Computer model slices and dices mountain climates

  • Chris Daly at a weather sensor in old-growth Douglas fir in the H.J. Andrews Experimental Forest in Oregon.

    Thomas Nash


On the face of a wind-swept cliff ... At the bottom of a frost-prone hollow ... Beneath the canopy of an old-growth tree ...

Oregon State University climatologist Chris Daly and his team have positioned their instruments in some oddball places here in central Oregon’s H.J. Andrews Experimental Forest. “The World Meteorological Organization would have a fit,” Daly says with a grin. And indeed, the readings he’s obtaining don’t always jibe with those taken by standardized weather stations. That’s just the point, he says.

Already, readings from this forest have helped fine-tune a pioneering computer model called PRISM, which starts with data from widely spaced weather stations, then uses topography as a guide to calculate temperature and precipitation for places in between. A standard workhorse for ecological modelers, PRISM (Parameter-Elevation Regressions on Independent Slopes Model) is regularly used to reconstruct climate patterns in complex terrain and generate maps that are used by a variety of researchers. PRISM will soon make its public debut as the tool behind the U.S. Department of Agriculture’s first revision of plant hardiness zones since 1990. And climate modelers use PRISM maps to increase the detail of their future projections.

PRISM’s power stems from the fact that it is a good 50 times more fine-grained than most regional climate models. While the latter generate data points spaced tens of miles apart, PRISM plots the landscape as a grid of boxes that measure just a half-mile on each side. So PRISM picks up many details that coarser models miss, like the rain shadow on the leeward slope of a moderately high ridge and the effects of the cool fingers of fog that find their way through low spots in California’s coastal mountains.

PRISM also quantifies the temperature inversions so typical of mountain valleys, including those of the Andrews forest. Topping out at 5,350 feet, this 15,800-acre reserve sits in the range of highly eroded volcanoes known as the Old Cascades. It is the epitome of craggy. Strip away the majestic Douglas firs and western hemlocks, and the place would be a Mordor-like heap of basalt and breccia, riven by narrow, steep-sided valleys. In winter, the valleys here are so dark that they sometimes get no more than an hour of direct sunlight a day. Unless a storm passes through to stir up the air, a cold-air pool that sets up at night can linger for weeks. As a result, the valleys are often more than 25 degrees F colder than the overtopping ridges.

Daly started working on PRISM in 1991, while he was an Oregon State graduate student, and he continues to make improvements in the system. Currently, he’s in the process of pushing PRISM to even higher resolution: Soon, PRISM should be able to zoom in on areas as small as 900 square feet, the size of a modest house. At that scale, precipitation does not vary that much, but temperature can exhibit stunning swings. As Daly sees it, “The more complex your landscape, the more options you as an organism have. So, if you’re a person driving in a car with the windows open, you’ll notice when you dip in and out of the cold air pooling in a valley. And if you’re a little bug that spends its days on one side of a rock, then going around to the other side could prove to be a complete revelation.”


Dancing with Climate Change

In California's White Mountains, scientists ponder the fate of beardtongues, bristlecone pines and butterflies in a rapidly warming world.