The Effect of a Short Observational Record on the Statistics of Temperature Extremes

Zeder, Joel; Sippel, Sebastian; Pasche, Olivier C.; Engelke, Sebastian; Fischer, Erich M.

doi:10.1029/2023gl104090

Cited by 5 publications

(3 citation statements)

References 39 publications

(67 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Results depend weakly on the covariate dataset used (ERA5 or GISTEMP, see figure 1). We note however an overestimation of the estimated RP compared to the actual RP, which becomes significant when RPs greater than 50 years are considered: this is probably partly due to estimation of RPs higher than the length of the time series, and partly due to the known tendency of maximum likelihood estimation to over-estimate return times (Zeder et al 2023). We found that from 1950-1979 (prior to the satellite era), in some regions with few observation data (parts of South America and Africa), a large number of extreme events with high RPs was found (see also figure 2).…”

Section: Resultsmentioning

confidence: 74%

Heat extremes linearly shift with global warming, with frequency doubling per decade since 1979

Vautard,

Barnes,

Philip

et al. 2024

Environ. Res. Lett.

View full text Add to dashboard Cite

Heat extremes have been increasing both in frequency and in intensity in most land regions of the world, and this increase has been attributed to human activities. In the last decade, many outstanding and record shattering heat extremes have occurred worldwide, triggering fears of a nonlinear behaviour or an “acceleration” in the development of heat conditions, considering the warming level when the event occurred. Here we show that the evolution of yearly temperature maxima, with return periods above 10 years, consistently shifts with global temperatures and does not significantly depart from this behaviour in recent years or decades when considered globally or at the scale of continents. This result is obtained by using a classical statistical event attribution technique, where the assumption that the distribution of block-maxima extremes linearly shifts with global warming is tested across years and world land regions. However, the pace of frequency change is large, with the probability of heat extremes exponentially rising and nearly doubling every decade since 1979, particularly when considering events with a return period of about 10 to 50 years in 2000. This makes the climate of a decade ago unrepresentative of today’s climate. Our results overall mean that we do not expect events like the recent outstanding extremes to undergo nonlinear changes, despite fast changes. They also show that assumptions underlying attribution techniques used in many recent studies are consistent with recent temperature trends.

show abstract

Section: Resultsmentioning

confidence: 74%

Heat extremes linearly shift with global warming, with frequency doubling per decade since 1979

Vautard,

Barnes,

Philip

et al. 2024

Environ. Res. Lett.

View full text Add to dashboard Cite

show abstract

“…It is a common practice in event attribution studies to perform "out of sample" statistical analyses where the event in question (often the maximum value in the observations) is excluded. It has recently been pointed out that this can introduce a selection bias (Miralles and Davison, 2023;Zeder et al, 2023). However, it can be the case that the maxima in question can be far outliers and cause very poor quality of fit for the rest of the extreme value sample data if included in the fitting data (Bercos-Hickey et al, 2022) casting doubts on the validity of the attribution statements.…”

Section: Discussionmentioning

confidence: 99%

On the uncertainty of long-period return values of extreme daily precipitation

Wehner,

Duffy,

Risser

et al. 2024

Front. Clim.

View full text Add to dashboard Cite

Methods for calculating return values of extreme precipitation and their uncertainty are compared using daily precipitation rates over the Western U.S. and Southwestern Canada from a large ensemble of climate model simulations. The roles of return-value estimation procedures and sample size in uncertainty are evaluated for various return periods. We compare two different generalized extreme value (GEV) parameter estimation techniques, namely L-moments and maximum likelihood (MLE), as well as empirical techniques. Even for very large datasets, confidence intervals calculated using GEV techniques are narrower than those calculated using empirical methods. Furthermore, the more efficient L-moments parameter estimation techniques result in narrower confidence intervals than MLE parameter estimation techniques at small sample sizes, but similar best estimates. It should be noted that we do not claim that either parameter fitting technique is better calibrated than the other to estimate long period return values. While a non-stationary MLE methodology is readily available to estimate GEV parameters, it is not for the L-moments method. Comparison of uncertainty quantification methods are found to yield significantly different estimates for small sample sizes but converge to similar results as sample size increases. Finally, practical recommendations about the length and size of climate model ensemble simulations and the choice of statistical methods to robustly estimate long period return values of extreme daily precipitation statistics and quantify their uncertainty.

show abstract

“…Here we see, besides the remarkably high agreement of the empirical and analytical latitudinal patterns, that analytical probability ratios tend to be slightly lower. This might be explained by the limitations of GEV-fits in representing the highest end of the tail, especially if the sample size is modest [40].…”

Section: Global Precipitation Record Shattering Probability Increases...mentioning

confidence: 99%

Surge in future record shattering precipitation driven by increasing variability

de Vries,

Sippel,

Zeder

et al. 2023

Preprint

View full text Add to dashboard Cite

Climate events that shatter previous records by large margins are a threat to society because critical infrastructure is often not designed to withstand such events. Climate change is expected to increase the probability of such events happening, but quantifying event probabilities is challenging due to natural variability and limited data availability, especially in observations and for events that would be "impossible" without climate change. Here, we quantify the changing probability of record shattering precipitation events in climate models using extreme value statistics. We show that CMIP6 models forced with SSP3-7.0 project much higher probability of record shattering extreme precipitation in a changing climate relative to a stationary climate by the end of the century. This holds for almost all the global land, in line with the expected intensification of extreme precipitation. For vulnerable regions in the tropics, the projected probability of shattering precipitation records anywhere in the region in a 5-year period exceeds 50% in the high warming SSP3-7.0 scenario. We demonstrate that increasing variability is an essential driver of near-term increases in precipitation record shattering probability. Overall, our results provide estimates of probabilities and intensities of unseen record shattering precipitation events in a warming world that are crucial to inform risk assessment and adaptation policies.

show abstract

The Effect of a Short Observational Record on the Statistics of Temperature Extremes

Cited by 5 publications

References 39 publications

Heat extremes linearly shift with global warming, with frequency doubling per decade since 1979

Heat extremes linearly shift with global warming, with frequency doubling per decade since 1979

On the uncertainty of long-period return values of extreme daily precipitation

Surge in future record shattering precipitation driven by increasing variability

Contact Info

Product

Resources

About