A large ensemble illustration of how record-shattering heat records can endure

Autor: James S Risbey, Damien B Irving, Dougal T Squire, Richard J Matear, Didier P Monselesan, Michael J Pook, Nandini Ramesh, Doug Richardson, Carly R Tozer
Jazyk: angličtina
Rok vydání: 2023
Předmět:
Zdroj: Environmental Research: Climate, Vol 2, Iss 3, p 035003 (2023)
Druh dokumentu: article
ISSN: 2752-5295
DOI: 10.1088/2752-5295/acd714
Popis: The record-shattering hot day in the Pacific Northwest in June 2021 is used to motivate a study of record-shattering temperature extremes in a very large hindcast ensemble. The hottest days in the Pacific Northwest in the large ensemble have similar large scale and synoptic patterns to those associated with the observed event. From the perspective of a fixed location, the hottest ensemble days are acutely sensitive to the chance sequencing of a dry period with a precisely positioned weather pattern. These days are thus rare and require very large samples (tens of thousands of years) to capture. The enduring nature of record-shattering heat records can be understood through this lens of weather ‘noise’ and sampling. When a record-shattering event occurs due to chance alignment of weather systems in the optimal configuration, any small sample of years subsequent to the (very unlikely) record event has an extremely low chance of finding yet another chance extreme. While warming of the baseline climate can narrow the gap between more regular extremes and record-shattering extremes, this can take many decades depending on the pace of climate change. Climate models are unlikely to capture record-shattering extremes at fixed locations given by observations unless the model samples are large enough to provide enough weather outcomes to include the optimal weather alignments. This underscores the need to account for sampling in assessing models and changes in weather-sensitive extremes. In particular, climate models are not necessarily deficient in representing extremes if that assessment is based on their absence in undersize samples.
Databáze: Directory of Open Access Journals