QUESTIONING THE ACCURACY OF K-AR AGES: EXAMPLE OF THE DATING OF TWO TRACHYTIC DOMES FROM THE GOLCUK (TURKEY)

The Golcuk volcano, located south to the city of Isparta (Turkey), built-up in three stages during the Pleistocene (Alici et al., 1998; Platevoet et al., 2008; Elitok et al., 2010). A reliable and accurate dating of the most recent volcanic stage is crucial to assess the potential hazards for the infrastructures and the populations. In this context, Platevoet et al. (2008) and Schmitt et al. (2014) applied respectively the K-Ar and the (U-Th)/He methods to date the youngest volcanic stage of the Glocuk and obtained contrasted results. Based on the K-Ar dating of two trachytic domes (Kucukcesme Tepe et Pilav Tepe), the post-caldera stage is dated between 52 +/- 2 ka and 24 +/- 2 ka by Platevoet et al. (2008). When the (U-Th)/He method is applied on zircons from the same trachytic domes, Schmitt et al. (2014) date the post-caldera stage around 14.1 +/- 0.5-12.9 +/- 0.4 ka. Schmitt et al. (2014) proposed that the K-Ar ages were erroneously too old, probably because of excess Ar-40*. We have tested this hypothesis by dating groundmass splits of these trachytes using the Ar-40/Ar-39 method. Our Ar-40/Ar-39 ages reproduce the (U-Th)/He ages on zircons and date the youngest volcanic stage of the Golcuk at 13.4 +/- 0.3 ka. Confirming the hypothesis of Schmitt et al. (2014), we propose that the error on the K-Ar ages is related to excess argon probably trapped in small fragments of clinopyroxene attached to the groundmass.