This talk has been automatically captioned. You can remove these by pressing CC on the video toolbar.

Name: Shrijita Bhattacharya

Talk Title: Extremes of the spatial impact of heat waves

Abstract: A region is said to be under a heat wave spell if its temperature remains above a given threshold for a prolonged span of time. Heat waves are often accompanied with some of the serious consequences like forest fires, droughts, impeded agricultural productivity and high mortality rates. As such, there has been an increasing amount of research to allow for the quantification, prediction and inference on heat waves. In this paper, we develop a statistical framework for modeling the areal impact of heat waves. A time series is constructed by considering the area of US under profound heat wave activity over moving windows of time. We extensively make use of the Pickands-Balkema-de Haan theorem to model the extremes of this time series using the generalized Pareto distribution (GPD). As a main contribution, we estimate the out-of-sample return levels for this time series as a function of covariates like the season (time of the year) and El Niño Southern Oscillation (ENSO) index. Sensitivity analysis of the proposed method with respect to factors like intensity level, duration and geographical location of heat wave events have been well studied. The proposed methodology has been evaluated in the context of United States Historical Climatology Network (USHCN) daily temperature records collected over a span of 100 years across the continental US.

This talk is a contributed talk at EVA 2021.