Public Release: 

How Does Your City Grow?

Penn State

Philadelphia, Pa. -- Microclimates caused by urbanization might not have much influence on regional climate, but they do impact local liveability, according to Penn State meteorologists.

Urban planners need a way to predict the growth of cities and assess changing microclimates, and the researchers have a computer model they hope will help.

"The model provides a probability that any location will be developed at each specific time," says Dr. Toby N. Carlson, professor of meteorology. "With this, we can eventually predict the types of land use and the microclimates that will form."

The model was developed by Keith Clarke at the University of California, Santa Barbara, but the researchers are applying it to the Philadelphia metropolitan area and southeastern Pennsylvania. An animation of the transitions from 1987 to 2050 can be found at

Carlson; David Ripley, research associate; and Traci Arthur, graduate student in meteorology, are using satellite image data to calculate land use types, such as forest crop, commercial, residential and water. They are also looking at amount of vegetation, impervious surface area and various surface climate parameters. Their objective is to predict all these variables using the Clarke model, which currently provides only a probability that a given surface will become developed.

The Penn State researchers examined satellite images over Eastern Pennsylvania between 1987 and 1996.

"In Pennsylvania, forest is remaining forest, but farmland is turning over into suburban and urban areas," Carlson told attendees today (Feb. 17) of the annual meeting of the American Association for the Advancement of Science. "Impervious surface area is increasing dramatically well outside older urban centers such as Philadelphia.

In the Clarke prediction model, downtown areas remain stable with commercial development creeping out along major roads. The growth of housing developments can be seen on smaller roads. The base map data for this simulation comes from satellite remote sensing data.

"In the satellite imagery, new housing developments sometimes look like deserts -- hot and dry -- when viewed from a meteorologic perspective," says Carlson. "Then lawns and trees are planted and the situation gets better, but it never gets back to the microclimate of the agricultural fields. It always stays warmer and dryer."

These changes take place because of the changes in land use. Asphalt and concrete roads, sidewalks and driveways create impervious areas from which no moisture evaporates. Normally, evaporating moisture cools the ground, but with pavement, there is no cooling and it becomes noticeably warmer. Land use changes also change runoff patterns, especially when large areas impervious to seepage occur. The increased runoff can cause local flooding.

"We would like managers and urban planners to be able to see what changes have already occurred and might occur in the future, and be able to calculate such things as surface temperature index, a measure of how warm an area is."

Carlson and his associates provide the Clarke prediction model with the local terrain, roads, bodies of water and protected areas and the land surface type as diagnosed from the satellite imagery.

"One thing we haven't done yet, is validate the Clarke model for Eastern Pennsylvania," says Carlson.

To validate the model, the researchers will feed in information from the past and run the model up to the present. If the model's simulation matches the present urban picture as diagnosed from current satellite imaging, then the model is validated. If it doesn't, then parts of the model are tweaked until the past creates the present.

EDITORS: Dr. Carlson may be reached at (814) 863-1582 or by email.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.