A new statistical study by scientists shows large mismatches in models predicting how carbon moves in and out of soil.

The study, published in the peer-reviewed journal Nature Communications on April 1, attempted to reconcile different measurements of the two main movers of carbon through the soil. Researchers from Virginia Tech University and the Pacific Northwest National Laboratory conducted the research, and both institutions funded the research.

To understand the atmosphere, scientists treat soil carbon like a bank account. Carbon is deposited by plants in the soil in the form of organic carbon matter. The study refers to the process by the term gross primary productivity (GPP). Bacteria, roots and other denizens of the soil biome withdraw carbon dioxide (CO2) into the atmosphere via soil respiration (R_s).

Other, smaller sources of CO2 like fire and dissolving organic carbon in water, also make up ways in which carbon moves out of soil.

Understanding how carbon moves through the soil and how long it stays there could have huge implications for estimating the true environmental value of carbon credits. Some companies and environmental groups have pushed carbon credits as a means to potentially reduce CO2 emissions and off-set more harmful gasses like methane and nitrous oxide. 

Moving carbon from the atmosphere to the soil could in turn fight climate change, credit supporters say.

However, other groups have derided carbon credits as "greenwashing." 

GPP and R_s are directly connected, but frequently measured in different ways. GPP is typically estimated using satellites, while R_s is typically measured by sampling a small geographic area and then scaling up.

Because the global values are enormous, scientists use the measure petagrams (Pg) to discuss changes. One petagram is equal to 1 billion metric tonnes, or about 1.1 billion tons.

Researchers used randomly selected values to get 10,000 potential R_s values drawn from scientific literature, and then estimated GPP values that would result.

The result was a hypothesized GPP range of 126-178 Pg. Measurements of GPP based on satellite information furnish a value of 95-131 Pg.

While some overlap exists, given that the ranges are generated using statistics, overlap between the projected range and the measured range was extremely unlikely.

When researchers attempted the opposite estimate, they came away with a hypothesized range for R_s of 60-78, far less than the up-scaled value of 79-96 Pg.

"Either the amount of carbon coming out of the atmosphere from the plants is wrong or the amount coming out of the soil is wrong," says Meredith Steele, an assistant professor at Virginia Tech's School of Plant and Environmental Sciences, in a statement.

Steele's Ph.D. student, Jinshi Jian, led the research team.

Click here for more No-Till News.