Rechercher
Fermer ce champ de recherche.

TRAINING SCHOOL - ATTRIBUTING IMPACTS OF CLIMATE CHANGE (I2C): CHALLENGES, METHODS AND PERSPECTIVES

Dates: 26-31 May 2024

Location: Maison Clément, Les Plantiers, France (Maps)

Format: in person

 

Contact: xaidaproject@gmail.com

ABSTRACT

A large number of extreme weather and climate events have occurred recently in Europe and around the world, causing damage to infrastructure and loss of life, particularly in developing countries. This has raised the question of the role of climate change in altering the likelihood or magnitude of a number of these events. The science of attributing extreme events has been? developed to answer these questions. With the many advances in this field, it is now possible to go beyond the attribution of climatic events and look at their impacts as well. This is one of the major challenges facing the attribution sciences. At the same time, the development of data processing techniques based on artificial intelligence algorithms is opening up the prospect of new techniques that could help tackle this challenge.

The school provides an introduction to some various aspects of attribution of extreme weather events and their impacts to climate change, including perspectives brought into the field by advances in machine learning. In addition to morning lectures, the afternoons are dedicated to student projects. There are 5 student projects;  each student works on the same project in groups for the duration of the school.

 

 

 

Key words
  • attribution

  • extreme events

  • impacts

  • machine learning

Welcoming 35 participants, including 22 doctoral and post-doctoral students, the event blended interdisciplinarity with innovative techniques and working methods to introduce everyone to the various aspects of attributing extreme weather events and their impacts to climate change.

 

The school enabled students to work on projects using innovative techniques and methods. Thus, each project had its own specificities, processes and domains:
– Project 1: Conditional attribution with storyline
– Project 2: Compound event analysis, bivariate statistical analysis of copulas
– Project 3: Crossing social and urban climate data
– Project 4: legal co-construction
– Project 5: conditional attribution by analogous.


This variety also encouraged students to take an interest in the work of each group, thus promoting exchanges.

Speakers
  • Mariana de Brito (UFZ) – Text mining in climate extremes research- from impacts to adaptation. FIND THE PRESENTATION HERE
  • Sabine Undorf (PIK) – Attributing the impacts of climate change. FIND THE PRESENTATION HERE
  • Rupert Stuart Smith (Univ. Oxford) – Climate science in court: leveraging scientific insight for legal accountability.
  • Samuel Rufat (Univ. Cergy)Social vulnerability and exposure to hazards and climate change impacts
  • Emmanuel Rouges (Univ. Reading)Energy meteorology: The challenges faced by the energy sector with increased renewable generation. FIND THE PRESENTATION HERE
  • Andreia Ribeiro (UFZ)Compound events and agriculture in a changing climate.
  • Aglaé Jézéquel (LMD-IPSL, CNRS) – Potential pitfalls of extreme events attribution applied to impacts FIND THE PRESENTATION HERE
Scientific Committee
  • Davide FARANDA (LSCE-IPSL, CNRS)
  • Tamara HAPPE (IVM, VU Amsterdam)
  • Aglaé JEZEQUEL (LMD-IPSL, CNRS)
  • Marlene KRETSCHMER (Uni. Leipzig)
  • Nora LINSCHEID (MPG Biogeochemistry)
  • Robin NOYELLE (LSCE-IPSL)
  • Dominik SCHUMACHER (ETH Zurich)
  • Pascal YIOU (LSCE-IPSL, CEA)

Organisation

Manon ROUSSELLE (IPSL)

Apolline SAUVIGNET (IPSL)

THE PROJECTS 

Find the pdf document here: Training School Project


Project 1: Exploring winter energy extremes using weather regimes 

 

Supervisor: Emmanuel Rouges 


Limiting climate change requires the transition to an energy network with a higher proportion of renewable generation. Renewable energy sources such as wind and solar are highly dependent on weather conditions. Combined with energy demand which partly relies on temperature, this signifies that balancing energy demand and supply will be increasingly weather dependent. Therefore, multiple studies have investigated the dependence of periods of high energy demand and low renewable generation on weather, and more specifically on weather regimes. Weather regimes are large-scale atmospheric patterns representing most of the low-frequency variability in the mid-latitudes, which impact energy related variables such as temperature, wind and incoming solar radiation. The winter of 1962-1963 is known as the coldest winter in Europe of the 20th century. It was characterized by a very high frequency of blocking regimes such as the negative North Atlantic Oscillation. Following this observation, within this project the relationship between weather regimes and winter energy conditions is investigated. This is done by setting the proportion of weather regimes during a winter season. Using 40+ years of modelled energy data for European countries and weather regimes, it is possible to create multiple iterations of winters. This allows to determine the relative influence of weather regimes on energy but also to identify potential worst-case scenarios. 


Project 2: Using global gridded crop model projections to identify changes in the relationship between climate and crop yields

 

Supervisor: Dominik Schumacher, Andreia Ribeiro & Raed Hamed

 

Agriculture is strongly influenced by extreme climate events, in particular by compound heat and drought during crop growing seasons. Climate change is expected to exacerbate agricultural impacts from such compound extreme events, varying across regions and crop varieties. In this project we aim to analyze future changes of climate-related agricultural impacts using a suite of global gridded crop models under different climate scenarios. We will explore the potential of copula theory to analyze the dependence structure between climate variables relevant for agriculture and crop yield simulations. The properties of copulas enable the robust estimation of tail dependencies and conditional probabilities of one extreme event given the occurrence of another extreme event. Initially, our focus will be on examining the dependencies of agricultural outcomes on individual factors such as dryness and heat using bivariate joint distributions, as well as exploring the dependence between extreme heat and dryness. Further work expanding the analysis to characterize the compound effects on agriculture is contingent on project progress during the training school. Ultimately, we aim to quantify how changes in climate will alter climate–crop relationships and how it will affect probabilities of crop failure in the future.

Preferred programming languages: R or python.


Project 3: Crossing urban climate models with social vulnerability in a French city: who is most at-risk during heatwaves and what changes becauses of climate change? 

 

Supervisor: Aglaé Jézéquel & Samuel Rufat

 

Heatwaves are getting more frequent, more intense, and longer in the context of climate change. They also have impacts on populations, including excess mortality. Densely populated areas are most at-risk, because the urban heat island effect amplifies extreme temperatures, especially at night. There are however heterogeneities in urban heat island effect within cities, for example related to the type of buildings,
their densities, presence or absence of vegetation. State-of-the-art urban climate models developed in the provide project (https://climateriskdashboard.climateanalytics.org/impacts/explore) simulate urban climate at a 100m resolution, and provide an ensemble of variables related to heatwaves. Mapping these variables can help to identify inhomogeneities and the most exposed neighborhoods in a given city for different scenarios of greenhouse gases emissions. Heatwaves also affect people differently based on their own vulnerability. Studies have shown that factors such as age and gender have an influence on the reaction to extreme heat. The habitation type and the income level also affect people’s ability to protect themselves from the heat. The French National Institute of Statistics and economic studies (INSEE) provides vulnerability indices (e.g. density of population, age, age of the buildings, number of low-income houses) at a 200m
resolution over France. The goal of this student project is to cross social vulnerability and urban heat island indices to identify the social characteristics of the neighborhoods with highest exposition to the heat
for different climate change scenarios.


Project 4: Practical attribution for a legal case: developing the scientific basis

 

Supervisor: Rupert Stuart-Smith & Johannes Wendland

 

I. Background of lawsuit
Pari Island, located north of the coast of Jakarta, is a paradise at risk of being lost: As a low-lying island with its highest point only 3 meters above sea level, Pari is at the frontline of climate change. For Pari’s roughly 1’500 inhabitants, this means they are threatened existentially in more than one way: Rising sea levels and floodings endanger not only their homes but ultimately their island and their way of living. Equally at risk is their economical existence, based mostly on fishing and tourism, as a changing climate appears to affect the abundance of fish and tourists are scared off by floods and the erosion of Pari’s beautiful beaches. Without mitigation of climate change and proper adaptation measures, the island community will sooner or later be forced to abandon their home.
This is why four Pari islanders decided to act and sue one of the world’s carbon majors, Holcim Ltd., based in Switzerland. A study by Richard Heede shows that
Holcim is responsible for 0,42 % of all global industrial CO2 emissions since the year 1750. In their lawsuit, the plaintiffs demand that Holcim (i) reduce its emissions by 43% until 2030 and 69% until 2040, (ii) contribute financially to climate change adaptation measures on Pari island, and (iii) redress proportionally their climate induced damages on Pari island. They argue that Holcim has added substantially to global climate change and, hence, must be held responsible for its impacts, as well. The four plaintiffs are supported through a public campaign called “Call for Climate Justice”,1 carried by three NGOs: HEKS in Switzerland, the European Center for Constitutional and Human Rights (ECCHR) in Berlin, and WALHI / Friends of the Earth Indonesia. Besides supporting the plaintiffs, the NGOs aim to draw attention to climate change, questions of climate justice, and climate change induced loss & damage, especially in the Global South. 

To succeed in court, the plaintiffs need to demonstrate that there is a causal link between climate change and the impacts they suffer. In other words: they need to show that the floodings, negative effects on fishing, and erosion can be attributed to climate change. While a study by Jochen Hinkel et al2 concluded that it is “virtually certain” that anthropogenic climate change has already caused damages on Pari Island, the study focused mostly on flood events and some uncertainties remain.


II. Possible questions of I2C for training school
1…Diminishing abundance of fish: The plaintiffs who are fishermen lament that it has become much harder to find fish and that some species have virtually disappeared. Relatedly, one of the plaintiffs explains that she cannot operate her fish farm of grouper fish anymore because the sea water is too warm, causing the fish to die. However, there is so far no study proving that these effects are due to climate change (and not other factors such as overfishing, pollution etc.). If it would be possible for a student project to
substantiate the role of climate change, this would be very useful for the lawsuit.
2. Sea level rise and floodings: Pari Island’s highest point is only three meters above sea level, the house of one of the plaintiffs lies 70 to 90 cm above the median sea level. This means that an increased median sea level has significant consequences for the island. The study by Jochen Hinkel et al found that the median sea level around Pair has increased 15 to 26 cm due to climate change. A flood on 4/5 December that flooded the plaintiff’s house with 20 cm of water and caused damage would hence not have occurred or have been
less severe without climate change. Due to gaps in the data, however, the study was only able to estimate the median sea level rise and the historic extreme sea levels around the island with a moderate level of certainty. A student project could possibly mitigate these uncertainties and establish the link to climate change more strongly.

3. Viability of life on Pari Island: While historic emissions will inevitably cause further sea level rise, the future of life on Pari depends on two factors: The extent of global warming in the future and suitable adaptation measures on the island. According to Jochen Hinkel et al, Pari Island could, under the right circumstances, in principle remain habitable over the next centuries. A student project might further specify this conclusion by assessing life on Pari under various global warming and adaption scenarios.
4. Erosion: According to media reports, Pari Island has lost 11% of its surface area between 2013 and 2021 due to erosion. For the plaintiffs, this means not only a loss of living space and an increased risk of floods but also that the beaches, which make their island attractive to tourists, might disappear. However, Jochen Hinkel et al note that satellite images do not confirm the reported loss of surface area and that potential erosion could be caused by other factors than sea level rise. A student project might be able to confirm whether there is erosion and whether it is linked to climate change.


Project 5: Fast-track attribution: how can we integrate impacts in the Climameter framework? 

 

Supervisor: Davide Faranda & Robin Noyelle

 

This project introduces students to ClimaMeter, an analytical framework developed to contextualize and analyze weather extremes within a climate context. ClimaMeter’s methodology is based on identifying similar weather patterns responsible for extreme events (analogues) and analyzing two distinct periods since 1979 to distinguish between natural climate variability and anthropogenic climate change influences. While ClimaMeter offers swift and
reproducible rapid attribution analysis, it faces limitations in unprecedented scenarios, lacks tailored features for impact attribution studies, and may encounter challenges in effectively communicating results to the general public. Students will delve deeper into ClimaMeter’s methodology, gaining insights into its contributions to climate attribution and extreme event analysis. Participants will explore ClimaMeter’s approach to attributing extreme weather events, learning how rapid attribution tools could bridge the gap between immediate contextualization of weather events and peer-reviewed attribution studies. Throughout this project, students will engage in learning the Python code necessary to detect analogues and will brainstorm potential enhancements to ClimaMeter’s attribution chain, focusing particularly
on implementing additional modules for impact attribution and refining communication strategies (e.g., incorporating new elements into the ClimaMeter figure). Through hands-on activities and collaborative discussions, students will gain practical insights into the challenges and opportunities of climate attribution rapid tools, ultimately contributing to a deeper understanding of the complex interplay between climate and weather extremes. 

This training school is powered by the ‘Climate Graduate School’ (IPSL-CGS) of the Insitut Pierre-Simon Laplace, sponsored by the National Research Agency (ANR) within the framework of the Investissements d’Avenir programme called ‘Academic School for Research’ (Ecoles Universitaire de Recherche – EUR), under the reference n° ANR-11-IDEX-0004 – 17-EURE-0006.

 

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101003469.