New article published on soil organic carbon estimation at multiple scales

A new open-access paper in Geoderma uses soil organic carbon (SOC) data from Hungary sampled in 1992 and 2010 to estimate SOC stock change at multiple spatial scales. The paper, Estimating soil organic carbon stock change at multiple scales using machine learning and multivariate geostatistics, estimates soil organic carbon stock change at scales that range from point level to the entire country.

At the country scale, the topsoil SOC stock was predicted to have increased with 14.9 teragrams (Tg), about 3% between 1992 and 2010. The increase is likely caused by the substantial afforestation programs that Hungary implemented during the past three decades.

“This may well be the solution to the quest for accurate estimation of soil organic carbon change.” Gerard Heuvelink

Authors include ISRIC senior researcher Gerard Heuvelink with lead author Gábor Szatmári  and László Pásztor (Institute for Soil Sciences, Centre for Agricultural Research, Hungary).

“Maps of soil organic carbon change over time are often too uncertain to detect significant changes, but this study nicely shows that uncertainties decrease dramatically when predictions are averaged over larger areas.” said Heuvelink. “This may well be the solution to the quest for accurate estimation of soil organic carbon change.”

Access the paper here: https://doi.org/10.1016/j.geoderma.2021.115356

Full citation:

Gábor Szatmári, László Pásztor,  Gerard B.M.Heuvelink. 2021. Estimating soil organic carbon stock change at multiple scales using machine learning and multivariate geostatistics. Geoderma. https://doi.org/10.1016/j.geoderma.2021.115356