“…Using this composite data set, we derived the second-order least squares regression for each isotope with elevation: [Liu et al, 2008b[Liu et al, , 2009 (Figure 5), the isotope elevation gradient is distinct from those along other boundaries of the Tibetan Plateau. On the south slope of the Himalaya and southern Tibet, the d 18 O w altitudinal gradients are higher (20.320.23&/100m) [Garzione et al, 2000;Zhang et al, 2002;Ding et al, 2009;Hren, 2009;Yao et al, 2009;Wen et al, 2012]. In contrast, on the northern and southeastern part of the Tibetan Plateau, the gradients are as low as 20.15 &/100 m Ding et al, 2009;Bershaw et al, 2012], and 0.19&/100 m [Hoke et al, 2014], respectively.…”