Prof. Carlo De Michele (Politecnico di Milano, Italy)

Dr. Emanuele Bevacqua (Helmholtz Centre for Environmental Research – UFZ, Germany)

Prof. Manuela Brunner (SLF, Switzerland)

Dr. Mariana M. de Brito (Helmholtz Centre for Environmental Research – UFZ, Germany)

Dr. Giorgia Di Capua (PIK, Germany)

Prof. Reik Donner (Magdeburg-Stendal University of Applied Sciences, Germany)

Prof. Fabrizio Durante (University of Salento, Italy)

Prof. Sebastian Engelke (University of Geneva, Switzerland)

Dr. Kai Kornnhuber (IIASA, LDEO-Columbia University)

Dr. Milan Palus (Czech Academy of Sciences, Czech Republic)

Prof. Simona Sacchi (Università degli Studi di Milano Bicocca, Italy)

Prof. Gianfausto Salvadori (University of Salento, Italy)

Dr. Lisa Thalheimer-Prezyna (United Nations University, Germany)

Student-Project – Instructors

Fabiola Banfi (Politecnico di Milano, Italy)

Taís Maria Nunes Carvalho (UFZ, Germany)

Beijing Fang (UFZ, Germany)
Timothy Lam (PIK, Germany)
Lisa Thalheimer-Prezyna (United Nations University, Germany)
Pauline Rivoire (University of Lausanne and EPFL, Switzerland)

Kevin Schwarzwald (Columbia University, USA)

Student Projects

Participants of the Training School are requested to choose from one of the following 6 student projects. During the two weeks there will be ample time for the groups to work on their projects, with the intended outcome of each project being a submittable manuscript in the months following the School.

Project 1: “Assessing causality in hot and dry events”

Supervision: Giorgia Di Capua and Timothy Lam 

The Mediterranean is a hot spot region for anthropogenic climate change, which has experience both enhanced warming and drying with respect to other regions in the world. Studying atmospheric causal drivers of compound hot and dry events in the region is crucial to identify predictability potential and make informed risk assessments. In this project, we aim to study compound hot and dry events applying causal discovery [Runge, 2018; Runge et al., 2019] and machine learning techniques to both reanalysis data and Coupled model intercomparison project (CMIP) historical and future projections. 

References

Runge J., 2018. Causal network reconstruction from time series: From theoretical assumptions to practical estimation. Chaos 1 July 2018; 28 (7): 075310. https://doi.org/10.1063/1.5025050

Runge J. et al., 2019. Detecting and quantifying causal associations in large nonlinear time series datasets. Sci. Adv. 5,eaau4996(2019). https://doi.org/10.1126/sciadv.aau4996.

Project 2: “Complex network analysis of compounding opposite hydrometeorological extremes”

Supervision: Reik Donner and Fabiola Banfi

Instantaneous co-occurrence of drought and flood conditions in different regions can be triggered by large-scale atmospheric circulation anomalies like blocking patterns, while successions of opposite extremes (even at the same location) may be related to the lifetime of such quasi-stationary states of the atmosphere. Based on available long-term observational and reanalysis data of relevant meteorological and hydrological characteristics, a systematic characterization of compounding opposite hydrometeorological extremes at continental to global scale is attempted by exploiting functional complex network analysis in combination with suitable co-occurrence statistics [Donner et al. 2017; Boers et al. 2019].

References

Boers, N., Goswami, B., Rheinwalt, A. et al. Complex networks reveal global pattern of extreme-rainfall teleconnections. Nature 566, 373–377 (2019). https://doi.org/10.1038/s41586-018-0872-x

Donner R.V., Wiedermann M., Donges J.F. (2017). Complex Network Techniques for Climatological Data Analysis. In: Franzke CLE, O’Kane TJ, eds. Nonlinear and Stochastic Climate Dynamics. Cambridge University Press; 2017:159-183.

Student project 3: “Compound Extremes, Risk Perception and Climate migration”

Supervision: Simona Sacchi and Lisa Thalheimer

While hazards may be only indirectly influenced by individual and group actions, exposure and vulnerability are directly shaped by the human factor. Consequently, risk estimation models could significantly benefit from psychologically informed insights into human precautionary behavior and adaptation strategies. These factors become even more critical in the case of compound events, where risk information is highly complex and difficult to process cognitively [Sacchi et al. 2022]. In such situations, determining the most appropriate behavioral response can be particularly challenging. Also, psychologically insights may have a key role in the decision processes driving the human migration in response to adverse weather conditions [Thalheimer, 2023].

In this project, we want to investigate empirically, through the development and realization of ad-hoc surveys, the connections between compound extremes and risk perception, and those between risk perception and climate migration.

References

Sacchi S., Faccenda G., De Michele C. (2022). Risk perception and behavioral intentions in facing compound climate-related hazards. iScience. 26(1): 105787 https://doi.org/10.1016/j.isci.2022.105787

Thalheimer, L. (2023). Compounding Risks and Increased Vulnerabilities: Climate Change, Conflict, and Mobility in East Africa. In: Walker, T., McGaughey, J., Machnik-Kekesi, G., Kelly, V. (eds) Environmental Migration in the Face of Emerging Risks. Palgrave Macmillan, Cham. https://doi.org/10.1007/978-3-031-29529-4_8

Student project 4: “A global database of the impacts of extreme events from text data”

Supervision: Mariana Madruga de Brito and Taís Maria Nunes Carvalho

In this project, we aim to extract detailed information on the impacts of extreme events (e.g., floods, storms, and droughts) based on Red Cross reports. By analyzing these reports, we will gather comprehensive data on both direct and indirect impacts on society and the environment. This includes metrics like fatalities, economic losses, and increases in migration. Since the Red Cross focuses especially on the Global South, we expect to address spatial biases common in global impact databases (e.g., EM-DAT focuses on Europe and the US). To extract structured information from the reports, we will utilize natural language processing (NLP) tools. Our approach will incorporate both supervised classification methods (e.g., Sodoge et al., 2023) and large language models (LLMs) (e.g., Carvalho et al., 2024).

References

Carvalho T. M. N. et al. (2024, under review) Beyond the surface: leveraging NLP to map global natural hazard impacts. https://openreview.net/forum?id=Jo4tqdIpfX  

Sodoge, J., Kuhlicke, C., & de Brito, M. M. (2023). Automatized spatio-temporal detection of drought impacts from newspaper articles using natural language processing and machine learning. Weather and Climate Extremes, 41, 100574.

Student Project 5: “Quantifying the economic impacts from compound extremes”

Supervision Kai Kornhuber, Pauline Rivoire and Kevin Schwarzwald

The increasing frequency and intensity of compound extreme events—where multiple extreme weather conditions occur simultaneously or in close succession—pose significant economic challenges globally, which so far have only been sparsely quantified [1] but are generally missing in the stress testing schemes of large financial institutions and the Insurance Sector. This project aims to quantify the economic impacts of such compound extremes on specific sectors using a range of climate and socioeconomic data sets and the employment of innovative statistical methods. The ultimate goal is to provide a holistic stress testing framework that adequately incorporates currently unaccounted risks posed by compound extreme events.

References

[1] Compound Risks: Implications for Physical Climate Scenario Analysis; Dolk, Mahul, Ranger, Ceglar, Kornhuber https://www.ngfs.net/sites/default/files/media/2023/11/07/ngfs_compound_risks_implications_for_physical_climate_scenario_analysis.pdf

Student Project 6: “”

Supervision: Manuela Brunner and Beijing Fang