HYDROLOGICAL PROCESSES Hydrol. Process. 23, 1927–1936 (2009) Published online 14 May 2009 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/hyp.7331 The links between the categorised Southern Oscillation indicators and climate and hydrologic variables in Turkey M. ¸ Caˇ gatay Karab¨ ork 1 * and Ercan Kahya 2,3 1 Dumlupınar University, Civil Engineering Department, Hydraulics Division, K¨ utahya, Turkey 2 Civil Engineering Department, Istanbul Technical University, Maslak, Istanbul, Turkey 3 Civil Engineering Department, American University of Sharjah, Sharjah, United Arab Emirates Abstract: Some previous global and regional studies have indicated teleconnection between the extreme phases of the Southern Oscillation (SO) and Turkish climate and hydrologic variables; however, they failed to suggest a strong correlation structure. In this study, categorised Southern Oscillation index (SOI) and Multivariate ENSO (El Nino Southern Oscillation) index (MEI) series were used to examine the far-reaching effects of the SO on temperature, precipitation and streamflow patterns in Turkey. These SO indicators were categorised into five subgroups according to their empirical distributions. Correlations between the categorised SO indicators and three analysis variables were computed using the Spearman’s rho from lag-0 to lag-4. Significance of calculated correlations was tested at the 0Ð01 level for station-based analysis and at the 0Ð05 level for regional analysis. Temperature records demonstrated significant correlations with the categorised SOI and MEI in nearly half of the entire stations. For some categories, precipitation and streamflow were found to be correlated with the SO indicators in some stations mainly in western Turkey. Regional analyses of temperature and precipitation revealed a clear and strong correlation structure with the categorised SO indicators on a large portion of Turkey. This was not concluded by the earlier pertinent studies. Besides, this study showed that significant correlations were obtained not only for the SO extreme phases (namely, El Nino and La Nina) but also for neutral and moderate phases of the SO. Plausible explanations for the observed teleconnection are presented. Copyright  2009 John Wiley & Sons, Ltd. KEY WORDS Southern Oscillation index; multivariate ENSO index; correlation; temperature; precipitation; streamflow; Turkey Received 30 April 2008; Accepted 24 March 2009 INTRODUCTION Water has been always a vital matter affecting people from many social standpoints in the Middle East because of a noticeable regional population increase in each year and high rate of consumption in irrigation practices. Con- sequently, the Middle East is as vulnerable as any part of the globe to any changes in available surface and ground water resources. Improved understanding of the teleconnection between remote conditions in the equa- torial Pacific and climate conditions on an area (e.g. the continental United States) caused better land sur- face hydrologic predictability, especially more accurate streamflow forecasts with longer lead times than tradi- tional methods (Hamlet and Lettenmaier, 1999; Wood et al., 2002). Moreover, Maurer and Lettenmaier (2003) showed that remote climate-forcing signals and initial land surface states provide a measure of predictability of runoff over the Mississippi River basin with considerable spatial variability in the degree of predictability. El Nino Southern Oscillation (ENSO) has been linked to many climate anomalies throughout the world. The * Correspondence to: M. ¸ Caˇ gatay Karab¨ ork, Dumlupınar University, Civil Engineering Department, Hydraulics Division, K¨ utahya, Turkey. E-mail: karabork@mail.dumlupinar.edu.tr changes in the phases of the Southern Oscillation (SO) are considered as the most notable signals in inter-annual climate variations (Ropelewski and Halpert, 1987; Phi- lander, 1990; Kahya and Dracup, 1993a). In last four decades, the extreme phases of the SO (namely, El Nino and La Nina events) and their influences on cli- mate and other earth systems processes have always intrigued researchers to make an investigation from a regional scale (e.g. Redmond and Koch, 1991; Kahya and Dracup, 1993a; Dracup and Kahya, 1994; Eltahir, 1996; Chiew et al., 1998; ¸ Sen et al., 2004) to a global scale (e.g. Bradley et al., 1987; Ropelewski and Halpert, 1987, 1989; Kiladis and Diaz, 1989; Chiew and McMohen, 2002). Due to the scarcity of data in the global-scale stud- ies covering the geography of Turkey (e.g. Ropelewski and Halpert, 1987, 1989; Kiladis and Diaz, 1989), the implications of the SO signals in Turkey were incon- clusive. However, the literature review of major studies concerning the Middle East region has shown detectable significant teleconnection between the SO and surface hydrologic and climate variables. Among those, Naze- mosadat and Cordery (2000) documented the El Nino impacts on the precipitation patterns of Iran. Following this study, Nazemosadat and Ghasemi (2004) quantified the SO–precipitation relation in Iran using precipitation Copyright  2009 John Wiley & Sons, Ltd.