Increased Risk of the 2014 Australian May Heatwave Due to Anthropogenic Activity

Abstract

An evaluation on the ability of the Community Earth System Model (CESM) model to simulate heatwave-relevant synoptic patterns over Australia was performed. The method of self-organizing maps was used (Kohonen 2001), which has been shown to be a powerful tool for evaluating model simulated synoptic variability in other studies (Cassano et al. 2006; Uotila et al. 2007; Gibson et al. 2015, manuscript submitted to Climate Dyn.). Each historical realization of the CESM ensemble was compared to ERAinterim over 1979–2004 in terms of the frequency of each synoptic pattern (node). A total of 20 nodes were computed, as shown in Fig. S31.1. Many of the nodes resemble the development of the observed synoptic system associated with the 19-day 2014 May heatwave (see Fig. S31.2 for snapshots of the observed system). In particular, node c4 was most relevant to the event occurring for 8 of the heatwave days, including 6 consecutive days (Fig. S31.2). Other similar nodes, also characterized by anticyclonic conditions over the Tasman Sea, occurred during the heat wave including nodes b4, c3, d3, d4. Overall, all realizations of CESM produced similar frequencies of each self-organizing map (SOM) node. Encouragingly, the frequency of node c4 (that most relevant to the heatwave in this study) in CESM was very similar to that of ERA-interim, while some of the other less relevant nodes were either slightly underestimated (nodes c3, d3, d4) or overestimated (b4) by the model (Fig. S31.3). The biggest deficiency in the model is its tendency to overstimulate the frequency of zonally elongated patterns (i.e., nodes a2 and b2), however, this is less of a concern as these patterns are not relevant to the heat event.