Coupling Of A Global 3D Atmospheric Chemistry Transport Model With An Atmospheric General Circulation Model And Its Applications

Global warming in the past century is a solid fact without question, which is thought to be mainly induced by human activities in a lot of related research work. In order to understand the mechanisms and laws of global climate changes better and to predict the future climate states more accurately, it is greatly needed to develop an earth system model which is expected to help us scientifically and completely disclose the complicated interactions among the atmosphere, the ocean, the land surface, the sea-ice and the biosphere. One of the most important parts of the earth system model development is to couple a global three-dimensional atmospheric chemistry transport model (ACTM) and an atmospheric general circulation model (AGCM). As the first step of the plan on the coupling of ACTM and AGCM in China, the model of ozone and related chemical tracers, version 2 (MOZART-2), an ACTM, has been successfully coupled with the Community Atmospheric Model, version 2 (CAM2), an AGCM, and thus the new off-line coupled model system consisting of MOZART-2 and CAM2 has been set up. Focusing on the evaluations and applications of the new coupled system, a series of research work has been made, and the main conclusions are drawn as followings: Comparison of simulated ozone mixing ratios with multi-year averaged surface observations over East Asia shows that the coupled model system of MOZART-2 and CAM2 can reasonably reproduce two typical patterns of seasonal cycles of the observed surface ozone, namely, the pattern of a summer minimum and a winter maximum appearing in the regions with clean oceanic boundary layers which is represented by Minamitorishima site, and the pattern of a spring maximum and a summer minimum mostly in the mid-latitudes areas in the Northern Hemisphere with remote clean atmosphere represented by Ryori site. Simulation studies demonstrate that photochemistry origin is the main contributor to the spring ozone maximum appearing in some surface observation sites. In the meantime, the model system can also capture the general features of seasonal variations of surface CO concentrations quite well, although it underestimates the CO concentrations at both Minamitorishima and Ryori. The underestimation is primarily associated with the emission inventory adopted and the model climate state in this study. Compared with the ozonesonde data at four sites in the Eastern Asian areas, the vertical gradient and magnitude of ozone concentrations are reasonably simulated by the model system except a little overestimation in winter, especially in the upper troposphere. The model system can also generally capture the seasonal, latitudinal and altitudinal variations of ozone concentrations over East Asia. Analysis indicates that the underestimation of tropopause height in February contributes to the overestimation of winter ozone in the upper and middle troposphere at Tateno. Comparison of model results with multi-year averaged surface observations over Europe shows that the seasonal variations of surface ozone mixing ratios are generally reproduced except a slight overestimation at Hohenpeissenberg and a little underestimation at Kollumerwaard; in addition, the magnitude and general seasonal variation features of surface CO concentrations are also be well captured. Compared with the ozonesonde observation data at Lindenberg in Germany and Sodankyla in Finland, the vertical gradient and magnitude of tropospheric ozone mixing ratios are well reproduced by the model system, especially in autumn when the simulated mid-troposphere ozone concentrations at these sites are in better agreement with the observations. Compared with the ozonesonde observations at Hohenpeissenberg in Germany and Ny Alesund in Norway, seasonal cycles of ozone concentrations at 300hPa, 500hPa and 800hPa can also be fairly reproduced, especially at 300hPa. Summarily, good performance of this model system on simulations of troposphericozone concentrations over Europe can lay a reliable basis for further study of Tran-Eurasian transport of pollutants. The research is carried out by further using this coupled system to investigate the influences of the European emission changes on Eastern Asian pollutant distributions, and the transport path and possible mechanism are also proposed. The study shows that during winter, spring and autumn, CO pollutants emitted from European surface emissons travel across the European continent or pass by the Arctic regions into the Eastern Asian regions mostly by means of advection, and then are transported upward into the mid-troposphere or upper-troposphere by vertical diffusion or convection; while in summer, they are mainly transported upward into the mid-troposphere or upper-troposphere with the help of convection, and then go into the Eastern Asian areas by means of advection with the Westerly Jet. Analysis for investigating the influences of different atmospheric models on MOZART-2 shows that there are some differences between MOZART+CAM2 system and MOZART+MACCM3 system in simulating the tropospheric ozone over East Asia and Europe. Comparison of model results with observations shows that the main differences are found to be in the upper troposphere especially around the tropopause,and MOZART+CAM2 system can produce more reasonable simulations of ozone concentrations than MOZART+MACCM3 system does at these altitudes. Analysis shows that the main reason would be ascribed to differences of the simulated tropopause height by the two model system, and differences of the simulated wind fields and temperature fields at these altitudes by MACCM3 and CAM2.