Loup Petitjean (MSc) joins us for 12 months

Biophysical models are great tools to predict the complex balance of competitions and facilitations processes and their impacts on agroforestry systems performances.

For example, the Hi-sAfe biophysical model predicts crop and tree yield as well as some environmental performances such as carbon storage or nitrogen safety net, as a function of soil, climate, associated species and management practices.
However, this model is very difficult to parameterize for a given situation and takes a long time to run, making it a good research model, but a poor help for farmers to design efficient agroforestry systems.

During his 1-year research contract, Loup Petitjean will strive to solve this issue, by developping a meta-model of Hi-sAfe specifically adapted to quickly answer questions such as “given my soil and climate, what is the maximum tree density that I should plant to strike a good balance between tree and crop productivity?”.
In order to do so, he will build a statistical model based on simulations of the initial model, so that the metamodel is less detailed than the initial model, but gives quicker answer to a more limited range of questions.