INTERNATIONAL JOURNAL OF CLIMATOLOGY Int. J. Climatol. (2010) Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/joc.2125 Short Communication Air flow and stability indices in GCM future and control runs J. A. Hanafin, a,b,c * R. McGrath, b T. Semmler, b S. Wang, b P. Lynch, a S. Steele-Dunne a,d and P. Nolan a a School of Mathematical Sciences, University College Dublin, Belfield, Dublin 4, Ireland b Met ´ Eireann, Glasnevin, Dublin 9, Ireland c Environmental Change Institute/School of Physics, National University of Ireland, Galway, Ireland d Water Resources Section, Faculty Civil Engineering & Geoscience, Delft University of Technology, Postbus 5048, NL-2600 GA Delft, Netherlands ABSTRACT: Indices have been used as indicators of synoptic-scale flow strength, shear vorticity, flow direction and static stability over Ireland and Britain. Changes in large-scale dynamic flow and static stability over the European region are expected because of shifting climate patterns, and investigation of how these indices change in future runs of global climate models allows us to estimate how this will affect storm frequency and intensity in the region. Analysis of frequency distributions shows an increase in westerly flows and decreases in most other flow directions, indicating an increase in rainfall for the region. The flow strength on days with strong winds increases in the future runs, as does the number of gale days. The future runs show not only an overall increase in atmospheric stability but also significantly larger areas with stronger instability during periods of extreme instability. Copyright  2010 Royal Meteorological Society KEY WORDS regional climate change; climate change modeling; general circulation models; atmospheric stability; Ireland; synoptic scale circulation patterns Received 31 December 2007; Revised 23 December 2009; Accepted 9 February 2010 1. Introduction Weather types or circulation patterns have been used extensively to characterize the larger scale atmospheric state into different classes for regional applications, from analysis of trends in these classes to statistical modelling of surface variables based on the classes. An analysis of observed rainfall over Ireland (Sweeney and O’Hare, 1992) found distinct regional patterns of occurrence and intensity of precipitation associated with different Lamb weather types. A number of different approaches have been taken to identify and define the classes, some based on dynamics and some based on statistical analy- ses. Subjective (Lamb, 1972) and objective (Jones et al., 1993; Goodess and Palutikof, 1998; Trigo and DaCa- mara, 2000) analyses of weather types have been based on mean sea level pressure (MSLP) alone, and also (James, 2007) on MSLP and geopotential height. Sin- gular value decomposition (Santos et al., 2007), rotated principal component analyses (PCA) (Kostopoulou and Jones, 2007), empirical orthogonal function (Zorita and von Storch, 1999) and classification and regression tree * Correspondence to: J. A. Hanafin, Environmental Change Insti- tute/School of Physics, National University of Ireland, Galway, Uni- versity Road, Ireland. E-mail: Jenny.hanafin@nuigalway.ie analyses (Zorita et al., 1995) of MSLP have been used. Other approaches to characterizing the atmospheric state prior to downscaling include using self-organizing maps of u and v vectors, humidity and temperature (Hewitson and Crane, 2006) and PCA (Wibig, 1999), correlation map-based classification (Lund, 1963) and fuzzy rule- based classifications (Bardossy et al., 2005) of geopoten- tial height. The aim of this study was to investigate projected changes in weather patterns over Ireland due to climate change, using flow and stability indices determined from re-analysis and global climate model (GCM) fields from 20th century and future runs. The flow indices in question were used in a companion study as predictor fields for a statistical downscaling model of daily precipitation as the indices were shown to be well correlated with rainfall in the region (Hanafin et al., 2010). It is of interest to also investigate projected changes in the indices themselves in GCM runs, complementing storm tracking studies (Semmler et al., 2008) by examining regional changes in the parameters which provide the conditions for heavy rainfall and severe storms to develop. Changes in the North Atlantic storm track, for example, will influence the frequency and intensity of frontal systems in this particular area and should be apparent from the Copyright  2010 Royal Meteorological Society