Description du sujet de thèse

• Scientific background :
Gradual anthropogenic warming and parallel changes in the major global biogeochemical cycles are slowly pushing forest ecosystems into novel growing conditions, with uncertain consequences for ecosystem dynamics and climate. Short-term forest responses to warming, drought and increased atmospheric CO2 and nitrogen deposition are relatively well understood and skilfully simulated by process models. However, process-based understanding and confidence on model projections of carbon and water cycles weaken towards longer time scales and to the future, mainly because long-term observations of key processes by which long-lived trees die or adapt to the new growing conditions are lacking. In particular, long-term belowground forest responses are shrouded in mystery. Even more easily observable aboveground forest growth trends, such as the so-called tree-ring divergence from warming at cool northern latitudes or the apparent absence of sustained CO2 fertilisation of growth widely predicted by ecosystem models are difficult to explain with current field observations and experiments.

The utilisation of tree-ring width, combined with their carbon (δ13C) and oxygen (δ18O) stable isotope composition, is increasingly recognized as a method to address the lack of direct observations on century-scale alterations in plant physiology and growth with global change. This approach can be combined with readily available short-term satellite and eddy-covariance observations of forest growth and carbon and water fluxes over the past few decades to provide a consistent long-term benchmark for land surface models, a resource that is just beginning to be exploited. The ERC-funded project CATES is pioneering the field by developing a new cross-disciplinary framework to constrain climate projections by jointly improving the simulation of forest growth and water use efficiency (WUE; the ratio of photosynthesis to transpiration) at long time scales (decades to century) using novel observational standards for historical growth and physiology derived from tree-ring data.

• Objective of the thesis :
The thesis is part of the CATES project and aims to introduce a process-based representation of xylogenesis or wood formation in the ORCHIDEE global land surface model to improve the simulation of biomass growth through the assimilation of readily available tree-ring width and density data. This work will contribute to the overall CATES scientific objective of reducing uncertainties in the simulation of feedbacks of modified tree growth and physiology on the terrestrial system.

• Requirements :
• Master degree (M2 in France).
• Demonstrated interest in Biogeosciences and numerical modelling.
• Previous experience in the development of xylogenesis models.
• Proven experience with the ORCHIDEE model or another land surface model.
• Analytical thinking, scientific rigor, and autonomy.
• Strong computing, programming and English writing skills.

• Approach :
Tree growth results from the formation of xylem tissue (xylogenesis), a fundamental physiological process that is not currently represented in global land surface models such as ORCHIDEE, despite the existence of specialized process-based models. Integrating xylogenesis into ORCHIDEE is critical to improve the representation of carbon allocation in trees and to enable the use of long-term observational data, such as tree-ring series, in model evaluation and data assimilation. The scientific objective of this thesis is to implement a process-based representation of xylogenesis in ORCHIDEE and to constrain simulations of forest biomass growth from seasonal to centennial timescales using tree-ring data. To achieve this, the project will employ a Bayesian data assimilation (DA) framework to combine process-based modelling with site-level observations, including tree-ring measurements and carbon fluxes. While this DA system has been used with eddy covariance data, incorporating xylogenesis introduces new challenges. Technically, the task involves embedding the biomass formation process into the carbon allocation scheme of ORCHIDEE and calibrating its parameters using tree-ring width and density data, alongside conventional short-term eddy-covariance data. This will allow rigorous validation of model performance and its environmental responses under past and ongoing global change. Most tree-ring data are already available, but the candidate will contribute to field sampling and measurement at specialized experimental sites.

• Expected results :
The model development and data assimilation processes should result in a calibrated model that is able to realistically simulate short-term forest responses to climate extremes (e.g., drought, warming) and long-term changes in biomass growth over the 20th century for different forest types.

Contexte de travail

The workplace will be the LSCE (www.lsce.ipsl.fr), a world-class institute on climate change research and Earth System modelling located on the plateau of Saclay, near Paris. It employs over 320 researchers from more than 30 different nationalities and diverse scientific disciplines. The student will be part of the MOSAIC (Modelling of Continental Surfaces and Interfaces) team in the LSCE and will be enrolled in the Doctoral School of Environmental Sciences of Île-de-France (SEIF) at the University of Paris-Saclay. This is a PhD program by Research and it benefits from a very dynamic research environment and world-class modelling infrastructure.

Contraintes et risques

Applications should be submitted through the CNRS job portal (please follow this hyperlink: https://emploi.cnrs.fr/) including (1) a curriculum vitae, (2) short statement of motivation (½ page) and (3) names, addresses, phone numbers, and email addresses of at least two references. More information about the CATES project is available here: https://jbarichivich.github.io/erc-cates.html