Livestock emissions, vehiculated by manure, have a strong impact both on air quality and climate change.The first is related to the so-called ammonia cycle. According to this, manure yields ammonia (NH3) in the atmosphere, which reacts with atmospheric nitric and sulphuric acids to form up to 50% of primary particulate, PM2.5. For the second impact, livestock-mediated greenhouse gas (GHG) emissions are considered a sizeable causative agent of climate change, with up to 3.75 Gt CO2-equivalent emitted yearly.The PhD student will develop advanced hybrid modelling techniques merging geostatistics and deep neural networks to build a data-driven statistical model of the impact of livestock on particulate matters at the EU level.An essential intermediate output will be the publication of an open access (FAIR) dataset harmonising all data entering the model and including air quality (EEA), Meteorology (ECMWF, ERA5) livestock emissions (Copernicus), land cover and land use (Copernicus). Harmonisation will be faced by change of support and data fusion statistical techniques.The model will follow a multiscale approach, able to provide small-scale impact maps and used to test various climate change and mitigation scenarios at the local and European levels.The challenges related to the large size of the European data set, will be faced using high dimensional statistical models, advanced computational statistics, numerical optimisation techniques and high performance computing.

The ideal candidate for this project is a student with a master's degree in statistical science, computer science, environmental engineering, environmental sciences or physics.In the first 12-18 months, the training (in Bergamo, Pavia and abroad) will focus on the following:- Atmospheric sciences for climate change and air quality dynamics- Databases and coding- Numerical optimisation- Frequentist and Bayesian statistical theory- Computational statistics- Advanced geostatistical models for large spatiotemporal data- Machine learning for spatiotemporal data- Advanced computational statistics and high-performance computing.This training will be based on classes and learning by doing.

The research project will be developed in close connection with the research group on environmental statistics at the Department of Economics of the University of Bergamo. The group is composed of prof. Alessandro Fassò, prof. Francesco Finazzi, prof. Michela Cameletti, Dr. Rodlfo Metulini, Jacopo Rodeschini (PhD student), and Alessandro Fusta Moro (Research assistant). The PhD student will also collaborate with the Agrimonia network (www.agrimonia.net), including environmental statisticians from universities in Bergamo, Milano Bicocca, Torino, and Leibniz University Hannover.At the department, the PhD student will be provided with a desktop position, computing facilities, library etc.