—roughly 11 degrees F—could translate to three more hours of uncomfortable weather in a day.

That’s above and beyond the extra discomfort that other residents living in less urban areas in the same region experience.

To do the study, scientists designed and ran a climate-urban model and analyzed the results for six summers, from 2008 to 2013. The study included large cities such as New York, Boston, Philadelphia, Washington, Atlanta, Miami, Chicago, Detroit, Houston and Dallas. The scientists looked at the heat stress index, which describes how the temperature feels when humidity is taken into account.

Then the team calculated “heat caution hours,” which they defined as hours when the heat index reached above 27 degrees C, or about 80 degrees F—the dividing line they used between comfortable and uncomfortable.

Hot Nights
Cities bake more than rural areas. During a heat wave, the gap is even more pronounced. (Composite image by Shannon Colson | Pacific Northwest National Laboratory)

The team observed a competing dynamic between temperature and humidity in cities. When it comes to comfort, cities have the advantage when it comes to humidity. There’s usually less humidity in urban areas because there aren’t as many plants, trees and other vegetation to give off moisture. But uncomfortable higher temperatures in cities are a more powerful factor, with an impact about five times that of humidity.

In all the cities studied, the air became more uncomfortable at night as temperatures rose, regardless of humidity. Beyond that, the effects of urban living varied a bit from region to region.

Regional differences

At night, urban areas in the Northeast, such as New York and Philadelphia, experienced the largest nighttime penalty, with the heat index increasing up to about 4.5 degrees C.

But in terms of the number of extra hours of uncomfortable weather, the effects were most pronounced in the southeastern United States. Residents of Atlanta, for instance, experience on average four to five extra hours a day during the summer of uncomfortable weather. People living in the Northeast are more likely to have two or three such hours.

In many cities, urban living makes the daytime hours less comfortable too. Philadelphia, Washington, New York and Chicago all experience a higher heat index and more heat stress during the daytime, compared to more rural regions nearby. But because of lower humidity that accompanied higher temperatures in cities, daytime comfort was slightly enhanced in the southern and eastern United States, including Dallas, Houston, Atlanta and Miami.

Caution Hours
As background temperatures go up, the number of uncomfortable hours in a day rise more in U.S. cities in the Southeast, like Atlanta and Miami, than in cities in the Northeast, such as Boston and New York. (Graphic courtesy of Yun Qian | Pacific Northwest National Laboratory)

“It’s important to note the differences between cities,” said coauthor T.C. Chakraborty. “What is useful for heat mitigation in Chicago, for example, may be different than what is useful in Miami. Even within one city, different neighborhoods can be very different—for example, an area along a river compared to a neighborhood with no grass or water nearby.”

Along with Qian, former PNNL postdoctoral researcher Chandan Sarangi, now at the Indian Institute of Technology in India, is a lead author of the paper. Other PNNL authors are Jianfeng Li, Ruby Leung and Ying Liu.

The research was funded by the DOE Office of Science.

# # #

Link for published paper here.

4.3 21 votes
Article Rating

Discover more from Watts Up With That?

Subscribe to get the latest posts sent to your email.

118 Comments
Inline Feedbacks
View all comments
Jeffery P
December 20, 2021 11:46 am

Didn’t Obama’s science advisor make a statement about stopping global warming by paint our roofs white? Wasn’t that an implicit endorsement of the idea that UHI is behind most of the alleged warming?

December 20, 2021 12:45 pm

Look at those middle temperature bubbles shaped by what I suppose is prevailing winds.

Clyde Spencer
December 20, 2021 8:19 pm

… fact that the differences between nighttime and daytime and among different cities is so substantial is somewhat surprising, …

Not to anyone who has been paying attention for the last ten or twenty years!

When it comes to comfort, cities have the advantage when it comes to humidity.

Not in Phoenix. Gas stations, bus stops, and home patios have misters. Most homes have pools. The area of golf courses and urban green-ways probably equals the area of schools.

DayHay
December 22, 2021 1:23 pm

Could someone provide an overlay that shows where all the temperature stations are located?

bdgwx
Reply to  DayHay
December 22, 2021 1:53 pm

Here is a pretty good interactive map that lets you select the stations and see the monthly temperatures for each.

Bindidon
December 24, 2021 6:49 am

It’s a bit late to post on this thread, but… mieux vaut tard que jamais.

I read in a comment above:

The UHI impacts nighttime temps. That means that the daily average of Tmax and Tmin goes up. The daily average going up causes the UHI to impact the global average temperature.

Tmax is a small player in the growth of daily, monthly, annual, and global average. Think Tmin instead.

To such rather superficial claims about UHI’s alleged influence on Tmin temperatures, I prefer to have a look at real data, and to compare

  • those stations in the US which are really the least ‘UHI suspects’, the USCRN network (about 130 active units)
  • with all other far more ‘UHI suspected’ stations available in the raw GHCN daily data set (about 4,000 units on yearly average).

*
To let anyone understand what I mean with ‘UHI suspects’, I post a link to the complete USCRN info, including photos and lat/lon/elev coordinates (caution: lat & lon are here very approximating values):

https://www.ncei.noaa.gov/pub/data/uscrn/documentation/site/photos/stationsbystate_lores.pdf-20161212

This of course does not mean that all currently active USCRN stations would be free of any UHI influence. A look at the very first entry in the list

63869 30.5485 -87.8757 29.0 AL_Fairhope_3_NE

clearly shows that the Fairhope station in Alabama might be more affected by UHI than is for example

26563 60.7237 -150.4484 86.0 AK_Kenai_29_ENE

which in comparison is really ‘in the middle of nowhere’: a look at Google Maps using the coordinates of the stations is significative.

*
What is less significative however, is the comparison, for CONUS+AK, of the USCRN average (of the hourly data) with that of those GHCN daily stations located in the same 2.5 degree grid cells as their USCRN mates.

The monthly averages – starting with 2007 and based on anomalies wrt the mean giving the most USCRN station data (2016-2020) – were split into Tmin and Tmax.

1) Tmin

comment image

2) Tmax

comment image

*
What the two graphs of course don’t show is that

  • while the linear Tmax estimates for USCRN vs. GHCN daily are, with 0.35 resp. 0.34 °C / decade, nearly identical,
  • those for Tmin are, with 0.43 resp. 0.55 °C / decade higher than the Tmax estimates, and especially higher for GHCN daily than for USCRN.

*
But… who tells us that these differences are due to UHI?

December 30, 2021 7:27 pm

Thanks, but the animation is NOCOMM, worse than superfluous.