Modeling Dust in East Asia by CESM and Sources of Biases

East Asian dust has a significant impact on regional and global climate. In this study, we evaluate the spatial distributions and temporal variations of dust extinction profiles and dust optical depth (DOD) over East Asia simulated from the Community Earth System Model (CESM) with satellite retrievals from Luo et al. (2015a, 2015b) (L15), Yu et al. (2015) (Y15), and Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) level 3 (CAL-L3) products. Both L15 and Y15 are based on CALIPSO products but use different algorithms to separate dust from non-dust aerosols. We find high model biases of dust extinction in the upper troposphere over the Taklamakan Desert, Gobi Desert, and Tibetan Plateau, especially in the summer (June-July-August, JJA). CESM with dust emission scheme of Kok et al. (2014a, 2014b) has the best agreement with dust extinction profiles and DOD from L15 in the Taklamakan Desert and Tibetan Plateau. CESM with the default dust emission scheme of Zender et al. (2003a) underpredicts DOD in the Tibetan Plateau compared with observations from L15 due to the underestimation of local dust emission. Large uncertainties exist in observations from L15, Y15, and CAL-L3 and have significant impacts on the model evaluation of dust spatial distributions. We also assess dust surface concentrations and 10 m wind speed with meteorological records from weather stations in the Taklamakan and Gobi Deserts during dust events. CESM underestimates dust surface concentrations at most weather stations due to the inability of CESM to capture strong surface wind events.

Evaluations of tropospheric aerosol properties simulated by the community earth system model with a sectional aerosol microphysics scheme

A sectional aerosol model (CARMA) has been developed and coupled with the Community Earth System Model (CESM1). Aerosol microphysics, radiative properties, and interactions with clouds are simulated in the size-resolving model. The model described here uses 20 particle size bins for each aerosol component including freshly nucleated sulfate particles, as well as mixed particles containing sulfate, primary organics, black carbon, dust, and sea salt. The model also includes five types of bulk secondary organic aerosols with four volatility bins. The overall cost of CESM1-CARMA is approximately ∼2.6 times as much computer time as the standard three-mode aerosol model in CESM1 (CESM1-MAM3) and twice as much computer time as the seven-mode aerosol model in CESM1 (CESM1-MAM7) using similar gas phase chemistry codes. Aerosol spatial-temporal distributions are simulated and compared with a large set of observations from satellites, ground-based measurements, and airborne field campaigns. Simulated annual average aerosol optical depths are lower than MODIS/MISR satellite observations and AERONET observations by ∼32%. This difference is within the uncertainty of the satellite observations. CESM1/CARMA reproduces sulfate aerosol mass within 8%, organic aerosol mass within 20%, and black carbon aerosol mass within 50% compared with a multiyear average of the IMPROVE/EPA data over United States, but differences vary considerably at individual locations. Other data sets show similar levels of comparison with model simulations. The model suggests that in addition to sulfate, organic aerosols also significantly contribute to aerosol mass in the tropical UTLS, which is consistent with limited data.