Quantile-based Bias Correction and Uncertainty Quantification of Extreme Event Attribution Statements

/in /by

Extreme event attribution characterizes how anthropogenic climate change may have influenced the probability and magnitude of selected individual extreme weather and climate events. Attribution statements often involve quantification of the fraction of attributable risk (FAR) or the risk ratio (RR) and associated confidence intervals. Many such analyses use climate model output to characterize extreme event behavior with and without anthropogenic influence. Climate models may have biases in their representation of extreme events.

There are two important results from this work. The first result is that we present a methodology tailored to bias correct the extreme events simulated by climate models. This is necessary to make better use of the CMIP5 database of climate models. The second, and more significant result, concerns a finding about the uncertainty in attributing the human influence on individual climate and weather extreme events. We present a method to estimate the lower bound of the change in risk of events when the upper bound includes infinity. We find that this lower bound, using this one-sided estimate, is insensitive to the magnitude of the extreme event. This important finding means that observational uncertainty is not critical to determining whether humans have had a significant influence on an actual extreme event. We demonstrate the methodology using the summer 2011 central US heatwave and output from the Community Earth System Model.

 

Citation: Soyoung Jeon, Christopher J. Paciorek, Michael F. Wehner (2016) Quantile-based Bias Correction and Uncertainty Quantification of Extreme Event Attribution Statements. Weather and Climate Extremes. Early online release. DOI:10.1016/j.wace.2016.02.001

http://www.sciencedirect.com/science/article/pii/S2212094715300220

Precipitation over 20 years

/in /by

 

Observed twenty year return values of seasonal maximum daily precipitation. Units: Inches/day. Understanding the seasonal cycle of extreme precipitation is a critical part of CASCADE objectives. In the US, the meteorology of extreme storms varies greatly across seasons. The stats and software teams have developed R and Python based tools enabling robust descriptions of extreme temperature and precipitation.

3rd US NCA figure 2

/in /by

Figure for US NCACASCADE team members Chris Paciorek and Michael Wehner prepared this figure describing changes in extreme temperature and precipitation for review articles in the Bulletin of the American Meteorological Society as source material for the 3rd US National Climate Assessment.

The figure set shows change over 1950-2007 in estimated 20-year return value (oC) for (a) upper tail of daily maximum temperature, max(Tmax) (b) lower tail of daily maximum temperature, min(Tmax) (c) upper tail of daily minimum temperature, max(Tmin) and (d) lower tail of daily minimum temperature, min(Tmin). Results are based on fitting extreme value statistical models with a linear trend in the location parameter to exceedances of a location-specific threshold (greater than the 99th percentile for upper tail and less than the 1th percentile for lower tail). Circles indicate z-score for the estimated change (estimate divided by its standard error), with absolute z-scores exceeding 1, 2, and 3 indicated by open circles of increasing size.

Source: Peterson, et al. (2013) Monitoring and Understanding Changes in Heatwaves, Coldwaves, Floods and Droughts in the United States: State of Knowledge, Bulletin of the American Meteorological Society June 2013, 821-834, DOI: 10.1175/BAMS-D-12-00066.1, Supplement DOI: 10.1175/BAMS-D-12-00066.2

3rd US NCA figure 1

/in /by

CASCADE team members Chris Paciorek and Michael Wehner prepared this figure describing changes in extreme temperature and precipitation for review articles in the Bulletin of the American Meteorological Society as source material for the 3rd US National Climate Assessment.  The figure shows changes in observed twenty year return value of the daily accumulated precipitation from 1948 to 2010. Units: inches. Only locations for which data from at least 2/3 of the days in the 1948-2010 period were recorded are included in this analysis. The change in return period threshold at each station is shown by a circle whose relative size portrays its statistical significance.

Source: Kunkel et al. (2013) Monitoring and Understanding Trends in Extreme Storms: State of Knowledge, Bulletin of the American Meteorological Society, 94, 499–514, 10.1175/BAMS-D-11-00262.1