Search Results
You are looking at 1 - 3 of 3 items for
- Author or Editor: Megan C. Kirchmeier-Young x
- Refine by Access: All Content x
Abstract
Extreme events are important to many studying regional climate impacts but provide a challenge for many “deterministic” downscaling methodologies. The University of Wisconsin Probabilistic Downscaling (UWPD) dataset applies a “probabilistic” approach to downscaling that may be advantageous in a number of situations, including realistic representation of extreme events. The probabilistic approach to downscaling, however, presents some unique challenges for verification, especially when comparing a full probability density function with a single observed value for each day. Furthermore, because of the wide range of specific climatic information needed in climate impacts assessment, any single verification metric will be useful to only a limited set of practitioners. The intent of this study, then, is (i) to identify verification metrics appropriate for probabilistic downscaling of climate data; (ii) to apply, within the UWPD, those metrics to a suite of extreme event statistics that may be of use in climate impacts assessments; and (iii) in applying these metrics, to demonstrate the utility of a probabilistic approach to downscaling climate data, especially for representing extreme events.
Abstract
Extreme events are important to many studying regional climate impacts but provide a challenge for many “deterministic” downscaling methodologies. The University of Wisconsin Probabilistic Downscaling (UWPD) dataset applies a “probabilistic” approach to downscaling that may be advantageous in a number of situations, including realistic representation of extreme events. The probabilistic approach to downscaling, however, presents some unique challenges for verification, especially when comparing a full probability density function with a single observed value for each day. Furthermore, because of the wide range of specific climatic information needed in climate impacts assessment, any single verification metric will be useful to only a limited set of practitioners. The intent of this study, then, is (i) to identify verification metrics appropriate for probabilistic downscaling of climate data; (ii) to apply, within the UWPD, those metrics to a suite of extreme event statistics that may be of use in climate impacts assessments; and (iii) in applying these metrics, to demonstrate the utility of a probabilistic approach to downscaling climate data, especially for representing extreme events.
Abstract
Arctic sea ice extent (SIE) has decreased over recent decades, with record-setting minimum events in 2007 and again in 2012. A question of interest across many disciplines concerns the extent to which such extreme events can be attributed to anthropogenic influences. First, a detection and attribution analysis is performed for trends in SIE anomalies over the observed period. The main objective of this study is an event attribution analysis for extreme minimum events in Arctic SIE. Although focus is placed on the 2012 event, the results are generalized to extreme events of other magnitudes, including both past and potential future extremes. Several ensembles of model responses are used, including two single-model large ensembles. Using several different metrics to define the events in question, it is shown that an extreme SIE minimum of the magnitude seen in 2012 is consistent with a scenario including anthropogenic influence and is extremely unlikely in a scenario excluding anthropogenic influence. Hence, the 2012 Arctic sea ice minimum provides a counterexample to the often-quoted idea that individual extreme events cannot be attributed to human influence.
Abstract
Arctic sea ice extent (SIE) has decreased over recent decades, with record-setting minimum events in 2007 and again in 2012. A question of interest across many disciplines concerns the extent to which such extreme events can be attributed to anthropogenic influences. First, a detection and attribution analysis is performed for trends in SIE anomalies over the observed period. The main objective of this study is an event attribution analysis for extreme minimum events in Arctic SIE. Although focus is placed on the 2012 event, the results are generalized to extreme events of other magnitudes, including both past and potential future extremes. Several ensembles of model responses are used, including two single-model large ensembles. Using several different metrics to define the events in question, it is shown that an extreme SIE minimum of the magnitude seen in 2012 is consistent with a scenario including anthropogenic influence and is extremely unlikely in a scenario excluding anthropogenic influence. Hence, the 2012 Arctic sea ice minimum provides a counterexample to the often-quoted idea that individual extreme events cannot be attributed to human influence.
