Precipitation δ18O over the Himalaya-Tibet orogen from ECHAM5-wiso simulations: Statistical analysis of temperature, topography and precipitation

Sebastian G. Mutz,Todd A. Ehlers,Jingmin Li,Christian Steger,Heiko Paeth,Martin Werner,Christopher J. Poulsen
DOI: https://doi.org/10.1002/2016JD024856
2016-01-01
Abstract:Variations in oxygen isotope compositions (delta O-18) 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 delta O-18 is crucial for interpreting geologic delta O-18 in this context. We use the ECHAM5-wiso global atmospheric general circulation model to calculate delta O-18 in precipitation (delta O-18(p)) for the present-day climate. In the region of the Tibetan Plateau, spatial variations of monthly means of delta O-18(p) are statistically related to spatial variations of 2m 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 delta O-18(p). In addition to correlation analyses, a cross-validated stepwise multiple regression is carried out using delta O-18(p) as the predictand, and topography and atmospheric variables (temperature and precipitation amount) as predictors. The 2m 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 delta O-18(p) spatial variance in the same regions. The 2m 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 delta O-18(p) pattern on and around the Tibetan Plateau can be explained by variation in 2m 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 delta O-18(p) may partially be explained by variations in precipitation rates.
What problem does this paper attempt to address?