Climate change requires that we better understand coming changes to the planet. Since climate is the

result of complex interactions between the atmosphere, oceans, polar ice, and life, it is very challenging to

predict future conditions, including shifting extremes in variability or fluctuations of climate. A fundamental diculty is that the solar radiation drives the climate system far away from thermal equilibrium, into a regime where the well established principles of thermodynamics developed in the 19th century cannot be reliably applied. Non-equilibrium statistical mechanics is a rather young branch of theoretical physics that aims to identify overarching principles governing the behavior of such strongly driven systems, an endeavor that has been remarkably successful in recent years. The statistical analysis of long time series, such as records of temperature or ooding, has a longer history but new mathematical ideas are now transforming the eld. So far there has been only limited interactions between researchers in climate science, non-equilibrium statistical mechanics, and time series analysis, despite important implications of the work in each of the three elds for the other. In the workshop and seminar held at MPIPKS, senior and junior scientists in all three elds engaged in dialogues that benetted from cross-fertilization.

55 participants from 11 countries attended, 19 of which were at the graduate student or early postdoc level. Invited speakers represented the three areas, including Petra Friederichs (Bonn, DE), Christopher Jarzynski (Maryland, USA), Udo Seifert (Stuttgart, DE), and Cecile Penland (NOAA, USA). The Workshop program comprised 19 invited and 24 contributed talks as well as 8 poster presentations, and 24 lectures were scheduled during three weeks of the Seminar. All the young scientists attending the Seminar were given the opportunity to deliver hour-long presentations of their work, thus allowing for detailed feedback from the more senior participants. This emerging generation of scientists will be the ones to bring the elds of climate science, statistical mechanics, and time-series analysis together.

The program was very successful in establishing new contacts between participants with dierent backgrounds and identifying key questions for future interdisciplinary work. Examples include the application of principles of nonequilibrium statistical mechanics to the optimization of subgrid parametrizations of climate models, and the use of dynamical models of intermediate complexity to provide a physical underpinning of time series analysis.

A public lecture, entitled \Neue Daten aus der Klimaforschung: Bekommen wir die Klimakrise noch in den

Gri?" was given by Prof. Stefan Rahmstorf (Potsdam, DE) to a full auditorium on 18 July. The many

questions that Prof. Rahmstorf received at the end of the talk was one indication of the keen interest by the

audience in the ideas that he presented.