Statistical science for understanding climate
and the Earth system


Brief description

The summer school will start with a few introductory lectures about the climate change problem, to give students a succinct overview of the main threads of climate change research, its latest headline findings and its most critical open questions. We will organize these introductory lectures according to the structure of the Intergovernmental Panel on Climate Change assessment reports, talking about  the physical  sciences;  impacts, adaptation and vulnerability; and  mitigation choices.

 These introductory lectures aim at presenting the climate change problem in all its complexity. In particular, it will be clear by the end of these lectures that the climate change problem is not only about changes in the physical climate system. In fact, the human (socio-economic) dimension of the problem can be in many cases even more critical.  The introductory lectures will discuss this aspect and touch briefly upon some of the methods relevant to studying impacts to human systems (econometric modeling, dose-response functions, causal inference, for example), but we will not make these the focus of our technical lectures. Rather, our course will focus on methods, example problems and datasets relevant to the study of Earth’s physical climate.

Studying the Earth’s climate often involves large and complex data sets collected over space and evolving over time. These data can be traditional observations made from weather stations or satellite instruments or the output from numerical (computer) simulations of the Earth’s climate system. A basic statistical element in this area are spatial fields for a variety of variables. These can be as simple as temperature at the surface but could be more complex measurements such as vegetation types or the concentrations of air pollutants. The challenge is to determine the structure, i.e. spatial dependence, in these fields and to predict at locations that are unobserved. This school will develop the statistical methods connected to Gaussian processes with an emphasis on applying these models to environmental data. Lectures will be paired with hands-on data analysis in R and there will also be the opportunity to work on a more extensive group project that tackles a climate related question.

In a broader context this school is about statistical methods to infer curves and surfaces from noisy and perhaps indirect measurements. Although we focus on geophysical applications, this school will develop function fitting methods that are general such as splines, Gaussian process regression, and the foundations of solving inverse problems.  The school will also introduce the students to some advanced statistical tools including sparse matrix methods for large linear systems, neural networks for functional approximation, and Monte Carlo methods for Bayesian computation.


Lectures: The school starts on Monday July 8 at 2pm, and ends on July 18 after lunch.
A detailed schedule will be made available in due time. 

Morning: 3-4 hours/day lectures
Afternoon: 2-3 hours/day supervised tutorials as well as individual and team work.

Moreover, there will be a poster session, where participants, upon previous request, may present their research. A welcome cocktail will be offered during the poster session. More detailed info to follow.

Room and board

The school is residential. Accommodation is included in the registration fees — please see more detailed info at
The organizing committee will take care of the reservation.
Working days’ lunches are included in the registration fees.

Attendance and final certificate

Full attendance of the activities of the summer school is mandatory for the participants.
Subject to a positive participation to the program, an attendance certificate will be awarded by Università Bocconi, mentioning that the 2024 edition of the Summer School is offered in collaboration with University of Oxford and Imperial College London.