“Weather & Climate Extremes and Machine Learning: Recent Development and Prospective”

Workshop organised by IMI in collaboration with CSIRO, held at the Department of Meteorology, Stockholm University, 19–20 June 2024.

Weather and Climate (W&C) extremes have serious impacts on the society, infrastructure and the environment. It has been observed that W&C extremes are increasing in intensity and frequency. These extremes are multivariate by nature, and therefore uni-dimensional methods commonly used in climate research come with weaknesses. Machine learning
(ML) is a relatively new method, which has been introduced recently in climate research.

This workshop will discuss recent development of multivariate extremes, e.g. archetypes, in combination with ML. We will bring together climate scientists and applied mathematicians/statisticians to discuss the development in the field with outlook for future directions.

Dates/times

Wednesday 19 June – Thursday 20 June.

Location

Rossbysalen

Department of Meteorology, Arrhenius Laboratories 16C, 6th floor, Frescati Campus

Programme

Wednedsday–Thursday

Wednesday 19 June 2024

09:00–09:30 Abdel Hannachi
Welcoming & programme

09:30–10:10 Gabriele Messori
Drivers and statistics of spatially remote, concurrent extratropical extremes

10:10–10:50 Torben Schmith
Benefits and drawbacks of different extreme value approaches

10:50–11:20 Coffee break

11:20–12:00 Kathrin Finke
Tropospheric response to stratospheric variability via lagged quantile regression

12:00–12:40 Frank Kwasniok
Linking large-scale atmospheric regimes with local weather variables using advanced clustering techniques

12:40–14:00 Lunch

14:00–14:40 Carlos Pires
Estimation of probabilistic copulas from nonlinear correlations and application to
teleconnections and atmospheric forecasting

14:40–15:20 Sebastian Mair
Representative Subsets for Archetypal Analysis

15:20–15:50 Coffee break

15:50–16:30 Didier Monselesan
On the representations of MJO and ENSO and their interactions

16:30–17:10 James Risbey
The identification of long-lived Southern Hemisphere flow events

17:10 Food + Fika + Discussion

Thursday 20 June 2024

9:30–10:10 Meriem Krouma
Assessment of the predictability of cold-wet-windy Pan Atlantic compound extremes

10:10–10:50 Zoubeida Bargaoui
Comparing drought identification using order statistics and archetypes based on remotely
sensed relative productivity index

10:50–11:20 Coffee break

11:20–12:00 Abdel Hannachi
AA, dynamical systems and the interior point algorithm

12:00–12:40 Dmitry Mukhin
Identification and analysis of mid-latitude atmospheric regimes with hidden Markov models

12:40–14:00 Lunch

14:00–14:40 Waheed Iqbal
Near-surface wind extremes over the Baltic and the North Sea

14:40–15:20 Vera Melinda Galfi
The typicality of weather and climate extreme events

15:20–15:50 Coffee break

14:40–15:20 Ludovico Giorgini
Advancements in modelling ENSO: From stochastic modelling to neural ordinary differential equations

15:20 Discussion

19:00 Dinner

Last updated: 2025-10-07

Source: MISU