Descripción del proyecto
In a warming world, extreme heat will become more likely, and more extreme. Extreme heat can lead to devastating socioeconomic and ecological impacts, especially when peak maximum temperatures occur together with other compounding stressors, such as high ambient humidity, lack of nighttime cooling, or persistent drought. When unprecedented record-shattering compound heat extremes occur, it raises the question of whether current climate models are able to sufficiently capture the risk and intensity of such unlikely but devastating extremes under present, and especially future climate conditions. Are climate model projections downplaying the risk and intensity of current and future heat due to missing or incorrect process representation? Can climate models sufficiently sample the most unlikely and extreme albeit still plausible events? And how would such utterly unlikely extreme compound heat develop? Which intensity, persistence, or compounding could it reach? These are the questions I will answer in TrueHeat, and by doing so I will produce the best-informed knowledge of the unlikely but plausible heat that we may come to experience in the near-term future, and how single instances of extreme heat can turn into their most devastating and unprecedented version.
To do this, I will first exploit the vast amount of existing large ensembles from different climate models that allows me to curate best-estimate future heat and heat-impact projections using superior evaluation and performance-weighting frameworks. Then I will use climate prediction ensembles to assess where chance brings the largest heat intensification, while determining why forecasted black swan heat extremes did not occur, or occurred in a less extreme form in reality. Lastly, I will generate novel targeted boosted large ensembles to quantify the role that chance plays in the intensification of extreme heat, and to fully sample the most unlikely but physically plausible worst-case heat storyline.