Bonanza Creek Experimental Forest, where Fan and colleagues calibrated their high-latitude peatlands model.
Climate change is not an equal opportunity threat. Certain areas of the world are more susceptible than others to the shifts in temperature, precipitation and other effects climate scientists predict to occur over the next century. Even when these vulnerable regions are remote and rarely visited by humans, how climate change affects these ecosystems could have dramatic consequences for the global population.
One such area of vulnerability is the high-latitude peatlands, the bogs, fens and boreal forests found in the northern parts of Canada, Alaska and Russia. Much of this frigid land carries a layer of permafrost year-round, but the vegetation and soil there accounts for the majority of Earth’s biomass and carbon storage – roughly double that of the much better known tropical forests. Climate change is also expected to be particularly dramatic in these areas, with some models predicting temperatures will increase by as much as 7.5 degrees Celsius over the next hundred years.
Because of this vulnerability, Argonne computer scientist Zhaosheng Fan is focusing his attention specifically on how climate will affect this unique ecosystem. In his talk at the Computation Institute on February 14th, Fan (Assistant Biogeochemical Modeler in the Biosciences Division at ANL) described how scientists created a model of these high-latitude peatlands, how they refined the model based on field experiments and the very serious warnings the model produced about how these remote areas might someday affect the rest of the world.
Fan and his colleagues built their model off of the Terrestrial Ecosystem Model, initially developed by researchers at the Marine Biological Laboratory in Woods Hole and later modified by researchers at the University of Alaska at Fairbanks for high-latitude regions. The peatland model adds features specific to this biome, including the unique structure and composition of the soil, the level of the water table and the sensitivity of the ecosystem to fires. Researchers were especially interested in the production of carbon dioxide and methane within the soil, and how much of those gases were released into the atmosphere under different weather conditions.
Constructing the model was the first step, but the researchers needed to check that their simulation actually lined up with reality. So a field experiment was set up in the peatlands of the colorfully-named Bonanza Creek Experimental Forest, located 20 miles southwest of Fairbanks, Alaska and full of fens and bogs. There, three controlled sections of peatlands could be observed: a control area and areas with experimenter-manipulated low or high water tables.
Scientists took long-term measurements of methane and carbon dioxide flux at the sites, and compared the results to what the model would have predicted given the weather conditions. When the predictions and measurements didn’t align, the model was recalibrated to produce results closer to reality.
After that process, the researchers could move from modeling the present to simulating the future. Three different emission scenarios from two global climate models were chosen to provide predicted weather conditions for the rest of the 21st century and fed into the peatlands model. For the first 50 years, the model predicts business as usual – as the Earth warms, photosynthesis increases, and the more carbon is sequestered inside the soil, providing something of a brake on the greenhouse effect. But somewhere around 2060, the model predicts that the peatlands will flip from being a carbon sink to being a carbon source, actually putting more carbon out into the atmosphere and potentially accelerating climate change in a positive feedback loop.
While uncertainty remains about the model’s predictions – and some important details, such as root dynamics and soil chemistry, are not yet built in – the results offer a warning about an under-the-radar consequence of climate change. While the predictions of most climate modeling are reported in terms of global impact, Fan’s model offers insight about a more localized effect…which nonetheless could ripple outward to affect the rest of the world.
