Beyond fractions and pools: Radiocarbon distributions as a constraint for soil organic carbon models

Laboratory quantification and computer modeling of soil organic carbon (SOC) and radiocarbon (¹⁴C) rarely align well. Diverse lab methods exist to separate SOC into “fractions” with operationally defined boundaries which include physical, chemical, and biological thresholds. Compartmental soil models, on the other hand, use homogeneous “pools” to group SOC by its rate of decay. However, the stochastic nature of organic matter decomposition makes it virtually impossible to isolate model pools through fractionation or to model fractions as pools. Radiocarbon is a powerful metric that integrates C fluxes in and out of a system, but fraction and pool ¹⁴C valuesrepresent an average which may be composed of widely different C ages. In order to advance and constrain our understanding of SOC dynamics, we need to go beyond mean ¹⁴C values of operationally defined pools and instead use radiocarbon distributions. Thermal (oxidative) fractionation of SOC produces continuous C data as a function of temperature (20˚C to 900˚C), and discrete ¹⁴C measurements by collecting evolved gas. By fitting splines to ¹⁴C data and weighing by C release over temperature, a continuous mass-weighted distribution of ¹⁴C can be estimated. Recent modeling advances can similarly estimate system ¹⁴C distributions, and models may thus be constrained for the first time by continuous lab data. This convergence is a first step beyond pools and fractions that may significantly increase constraining power. We will present lab and mathematical methods, limitations, and outlooks for advancing soil SOC modeling.