Modeling the Asian Aridity during the Early Cenozoic

Guo et al., 2008; Licht et al., 2016; Li et al., 2018a). Previous modelling studies focus mainly on the monsoon climate during the early Cenozoic, and studies for the Asian aridity are still limited (e.g., Huber and Goldner, 2012; Zhang et al., 2012; Li et al., 2018b). Here Asian aridity during the early Cenozoic is investigated through climate modelling by changing atmospheric CO2 concentration, orbital parameters, and topography. Six numerical experiments were conducted (Table 1) using the Community Atmosphere Model version 4 (CAM4) (Neale et al., 2013). Due to the long simulation time required by highresolution, fully coupled, atmosphere–ocean models under ~50 Ma BP boundary conditions, we used simulated sea surface temperatures (SSTs) by running the lower resolution (horizontal resolution of ~3.75°) version of the Norwegian Earth System Model (NorESM-L) (Zhang et al., 2014) under ~50 Ma BP paleogeography (Scotese, 2001) and an atmospheric CO2 concentration of 1120 or 560 ppmv for the 2,200 modelled years. We employed the high-resolution CAM4 (horizontal resolution of ~1°) to simulate atmospheric conditions using the same paleogeography, atmospheric CO2 concentration and associated simulated SSTs computed from the NorESM-L. In detail, experiments E_1120 and E_560 were set to different atmospheric CO2 concentrations (Table 1). Experiments E_560_6k and E_560_21k were conducted to check climate sensitivity to changes of orbital parameters (Table 1). Experiment E_560_notopo was designed to evaluate the climate effects of topography, in which global topography was lowered (Table 1). Finally, experiment E_modern was conducted under modern boundary conditions for comparison (Table 1). Aridity was measured using the aridity index (AI) (e.g., Ahani et al., 2013; Fu and Feng, 2014) that is the ratio of annual precipitation to potential evapotranspiration (PET). We applied the Penman-Monteith algorithm to estimate PET (e.g., Allen et al., 1998; Fu and Feng, 2014). The AI is widely used to quantify aridity worldwide (e.g., Fu and Feng, 2014). Here, dry regions are defined having an AI smaller than 0.65, and are further classified into dry-subhumid (0.5 ≤ AI < 0.65), semi-arid (0.2 ≤ AI < 0.5), arid (0.05 ≤ AI < 0.2), and hyperarid (AI < 0.05) types. Wet regions are defined as having an AI greater than 0.65. Model results indicate that the mid-latitude Asia always has dry regions under ~50 Ma BP boundary conditions with dry regions located mainly between 20°N to 40°N (Figs. 1a−1e). By comparison, modern dry regions are located mainly at north and west sides of the Tibetan Plateau (TP) and limited between 20°N to 40°N in East Asia (Fig. 1f). Thus, from the early Cenozoic to the present day, changes in boundary conditions — particularly the development of the TP and surrounding topography — led to northward migration of Asian dry regions. During the early Cenozoic Asian aridity was affected significantly by the changes in atmospheric CO2 concentration, orbital parameters, and topography. Along with the decreases of atmospheric CO2 concentration (from 1120 ppmv to 560 ppmv), Asian dry regions expanded and their aridity strengthened (Figs. 1a and 1b). Arid areas, in particular, expanded significantly. Moreover, Asian aridity weakened (strengthened) when the Northern Hemispheric summer received more (less) solar insolation (Figs. 1c and 1d), indicating that orbital forcing significantly modulated Asian aridity during the early Cenozoic. Without the topography, Asian aridity would weaken, with fewer arid areas (Figs. 1e vs. 1b). This result indicates the development of Asian topography has played an important role on the development of Asian aridity during the Cenozoic. Modeling the Asian Aridity during the Early Cenozoic