Incorporating uncertainty into predictive models of greenhouse gases emissions from cattle

Starting date : 01/11/2021. Funding: ABIES, INRAE PHASE

Modelling the production of greenhouse gases (GHG) from livestock production systems is central for the development of GHG national inventories and for the development of GHG mitigation strategies. The prediction of GHG is based on various data regarding individual animals at a variety of scales (daily time-step, average data), with multiple variables related to animal characteristics and production. The full process of data collection and model construction is subjected to uncertainties that impact the accuracy of the model predictions. This project will integrate uncertainty and time-scale (dynamic vs static) and space-scale (animal, farm) analysis to identify sustainable livestock production systems with low environmental footprint. The objective is to enhance the interpretation of GHG predictions, identify limitations and propose model improvements. To this end, Bayesian techniques and sensitivity analyses will be the core approaches to be deployed in the project.