Domenico Giaquinto


PhD in: MERC
Ciclo: XXXVI

Titolo progetto: Investigating climate compound events via a combined multi-layer networks and multi hazard assessment framework

Supervisor(s): Prof. Jürgen Kurths and Prof. Warner Marzocchi

  • Scarica il CV

  • Investigating climate compound events via a combined multi-layer networks and multi hazard assessment framework.

    Climate science is one of the most complex field of study: different chemical, biological and physical processes take place among several environmental compartments, each one characterized by their own variables and dynamics. There is a scientific consensus on the importance of studying climate in order to understand what is the role of humankind in such process, in particular of climate change, what are the major threats both for society and ecosystems and what can be done to mitigate or avoid the most dangerous future scenarios.

    Our research project focuses on climate compound events, defined as extreme impacts that depend on multiple statistically dependent variables or events. Indeed, since climate-related disasters are usually caused by compound events, our aim is to develop new tools to understand, analyze, quantify and possibly predict multiple climate extremes, which are increasing over time due to climate change, threatening natural ecosystems and human societies given their strong clashes. We believe that multi-layer networks (MLNs) could be a useful tool to embed compound events into the framework of complexity science. Underlying interaction like synchronization among spatial regions between two different extremes are likely to be revealed thanks to the correct formulation of the network and the use of the right indices. To the best of our knowledge this is the first time compound events are studied with this methodology. Moreover, we believe that these evaluations should be supported by a rigorous multi-risk assessment analysis, filling the gap between complexity and risk science.

    The general framework of this project is a multi-layer network to interpret climate compound events. The MLN will exhibit different layers, one for each extreme, with intra-layer interactions or “homogeneous interactions” between structures referred to the same event and inter-layer interactions or “non-honogeneous interactions” between structures of distinct extreme. To characterize these interactions, both network and multi-hazard techniques of analysis are used. One aim is to build a framework in which these two science field could possibly meet, combining the strengths of the two and overcoming their single limitations. In a more general manner, there is the need to inter-compare the respective capacity of these techniques to quantitatively characterize the patterns of the studied compounds. These methodological results will enable us to analyze the critical points related to extremes, the key regions responsible for generating climate regime changes, to reveal global teleconnection’s patterns affecting climate, to study the evolution of global climate patterns and finally to develop tools for predictability and forecast of extreme climate events. On top of all of this, we want to contribute to the research in this field by offering a rigorous methodology to quantitatively evaluate uncertainties when it comes to network science and climatic phenomena. The development of all these aspects would be the basis on which to build possibly a reliable forecasting procedure.