Very rare heat extremes: quantifying and understanding using ensemble re-initialization

Abstract

Heat waves such as the one in Europe 2003 have severe consequences for the economy, society, and ecosystems. It is unclear whether temperatures could have exceeded these anomalies even without further climate change. Developing storylines and quantifying highest possible temperature levels is challenging given the lack of long homogeneous time series and methodological framework to assess them. Here, we address this challenge by analysing summer temperatures in a nearly 5000-year pre-industrial climate model simulation, performed with the Community Earth System Model CESM1. To assess how anomalous temperatures could get, we compare storylines, generated by three different methods: (1) a return-level estimate, deduced from a generalized extreme value distribution, (2) a regression model, based on dynamic and thermodynamic heat wave drivers, and (3) a novel ensemble boosting method, generating large samples of re-initialized extreme heat waves in the long climate simulation.All methods provide consistent temperature estimates, suggesting that historical exceptional heat waves as in Chicago 1995, Europe 2003 and Russia 2010 could have been substantially exceeded even in the absence of further global warming. These estimated unseen heat waves are caused by the same drivers as moderate observed events, but with more anomalous patterns. Moreover, altered contributions of circulation and soil moisture to temperature anomalies include amplified feedbacks in the surface energy budget. The methodological framework of combining different storyline approaches of heat waves with magnitudes beyond the observational record may ultimately contribute to adaptation and to the stress testing of ecosystems or socio-economic systems to increase resilience to extreme climate stressors.

An open workflow for the study of unseen weather extremes

“…The UNSEEN approach reveals a strong trend in temperature extremes over the last 40 years, which has increased the likelihood of events like the August 2020 temperature anomalies in the present climate (about <1% 40 in 2020), whilst it was virtually impossible in 1981. This trend is consistent with record-breaking or 'recordshattering' temperatures being expected to occur more frequently in a rapidly warming climate (Coumou et al, 2013;Fischer et al, 2021;Power and Delage, 2019). This case study shows how the UNSEEN-trends method (Kelder et al, 2020) may inform policies around adapting to the likelihood of present climate extremes rather than past events.…”

Health Risks to the Russian Population from Temperature Extremes at the Beginning of the XXI Century

“…Long-term changes in the Earth's energy balance increase the frequency, intensity and duration of temperature extremes [3,14,16,19,[26][27][28]33,35,36,40,41], the probability of compound and cascading events [1,2,6,13,16,17,20,22,24,29,35,41], including wildfires [17,32,34]. Their accelerating trends are predicted under certain scenarios of greenhouse gas emissions [2,10,16,27,33,38,40]. Historically, extreme weather and climate events are generally rare across the globe, with a time interval between events that allow human and natural systems to recover from the impacts experienced [30,37,39].…”

“…As an example, the danger of heat extremes for public health was particularly evident in August 2003 in Western Europe, when the heat wave had caused over 70,000 additional deaths [1,27,28,47,68,83,119]. In 2010, a "mega heatwave" with maximum temperatures up to 40°C and the combined effect of heat and smoke from numerous wildfires, covered Eastern Europe, the European part of Russia and southern Siberia [47]; estimates for the death toll were 55,000 in Russia and 11,000 in Moscow, with 15 billion USD of total economic loss, which was near 1% Russian gross domestic product [5,10,20,21,23,24,28,35,42,47,51,61,93,108,114,128,138]. According to World Health Organization, in the European region (43 countries) in the period from 2071 to 2099, heat waves could lead to 47-117,000 additional deaths annually [77].…”

An open workflow to gain insights about low‐likelihood high‐impact weather events from initialized predictions

“…Furthermore, UNSEEN is one tool within many available tools to study plausible low‐likelihood high‐impact weather events. There is scope to assess the mutual benefit of various approaches, as is common for event attribution (Philip et al, 2020; van Oldenborgh et al, 2021), including ensembles of opportunity (e.g., King et al, 2017; Lewis et al, 2017), single‐model initial‐condition large ensembles (e.g., Suarez‐Gutierrez et al, 2020a, 2020b), ensemble reinitialization methods (e.g., Gessner et al, 2021), targeted large ensemble experiments (e.g., Guillod et al, 2017; Mitchell et al, 2017), pooling of observations (e.g., Berghuijs et al, 2017; Robinson et al, 2021), long archives (Hawkins et al, 2019; Murphy et al, 2020), paleoclimatic records (Yan et al, 2020), and statistical weather generators (Brunner & Gilleland, 2020; Wilks & Wilby, 1999; Yiou, 2014). Seasonal and decadal prediction systems may, furthermore, contribute additional lines of evidence to event attribution statement if their trend estimates can be extrapolated to represent pre‐industrial climates.…”

“…Many approaches have been developed to reduce sampling uncertainties, ranging from traditional statistical weather generators (Brunner & Gilleland, 2020; Wilks & Wilby, 1999; Yiou, 2014), extreme value approaches (Coles, 2001; Katz, 2013), and dynamical systems theory (De Luca et al, 2020; Faranda et al, 2017); through pooling of observations (e.g., Berghuijs et al, 2017; Robinson et al, 2021), the use of long archives (Hawkins et al, 2019; Murphy et al, 2020), and paleoclimatic records (Yan et al, 2020); to probing ensemble members from weather and climate models (Box 1). Ensembles of opportunity (e.g., King et al, 2017; Lewis et al, 2017), single‐model initial‐condition large ensembles (SMILEs) (e.g., Suarez‐Gutierrez et al, 2020a, 2020b), ensemble reinitialization methods (e.g., Gessner et al, 2021), and targeted large ensemble experiments (e.g., Guillod et al, 2017; Hall et al, 2019; Mitchell et al, 2017), have all been used for the study of low‐likelihood high‐impact hydro‐climatic events.…”