ALBERT

Description: Variations in oxygen isotope compositions (δ18O) provide insight into modern climate and past changes in climate and topography. In addition, in regions such as Tibet, geologic archives of isotope ratios record climate change driven by plateau uplift and therefore also provide information about the surface uplift history. A good understanding of modern-day controls on δ18O is crucial for interpreting geologic δ18O in this context. We use the ECHAM5-wiso global atmospheric general circulation model to calculate δ18O in precipitation (δ18Op) for the present-day climate. In the region of the Tibetan Plateau, spatial variations of monthly means of δ18Op are statistically related to spatial variations of 2 m air temperature and precipitation rate, as well as to topography. The size and location of investigated regions are varied in our study to capture regional differences in these relationships and the processes governing the modern δ18Op. In addition to correlation analyses, a cross-validated stepwise multiple regression is carried out using δ18Op as the predictand, and topography and atmospheric variables (temperature and precipitation amount) as predictors. The 2 m air temperature and topography yield the highest spatial correlation coefficients of 〉0.9 and 〈 0.9, respectively, throughout most of the year. Particularly high correlation coefficients are calculated for the region along the Himalayan orogeny and parts of western China. The predictors explain 〉90% of the δ18Op spatial variance in the same regions. The 2 m air temperature is the dominant predictor and contributes 93.6% to the total explained spatial variance on average. The results demonstrate that most of the δ18Op pattern on and around the Tibetan Plateau can be explained by variation in 2 m air temperature and altitude. Correlation of the dependent predictors indicate that in low-altitude regions where topography does not determine temperature variability, the high correlation of temperature and δ18Op may partially be explained by variations in precipitation rates.