The second version of Chinese Academy of Sciences Earth System Model (CAS‐ESM 2) is described with emphasis on the development process, strength and weakness, and climate sensitivities in simulations of the Coupled Model Intercomparison Project (CMIP6) DECK experiments. CAS‐ESM 2 was built as a numerical model to simulate both the physical climate system as well as atmospheric chemistry and carbon cycle. It is a newcomer in the international modeling community to provide sufficiently independent solutions of climate simulations from those of other models. Performances of the model in simulating the basic states of the radiation budget of the atmosphere and ocean, precipitation, circulations, variabilities, and the 20^th Century warming are presented. Model biases and their possible causes are discussed. Strength includes horizontal heat transport in the atmosphere and oceans, vertical profile of the Atlantic Meridional Overturning Circulation; weakness includes the double Inter‐Tropical Convergence Zone (ITCZ) and stronger amplitude of the El Niño‐Southern Oscillation (ENSO) that are also common in many other models. The simulated the 20^th Century warming shares a similar discrepancy with observations as in several other models—less warming in the 1920s and stronger cooling in the 1960s than in observation—at the time when there was a steep increase of anthropogenic aerosols. As a result, the 20^th Century warming is about 60% of the observed warming despite that the model simulated a similar slope of warming trend after 1980 to observation. The model has an equilibrium climate sensitivity of 3.4 K with a positive cloud feedback from the shortwave radiation.

The dust emission parameterization in the Chinese Academy of Sciences Earth System Model version 2 (CAS‐ESM2) is described with emphasis on the implementation process and simulations of global dust cycle and East Asia dust event statistics. The parameterization is based on the scheme of Shao (2004; http://doi.org/10.1029/2003JD004372) which considers two major dust emission mechanisms, namely, saltation bombardment and aggregation disintegration. Shao (2004; http://doi.org/10.1029/2003JD004372) scheme was well tested against field observations and has been widely used in regional dust modeling. Here this scheme is implemented into a global climate model for the first time and thus provides an independent solution for simulating global dust emissions. With this scheme, CAS‐ESM2 reasonably simulates the main dust emission regions in the Earth and reproduces the observations of dust deposition flux and surface dust concentrations at most stations. Compared to the synoptic records of dust events, the model also captures the general patterns and seasonal variations of dust activities in East Asia. However, the model underestimates the frequency of strong dust events (instantaneous surface dust concentrations >1,000 μg m⁻³) due to the weaker surface winds simulated by the model. The model tends to simulate much weaker and longer‐lasting dust events in Eastern Sources (35–49°N, 94–126.5°E) of northern China, suggesting the weaker and slower‐moving synoptic weather systems associated with dust events in the model. Overall CAS‐ESM2 performs well in simulating the key aspects of global dust distribution and East Asian dust events, yet some biases remain to be improved.

Orographic drag is an essential process for numerical weather predictions in complex terrain regions, which depends on the inflow direction. In this study, we define the orographic asymmetry vector (OAV) for a coarse grid as the normalized vector between the grid's center point and its center of mass, and the orographic asymmetry in a flow direction—which describes the inclination direction and the extent of the subgrid terrain—is calculated as the projection of OAV on this direction. Calculations of the effective orographic length (OL) and the model grid length λ are extended to all flow directions. A new orographic drag scheme, which considers the effect of orographic anisotropy in all directions, is then developed based on the OAV projection and the extended OL and λ for any given direction. Sensitivity tests of the orographic drag under the new scheme are conducted using a 5 m/s vertically uniform wind along different directions for four coarse grid points in typical mountain regions. The new scheme is shown to provide a more continuous transition of the orographic parameters and the resulting stress as a function of flow direction than piecewise transition of schemes with only eight directions. The predicted momentum flux profile of the new scheme is compared with mountain‐wave simulations obtained from the integrated modeling system IAP‐WRF (Institute of Atmospheric Physics‐Weather Research and Forecasting Model) for the Rocky Mountain. The new scheme is shown to predict an overall narrower stress scatter about the reference simulation than the old scheme.

A reasonable representation of orographic anisotropy in earth system models is vital for improving weather and climate modeling. In this study, we implement the orographic drag scheme, including 3‐D orographic anisotropy (3D‐AFD), into the Chinese Academy of Sciences Earth System Model version 2 (CAS‐ESM 2.0). Three groups of simulations named sensitivity run, medium‐range forecast, and seasonal forecast respectively are conducted using the updated CAS‐ESM model and validated against station observation and reanalysis data. These simulations are run using the three schemes—3D‐AFD scheme, the 3D orographic anisotropy scheme for the eight‐direction (3D‐8x), and the 2D isotropic scheme (2D)—to compare their performance in CAS‐ESM 2.0. Sensitivity runs show that the 3D‐AFD provide more continuous transition of calculated drag as function of wind direction than the 3D‐8x, while the drag calculated using the 2D scheme show no variation with change of wind direction. Enhanced drag in the medium range and seasonal forecast using the updated CAS‐ESM alleviates part of the winter wind speed bias over the Tibetan Plateau (TP) and the cold bias over TP and the Siberian polar region. It is shown that the 3D‐AFD scheme alleviates more bias than that of the 3D‐8x scheme (by wind speed reduction of 1 ∼ 2 m/s and temperature of 1 ∼ 2 K) especially in the seasonal forecast. The results suggest that reasonable representation of the orographic anisotropy is important in climate modeling.