Evaluating heat-related mortality in Korea by objective classifications of ‘air masses’

✍ Scribed by Jan Kyselý; Radan Huth; Jiyoung Kim

• Publisher: John Wiley and Sons
• Year: 2009
• Tongue: English
• Weight: 692 KB
• Volume: 30
• Category: Article
• ISSN: 0899-8418
• DOI: 10.1002/joc.1994

✦ Synopsis

Abstract

Objective classifications of weather types (‘air masses’, AMs) are a frequently used tool for evaluating and predicting increased mortality due to heat stress in summer. The air‐mass‐based approach takes into account the entire weather situation rather than single elements; it identifies ‘oppressive’ AMs associated with elevated mortality in a given location/area, and applies regression models within the oppressive AMs in order to account for and predict excess mortality. Principal component analysis and cluster analysis are used to define the AMs. This study examines the applicability of the method in South Korea. We focus on methodological issues that concern (1) the selection of input meteorological variables, (2) the way the input variables are treated (averaged/pooled station data), (3) the number of principal components retained for the cluster analysis, and (4) the number of clusters (AMs) formed. The oppressive AMs are examined with respect to the mean relative excess mortality above baseline, and the coverage of days with large excess mortality. We find that results strongly depend on the settings of the classification procedure, and general rules concerning the most appropriate methodology for the identification of oppressive AMs are difficult to be formulated. As the coverage of days with large excess mortality by the oppressive AM is the most important criterion for the application into predicting elevated mortality risks, the oppressive AMs that cover a small fraction of such days do not appear to be useful in spite of the large mean excess mortality. For South Korea, the classifications based on only two input variables, air temperature and dew‐point deficit, pooled input data, and two retained principal components are superior for identifying conditions associated with large excess mortality. Both meteorological and non‐meteorological parameters are found to be important predictors in regression models for excess mortality within the oppressive AMs of the selected classifications. Copyright © 2009 Royal Meteorological Society