Understand and use the estimation of soil organic carbon persistence by Rock-Eval® thermal analysis

Soutenance de Thèse, aussi en visio https://zoom.us/j/4838967199 Abstract: One of the most important solutions to climate change lies literally right under our feet. Soils store twice the amount of carbon that is found in atmosphere and vegetation combined. They act as a buffer between solid earth and atmosphere and exercise a major control on the atmospheric concentration of CO2 through the release or sink of greenhouse gases. Moreover, organic carbon in soils in the form of organic matter is essential to soil health and fertility, to nutrient availability and water quality. My work is centred around the most valuable tool at our disposal for understanding and predicting the evolution of this reservoir in the future: soil organic carbon (SOC) dynamics models. A missing key influencing the accuracy of SOC model projections and a major challenge in soil science is our ability to estimate the proportion of SOC that will remain unchanged over projection-relevant timescales. This important amount of carbon that has been present in soils for centuries or millennia, and is therefore considered to be “stable”, can vary greatly from one location to another. The goal of my thesis project was to explore a new approach based on thermal analysis of SOC and machine learning, to characterise SOC, estimate the proportion of “stable” carbon in soil samples, and eventually use this information to improve the accuracy of SOC dynamics models. In a second step, I focused on the Rock-Eval® thermal analysis technique in the heart of this approach to understand better the important information it offers, based on model laboratory experiments. The main results of my thesis consist, on the one hand, of a complete and validated operational approach improving the accuracy of SOC models with a clear and significant value for “climate-smart” soil management. On the other hand, an experimental part offers new insights into the working principle, limitations and possibilities of the Rock-Eval® thermal analysis technique. Jury: Daniel RASSE (NIBIO): Rapporteur / Eric VERRECCHIA (Université de Lausanne): Rapporteur / Claire CHENU (AgroParisTech, INRAe): Examinatrice / Josette GARNIER (CNRS): Examinatrice / Thomas EGLIN (ADEME): Invité / François BAUDIN (Sorbonne Université): Directeur de thèse / Pierre BARRÉ (LMDENS): Co-directeur de thèse / Lauric CÉCILLON (INRAe): Co-directeur de thèse