Environmental Intelligence

Eric Barron
May 01, 2002

"It is practically impossible to find an element of commerce that is not connected to weather and climate, whether it's the production of winter coats or the sale of ice cream cones," said Eric Barron in the sixth of this year's Lectures on the Frontiers of Science. Likewise, weather affects transportation, agriculture, and defense, even how well we preserve ecosystems like the Everglades.

deer in snow covered forest
Charles Fergus

We've learned to forecast the weather. We can predict a late spring storm in central Pennsylvania. Can we learn to do the same for climate change?

"Our ability to predict the weather is growing," he added, "and when we make predictions, decision makers are taking advantage of them and they're demanding more and more."

Barron, a distinguished professor of geosciences and director of the Environment Institute in Penn State's College of Earth and Mineral Sciences, sees a trend.

In recent years meteorologists and other scientists have "been expanding the forecasting family" to include air quality in cities, lightning strikes, energy demand, ultraviolet radiation, and El Niño years. At the same time, the "time scales and space scales of interest are expanding dramatically," he said, "from deciding whether to take an umbrella to work to deciding whether to spray for mosquitoes or what crop to plant next spring. And people who intelligently use this information are making a better profit."

Can we take this forecasting expertise and apply it to how climate change will affect a particular spot on the globe 100 years from now?

We can, Barron believes, if we build an Environmental Intelligence Center.

"Think about a war room," Barron said. "Don't start thinking about protagonists and antagonists, think about how the military approaches a problem." Barron had met a brigadier general who focuses on how weather and climate affect military decisions. The general described what the military calls a "joint info-sphere," combining military intelligence, air surveillance, observations from the ground, and satellite data, and covering such issues as terrain, weather, atmospheric conditions, logistics, utilities, cultural features, and the whereabouts of enemy forces. "They want information 'from the mud to the sun,' they say, and it's all pulled by a decision maker—the general. It's not pushed by what I, the scientist, can do in making predictions for the future. It's pulled by the need to anticipate and exploit the weather to complete a certain mission."

To understand climate change, Barron said, we want "from the mud to the sun" coverage, too—or at least from the frog to the atmosphere. According to the 2000 report of the U.S. Global Change Research Program, Climage Change Impacts on the United States: The Potential Consequences of Climate Variability and Change, scientists studying climate change need to pay more attention to multiple stresses—such as when land use patterns, drought, and disease combine to cause physical deformities in frogs.

"Why can't we do that?" Barron asked. "We can't integrate all the observations.

"I can't tell you how many observations are out there," he added. The U.S. Geological Survey has water-monitoring stations. The EPA monitors air quality. The Weather Service collects atmospheric data from aircraft and satellites and balloons. Other groups have data sets from ice cores and tree rings and sea surface temperatures and ocean currents. "Imagine what the data sets are for human health," Barron said. "Point-of-service, collected at a doctor's office, aggregated by county. Imagine what the data sets look like if you are trying to do mammal counts or bird counts, to tell their distribution.

"Right now, each one of those different observations is in a different bin." All are governed by a different agency, collected with a different focus, and stored in a different place. "And we can't put them together."

In addition to acquiring and organizing it all, there's a problem of combining this information with the computer models used to predict climate change. Said Barron, "We don't have a common modeling framework."

While the Weather Service can predict temperatures for one town or another, the Global Circulation Models used to predict climate change necessarily have a coarser resolution. They're already taxing the capabilities of the fastest computers, taking three billion calculations for each prediction. Even so, Barron said, "we just barely have the oceans and the atmosphere in there and we're trying to add the ice.

"Pennsylvania has two grid points in that modeling framework. How in the world am I going to get down to the level of that ecosystem?"

We physically can't predict climate change on a global scale and end up with a forecast of any use to the governor of a state or the CEO of a corporation, yet globally is how scientists have approached "global" warming for 20 years.

How should they be thinking about it? More like a regional weather forecast, Barron believes. "It becomes a tractable problem if we look at it on a regional scale," he said. "It won't wipe out our computer resources. You can take the framework of a high-resolution weather model and hang an ecosystem model on it." With a cohesive system of collecting observations and of enabling access to information, it would become a regional Environmental Intelligence Center.

Cutting a climate model down to size in this way would allow forecasters to ask a wider set of questions. Predictions could be made for the short-term, for a season or a year ahead, or for decades or centuries to come, and then compared with historical trends. Decision makers could ask questions about their specific part of the globe and get the answers they need to make a useful decision. For instance, how far will West Nile virus spread next summer? If the winter is warmer than usual, mosquitoes that are usually killed off by the cold can successfully overwinter. If the spring is unusually dry, birds will flock to the remaining available waterholes, where they most likely will come in contact with and be infected by the mosquitoes congregating there to breed. If the birds are of certain species, they might travel from field to field, rather than staying in thick woods, and so will come into contact with horses. If the late spring or early summer is rainy, the mosquito population will then soar. Each of these factors affecting the spread of the disease is governed by climate. A regional Environmental Intelligence Center that could integrate observations of weather and climate, land use and recreation patterns, pollutants, the behavior and density of animal vectors, and the various kinds of insecticides could predict that spread, letting healthcare workers take preventative measures.

"You're always going to make decisions based on the information you have," Barron noted. "If we can make the information set richer and more reliable and more useful, we'll end up with more good decisions than bad decisions.

"My view is that the ability to anticipate is what makes knowledge really valuable," Barron said. "Think of knowledge as a way to promote both economic activity and environmental stewardship. Plus, being able to anticipate is a powerful learning tool. You learn if your predictions were right and what it takes to improve on them."

A key part of Barron's Environmental Intelligence Center is what he calls the "vigorous connection" between scientists and decision-makers. "We're starting to recognize that climate scientists shouldn't be just out there playing in our ponds. We have to think of ourselves as serving society. Just throwing knowledge over the transom doesn't seem to be working."

He gave an example: Several years ago, he was invited to join a study of the transition from Neanderthals to modern humans in Europe during the Ice Ages. "I was the pet climate modeler for this group of archeologists and anthropologists," he recalled. "It was an enormously fun experience to produce climate model simulations for the time period 30- to 50,000 years ago, and I kept producing what I like to produce: Here's the average winter, here's the average year, here's the average precipitation for the summer." His colleagues were not terribly interested in his results. One day, sitting in their shared office, Barron said, he started to listen to them. Neanderthals, he learned, didn't have a lot of hair on their bodies. They had no tools to speak of. Their teeth showed signs of wear as if they'd been chewing on furs. "I realized that the archeologists and anthropologists were talking about how long it takes for the skin to freeze if it's not clothed. All of the sudden I realized that none of these things I've been flying over the transom was really what they wanted. They wanted to know wind chill factors." He produced the maps, and the team discovered, he said, "that the tool cultures of Europe line up with the wind chill factors. The advanced tool cultures were associated with areas of deeper wind chill. They were more able to stand it."

That kind of "decision-maker pull," Barron concluded, is what will transform the current science of climate change into the new sustainability science. "This transition from will climate change occur to how significant is it is the same thing as saying that we're moving to a phase where we're asking the question, How can science be of service to society?

"Science should always be in service to society, but what I'm claiming is our approach to environmental problems and to climate change in particular should be much more deliberate.

"We need decision-maker pull. A decision-maker is anyone from you, when you decide whether or not to take an umbrella to work, to the President. What do you really want to know about the system?

"And we need to deliver those products."

Eric Barron, Ph.D., is director of the Environmental Institute in the College of Earth and Mineral Sciences and distinguished professor of geoscience. He currently chairs the Board on Atmospheric Sciences and Climate of the National Research Council and is a member of the council's committees on Global Change Research.

Last Updated May 01, 2002