News Release

This puzzle has been dubbed the "evaporation paradox." But a solution has surfaced.

Environmental engineers from the Johns Hopkins and Cornell universities presented their explanation in a paper published in the Nov. 5 issue of Nature. Marc B. Parlange (pictured at right) and Wilfried Brutsaert say that evaporation pan records do indeed support predictions made by the climate computer models that are used to simulate the impact of increased greenhouse gases. The paradox arose, they say, because many researchers have misinterpreted the evaporation measurements.

"The evaporation pans are a great tool to study climate," Parlange says. "But their data have to be used correctly."

The two scientists decided to unravel the paradox after Parlange, a professor in Hopkins' Department of Geography and Environmental Engineering, heard a Russian hydrology researcher refer to it during a symposium on water resources and global warming, held last spring at Hopkins. The Russian researcher had no clear explanation for some evaporation figures that did not seem to align with other climate change indicators. Parlange raised the issue with his mentor, Brutsaert, during a lunchtime run while visiting the Cornell campus in Ithaca, N.Y. Brutsaert (pictured at left), a professor of civil and environmental engineering at Cornell, had been Parlange's doctoral advisor.

To probe the paradox, the researchers focused on the Earth's hydrologic cycle, in which water falls to the ground as rain or snow, soaks into the soil or moves into bodies of water, then evaporates to form clouds that start the cycle once again. Over the past two or three decades, data collected in parts of the United States, the former Soviet Union, India and Venezuela have indicated that the pace of this hydrologic cycle is picking up. In other words, it has been taking less time to complete each cycle of precipitation through evaporation. Scientists disagree on the reason. It may be a natural phenomenon or one that humans have triggered by releasing more carbon dioxide. But the existence of this accelerated water cycle is now widely accepted.

Nevertheless, many climate researchers have been unable to explain why evaporation figures have decreased in areas where rainfall and cloudiness have become more prevalent. If less water vapor is moving into the atmosphere, how could more rain and clouds be forming?

"My suspicion," Parlange says, "is that some researchers have simply avoided talking about this. But it's been an awkward problem that has bothered scientists studying climate change."

To explain the paradox, Parlange and Brutsaert looked at the way evaporation is recorded. Weather stations throughout the world usually place water-filled metal pans, about a foot deep and three feet in diameter, outdoors on wooden platforms. Each day, a technician measures how much water has disappeared from the pan.

But Parlange and Brutsaert found that decreasing pan evaporation does not necessarily mean that less evaporation is occurring in the surrounding landscape. A key reason is that the interpretation of the pan measurements has not taken into account the role of humidity in the air or the moisture that is already present in or missing from the surrounding landscape. For example, a pan of water placed outside in a hot, dry desert would evaporate very quickly. But put the same pan in a cool rain forest, and the water would evaporate much more slowly. Thus, evaporation figures may drop in some areas merely because more rain and snow have saturated the terrain. This "regional land-surface moisture" must be factored in when climate researchers interpret raw measurements from evaporation pans, the researchers say.

When it is, the paradox disappears, and evaporation rates fall in line with other signs of climate change, the researchers say. In their Nature article, Parlange and Brutsaert assert that "in many situations, decreasing pan evaporation actually provides a strong indication of increasing terrestrial evaporation."

Related Web Sites Marc Parlange's Home Page http://www.jhu.edu/~dogee/parlange.html Wilfried Brutsaert's Home Page http://www.cee.cornell.edu/fac/Brutsaert.htm