Li xiang yu successfully passed through the doctoral dissertation defense on May 8th 2015 at the 101 conference room of the IAP scientific research building. His doctoral thesis title is the Westerlies and Global Monsoon in Mid-Pliocene Warm Period from Simulations

    As ones of the most important components of the global circulation system, westerlies and global monsoon system are relation to each other and have impact on ecosystem, water resource and humans’ activity through the influence on the weather and climate in the mid-latitude and monsoon regions, and have been paid close attention from the past, present, and future climate research fields. The mid-Pliocene warm period, approximately 3.0 to 3.3 million years ago, is an interval of relatively warm climate in the Earth’s history. Study the the past westerlies and global monsoon change in mid-Pliocene warm period would provide guidance for coping with climate change under the future warming senarios. A number of related researches of paleoenvironment reconstruction and climate simulations have been undertaken so for. Previous research on the change of westerlies and monsoon in earth history is mainly based on geological evidences, and at the same time shows large difficulties and uncertainties. The inherent unreliability and multi-interpretation of proxy data dampen their effectiveness in paleoclimate reconstruction. On the other hand, the spatial distribution of proxy data is rather sparse. And there is short of quantitative analysis on paleoclimate dynamical machanism and call for the paleoclimate modeling method. Though several modeling work have performed on past monsoons but mainly on regional East Asian Monsoon, and the paleomonsoon variation in mid-Pliocene from the view of global monsoon system is to be investigated. In this dissertation, the model results from the Pliocene Model Intercomparison Project were used to analyze the variation in mid-Pliocene in the annual and seasonal westerlies, and the global monsoon area and monsoon precipitation, as well as the underlying dynamical mechanism. At the same time, inter-model and model–data comparisons had been conducted. The main conclusions are summarized as follows:

    1. The characteristics and the possible underlying mechanism of the variation of annual westerlies in mid-Pliocene had been analyzed. Although there was an inter-model scatter, the mid-Pliocene westerlies in the Southern Hemisphere and Northern Hemisphere shifted poleward compared to the reference period, accompanied by a meridional dipole pattern of zonal winds which strengthened on the poleward flank of the midlatitude westerly jet and weakened on the equatorward flank. As a result of the atmospheric thermal structure adjustment, the meridional temperature gradient altered accordingly in the midlatitudes, with larger values on the poleward flank of the midlatitude westerly jet and smaller ones on the equatorward flank, and played an important role in the roughly in-phase variation of the zonal wind anomalies in the troposphere through thermal wind balance, On the other hand, the poleward shift of the westerly jet with a dipole pattern in the midlatitudes corresponded to the weakened Hadley cell and the poleward shift of the mean meridional circulation (both the Hadley and Ferrel cells). A small set of convincing geological evidences from the North Pacific deep-sea sites supported the major features of the simulated North Pacific westerlies, and confirmed the simulated weakening of the mid-Pliocene zonal winds on the equatorward flank of the westerly jet.

    2. The seasonal characteristics of the mid-Pliocene changes in westerlies had been examined. Compared with the reference period, the Northern Hemisphere westerlies shifted poleward in winter, spring and autumn, and moved poleward (equatorward) in the middle and higher (lower) troposphere in summer. The variations of seasonal westerlies in mid-Pliocene warm period were roughly in phase with the change of the meridional temperature gradient and mean meridional circulation. The ensemble mean of atmospheric general circulation models (AGCM-MME) and all models (MME) simulated a poleward shift of the Southern Hemisphere westerlies, while the ensemble mean of coupled atmosphere–ocean general circulation models (AOGCM-MME) showed no discernible meridional shift of the southern westerlies. Though both AGCM-MME and AOGCM-MME simulated meridional dipole pattern of westerly variation, the former (latter) showed that an obvious (a little) weakening of the zonal wind along the equatorward flank of westerly jets and a small (large) strengthening along the poleward flank of westerly jets. Summer westerly change in the Northern Hemisphere revealed certain features different from that of both annual and other seasonal characteristics. The zonal mean result showed the Northern Hemisphere westerlies shiftted poleward (equatorward) in middle and higher (lower) troposphere, with same circumstance happening in the North Pacific westerlies. The North Atlantic westerlies shifted equatorward (poleward) during summer (winter, spring and autumn). Caution should be taken to simply apply the annual westerly variation to summer, and to directly mix the hemisphere features with regional ones. Since it is difficult to reconstruct variation of the past summer westerlies based on eolian size, the information of summer westerly change in mid-Pliocene from model results can be used as a reference in paleoclimate reconstruct research.

    3. The variations of global monsoon area and monsoon precipitation in mid-Pliocene have been analyzed. It showed that in the mid-Pliocene warm period the global monsoon area (GMA), monsoon precipitation (GMP), and monsoon precipitation intensity (GMPI) reduced relative to the reference period, with an average of 6.4%, 8.0%, and 1.7%, respectively. In ocean (land) monsoon areas, both the GMP and GMA in ocean domain decreased (increased) with an average of 24.9% and 27.2% (6.4% and 6.0%), and the GPI decreased weakened by 3.0% (varied little). The reduced GMP is best characterized by more precipitation in land monsoon domains and less precipitation in ocean monsoon domains, an out-of-phase variation emerged in each hemisphere. The increased land monsoon precipitation was mainly derived from more monsoon precipitations in northern Africa, Asia and northern Australia. It is the variation of GMA not the GMPI that make contribution to the GMP change. Variations of temperature and adjustments of sea level pressure gradient between hemispheres, land and sea, and close regions, induced modification in the water vapor flux and its convergence, correspondingly leading to the monsoon precipitation changes. Simulated summer and annual precipitations were consistent with the reconstructed evidences in northern and western China, northern Chinese Loess Plateau (Jinbian section), west Yunan province, northern Africa and Australia. In central Chinese Loess Plateau (Chaona and Lingtai sections), model and geological evidences agreed with each other in annual precipitation. Because of the controversy between geological evidences, it is not able to judge whether the simulated and reconstructed summer precipitation changes are consistent or not in central Chinese Loess Plateau.
    4. The possible reasons for the spread between AGCMs and AOGCMs in changes of mid-Pliocene westerlies and global monsoon had been investigated. The different treatments of sea surface temperatures and sea ice between AGCMs and AOGCMs were speculated to impact greatly on the simulated temperature structure and zonal winds, and thus result in the different modifications of the westerly anomalies and global monsoon between the two model groups. The AGCMs showed a clear poleward shift of westerlies with a dipole structure in both hemispheres. The AOGCMs simulated a similar poleward shift of the Northern Hemisphere westerlies with a dipole structure, but no discernible meridional shift of the Southern Hemisphere westerly jets. And the weakening of the zonal wind along the equatorward flank of westerly jets is less obvious in AOGCMs than in AGCMs. On the other hand, the variations of monsoon precipitation in regions of southern South Asian and East Asian monsoon area and northern Madagascar Island in Africa are opposite between AGCM-MME and AOGCM-MME, with the former simulated negative signals while the latter showed positive ones. In addition, though both simulated an increased monsoon precipitation in northern Africa, northern the South Asian, and northern Australia, the magnitude is larger in AGCM-MME results than in AOGCM-MME.