Experiments in the permafrost zone near Denali simulate a warmer North.

Struggling to keep my balance, I teeter along a narrow plank way that wends through the rolling foothills near Denali National Park and Preserve. Just ahead, Northern Arizona University ecologist Ted Schuur, a lanky 6-footer, leads the way to Eight Mile Lake, his research field site since 2003. Occasionally I slip off the planks onto the squishy vegetative carpet below. The feathery mosses, sedge and diminutive shrubs that grow here — Labrador tea, low bush cranberry, bog rosemary — are well-adapted to wet, acidic soils.

The objective is to keep a running tally of CO2 as it’s inhaled and exhaled by plants and soil microbes, but not merely in the here and now. By artificially warming selected patches of tundra, Schuur open-air experiment aims to mimic the future, when air temperatures in Alaska are expected to be significantly higher. By 2100, the state is projected to see an additional warming of at least 4 to 5 degrees Fahrenheit over what’s already occurred, and that’s under the most optimistic scenario. Already, the tundra here is leaking carbon dioxide to the atmosphere, according to recent satellite-based measurements. The question Schuur is hoping to answer: As the region continues to warm, just how much more carbon dioxide will it contribute to the global pool?

Rounding the top of a knoll, we look down on an expanse of tundra that bristles with so many sensors and cables that it resembles an outdoor ICU ward. At the center of the site stands a gas-sensing tower that sniffs out the carbon dioxide drifting through the air from as far away as a quarter of a mile. At ground level, poly carbonate chambers placed atop the tundra whoosh as their tops periodically shut, then open, then shut again. Their job, I learn, is to trap the carbon dioxide rising from the surface and shunt it to an instrument that measures the amount.

Along with terrestrial and aquatic plants, the soil microbes that decompose organic matter are major players in the global carbon cycle. In the lingo of climate science, plants are “sinks” for carbon. Through the sunlight-driven process of photosynthesis, they lock up more carbon dioxide than they release, thus keeping it out of the atmosphere. By contrast, soil microbes that decompose organic matter are “sources” that burp out micro-bubbles of CO2 night and day, winter and summer.

The impacts of this manipulation are many. Triggered by the extra warmth, subsidence caused by slumping permafrost has lowered the surface of the experimental plots by several feet. The depth to which the soil thaws at the end of summer has likewise increased, indicating that the top layer of what used to be permafrost has added more organic matter to the microbial dining table.

Schuur draws my attention to the stack of drift-catching snow fences that, come October, researchers will array around half a dozen experimental plots, then laboriously remove again in April. Snow is an excellent insulator, he explains: “It’s like a giant blanket.” Beneath the drifts, Schuur and his colleagues have found, the ground can stay a good 3 to 4 degrees Fahrenheit warmer than it does in the unenforced control plots, thereby accelerating the warming that occurs in spring.

Given the present rate of temperature rise, the imbalance between plant uptake and microbial release of CO2 may well grow. By the end of the century, Schuur says, the amount of carbon the world’s permafrost zone transfers to the atmosphere each year could be in the range of 1 billion tons, comparable to the present-day emissions of Germany or Japan.

How much of the carbon coming out of permafrost will be transformed into methane? That’s another question Schuur is starting to tackle, for while methane is less abundant than CO2, it has 30 times the heat-trapping power over the course of a century. On the way back to the car, Schuur points out a clump of cotton grass whose partly hollow stems pipe methane into the atmosphere. “What matters is not that carbon goes in and out,” he says. “The important question is, what’s the net effect?”

Still unaccounted for, though, is the significant amount of carbon that appears to have vanished from underlying soils — about 20 times the amount Schuur and his colleagues have detected in the air. “Wow,” Schuur remembers saying to himself when he realized the size of the discrepancy. “This is a surprise.” Perhaps water seeping down slope is ferrying the missing carbon into streams, rivers and lakes, including Eight Mile Lake, or shunting it to swampy, oxygen-poor pockets of soil ruled by microbes that convert carbon to methane.