M2 internship: Multi-scale relocation issues

Background

Lopez-Ridaura et al (2005) propose a formal framework, based on linear programming, for analyzing trade-offs on multi-scale and multi-criteria problems. The method consists in optimizing a given indicator at a given scale under the constraints of other indicators. We then look at what we gain and what we lose on each criterion by changing the constraints or the optimized indicator. The method can be used :

• For a given scale: for example, 1) we seek to maximize the value of regional agricultural production without constraints, 2) we maximize it under the constraint of regional food autonomy, 3) we look at what we have gained and lost.
• To analyze the trade-offs between scales (what interests us most here): for example, we maximize the value of regional production by setting a food autonomy constraint first at the regional level, then sub-regional, then municipal (thus increasingly constraining) and we look at what we lose at the global level because of this local constraint.

Objectives

The aim of the training course is to try to transpose this method to the problems, tools and data developed by STEEP to analyze production chains. The main question is to identify what kind of hypotheses are needed for this. One can think for example of the question "what scale of relocation for the flour mills" or "for food autonomy" (in raw materials? in processed products?) or "for the methanizers" or "for the manufacture of ball bearings". If the framework referred to is mathematical, the central question of the internship will rather call upon other disciplines (in particular economics). The final objective is to validate or not this analytical framework as the main research avenue to pose the problem of relevant relocation scales.