Late Cenozoic uplift of western Turkey: Improved dating of the Kula Quaternary volcanic field and numerical modelling of the Gediz River terrace staircase

A set of 13 new unspiked K-Ar dates has been obtained for the Quaternary basaltic volcanism in the Kula area of western Turkey, providing improved age control for the fluvial deposits of the Gediz River that underlie these basalt flows. This dating is able, for the first time, to resolve different ages for the oldest basalts, assigned to category ß2, that cap the earliest Gediz deposits recognised in this area, at altitudes of ~140 to ~210 m above present river level. In particular, the ß2 basalt capping the Sarni{dotless}ç Plateau is dated to 1215 ± 16 ka (± 2?), suggesting that the youngest underlying fluvial deposits, ~185 m above present river level, are no younger than marine oxygen isotope stage (MIS) 38. In contrast, the ß2 basalt capping the adjacent Burgaz Plateau is dated to 1014 ± 23 ka, suggesting that the youngest underlying fluvial deposits, ~140 m above present river level, date from MIS 28. The staircase of 11 high Gediz terraces capping the latter plateau is thus dated to MIS 48-28, assuming they represent consecutive ~40 ka Milankovitch cycles, although it is possible that as many as two cycles are missing from this sequence such that the highest terrace is correspondingly older. Basalt flows assigned to the ß3 category, capping Gediz terraces ~35 and ~25 m above the present river level, have been dated to 236 ± 6 ka and 180 ± 5 ka, indicating incision rates of ~0.15 mm a- 1, similar to the time-averaged rates since the eruptions of the ß2 basalts. The youngest basalts, assigned to category ß4, are Late Holocene; our K-Ar results for them range from zero age to a maximum of 7 ± 2 ka. This fluvial incision is interpreted using numerical modelling as a consequence of uplift caused by a regional-scale increase in spatial average erosion rates to ~0.1 mm a- 1, starting at ~3100 ka, caused by climate deterioration, since when a total of ~410 m of uplift has occurred. Parameters deduced on this basis from the observed disposition of the Early Pleistocene Gediz terraces include the local effective viscosity of the lower crust, which is ~2 × 1018 Pa s, the Moho temperature of ~660 °C, and the depth of the base of the brittle upper crust, which is ~13 km. The thin lithosphere in this area results in high heat flow, causing this relatively shallow base of the brittle upper crust and the associated relatively thick lower-crustal layer, situated between depths of ~13 and ~30 km. It estimated that around 900 ka, at the start of the ~100 ka Milankovitch forcing, the spatial average erosion rate increased slightly, to ~0.12 mm a- 1; the associated relatively sluggish variations in uplift rates are as expected given the relatively thick lower-crustal layer. This modelling indicates that the growth of topography since the Pliocene in this study region has not involved a steady state. The landscape was significantly perturbed by the Middle Pliocene increase in erosion rates, and has subsequently adjusted towards-but not reached-a new steady state consistent with these increased erosion rates. It would not be possible to constrain what has been occurring from the Middle to Late Pleistocene or even the Early Pleistocene uplift response alone; information regarding the starting conditions is also essential, this being available in this region from the older geological record of stacked fluvial and lacustrine deposition. This result has major implications for the rigorous modelling of uplift histories in regions of rapid erosion, where preservation of information to constrain the starting conditions is unlikely. © 2006 Elsevier B.V. All rights reserved