Bridging Human and Machine Knowledge via Explainable AI for Weather and Climate applications

Abstract: Earth’s climate is changing rapidly under the effect of global warming, leading to more frequent and severe extreme weather events [1,2]. These weather extremes, in turn, are exacting heavy socioeconomic and environmental tolls [3], prompting an urgent need for better understanding and predicting them. In this talk, we present some recent results obtained for the tropical Indo-Pacific region, using human-understandable methods (or taking a human view), namely dynamical system theory. In particular, we show that changes in weather patterns are leading to more weather extremes, namely heatwaves and extreme precipitation. We then present the use of explainable AI tools (i.e., machine view) to investigate the onset and precursors of these extremes. More specifically, we try to bridge existing human knowledge (human view) and “AI knowledge” (machine view) to better understand the behaviour and predictability of weather extremes. We also briefly show ClimaEmpact, a new AI-based tool for understand the impact of extreme weather events.

[1] "Changes in tropical Indo-Pacific weather patterns aggravate regional extremes”. C Dong, R Noyelle, G Messori, A Gualandi, L Fery, P Yiou, M Vrac, F D’Andrea, SJ Camargo, E Coppola, G Balsamo, C Chen, D Faranda, G Mengaldo (1st round of revisions)

[2] "IPCC, 2021: Summary for Policymakers. In: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change”. VP Masson-Delmotte, et al. (2021) 

[3] "Evidence for sharp increase in the economic damages of extreme natural disasters”. M Coronese, F Lamperti, K Keller, F Chiaromonte, A Roventini, PNAS (2019).

Bio: Dr Gianmarco Mengaldo is an Assistant Professor in the Department of Mechanical Engineering at National University of Singapore (Singapore), and an Honorary Research Fellow at Imperial College London (United Kingdom). He received his BSc and MSc in Aerospace Engineering from Politecnico di Milano (Italy), and his PhD in Aeronautical Engineering from Imperial College London (United Kingdom). After his PhD he undertook various roles both in industry and academia, including at the European Centre for Medium-Range Weather Forecasts (ECMWF), the California Institute of Technology (Caltech), and Keefe, Bruyette and Woods (KBW). Dr Mengaldo’s adopts an interdisciplinary approach integrating mathematical and computational engineering to study complex systems that arise in applied science. His current research interests involve (i) explainable AI, both theoretical and applied, (ii) the intersection between AI and domain knowledge, (iii) data-mining technologies for coherent pattern identification, and (iv) high-fidelity multi-physics simulation tools. Dr Mengaldo’s main application areas include engineering, geophysics, healthcare, and finance.