How climate change and variability affects society is a question receiving increased attention by research community. With the advent of ‘Big Data’, historical observations or pseudo-data generated by processed based numerical models (such as Global Gridded Crop Models -GGCMs-in agriculture), are increasingly being employed to empirically estimate reduced-form bio- physical responses to meteorology. The resulting fitted response surfaces, when forced by Earth System Model (ESM) simulations of future climate, function as computationally tractable statistical emulators of climatic shocks to evaluate the broader energy and economic implications of the sectoral climate change impacts. The talk will introduce a simple and flexible statistical crop yield emulator built on ~2 million agro-climatic observations at a fine scale global gridded resolution of 0.5 deg (~55 km). Some insights on the potential application of cluster analysis will be also discussed with a broader aim of exploring synergies between traditional econometrics and machine learning methodologies.
Based in Venice, Malcolm Mistry is currently a post-doctoral researcher at the Department of Economics, Ca’ Foscari University, and a researcher at the Euro-Mediterranean Center on Climate Change (CMCC). His prior affiliations include research collaborations at Fondazione Eni Enrico Mattei(FEEM, Venice) and Barcelona Supercomputing Center (Spain).
Malcolm’s research focuses on examining impacts of climate change and variability using reduced form statistical models, aka. emulators. His other research interests focus on development of climate extreme indices, application of machine learning tools for modelling impacts of climate change, and the variability in Indian Summer Monsoons (considered as a lifeline for the farming community in South-East Asia).
Malcolm holds a PhD in Science and Management of Climate Change from Ca’Foscari University of Venice, and a Master’s in Weather and Climate Modelling from the University of Reading (U.K.).