1961-1997年华北春季沙尘暴频次持续减少，近10年又有增多趋势，整个过程中年际变率非常明显，如2006年发生沙尘暴18次，而2003年只有2次。这种长期减少趋势以及显著的年际变率，其背后隐藏的气候过程是值得思考的。以往工作主要基于沙尘暴频次原始序列进行分析，由于沙尘暴频次存在长期趋势，所得研究结果更多情况下是反映年代际尺度上对沙尘暴频次的影响，对于华北沙尘暴频次年际变率的特征和成因尚未得到充分研究。文中重点分析影响华北沙尘暴频次年际变率的主要原因，以及与其他气候要素的关联机制，最后利用模式模拟加以验证。 春季东亚近地面天气扰动是导致华北春季沙尘暴频次年际变率的主要原因，近地面天气扰动的长期减弱，是造成华北沙尘暴频次长期减少的重要原因。当东亚中高纬对流层高层天气变率增强，导致近地表天气扰动增强，华北沙尘暴频次偏多，这种东亚对流层高层扰动可能与欧洲北部的对流层高层天气扰动有关，而且是通过中高纬西风传输和引起的。春季东亚天气扰动偏强时所对应的环流背景是：对流层高层华南上空副热带西风急流显著增强，同时中亚上空以及北太平洋上空西风强度偏强，300 hPa高度上贝加尔湖西侧北风偏强，贝加尔湖以东南风偏强；对流层中层主要表现为东亚大槽偏强；对流层低层表现为我国华北以及日本上空存在气旋型环流异常，相应的蒙古气旋或东北低压活动多影响华北地区，华北上空为显著的西北风异常。这种环流形势很容易形成沙尘暴天气。 欧亚地表温度异常与东亚近地面天气扰动存在一定联系。当3月欧亚大陆中高纬度南北地表温差加大，与之相伴随的是中高纬度西风加强，由此引起东亚地区中纬度西风偏强，西风急流垂直环流加强，动量下传加强和锋面活动增强，进而引发大风天气和沙尘暴天气。当4-5月西伯利亚南部地表温度负异常，伴随副热带西风急流偏强，位置偏北，同时中高纬度西风路径偏南，贝加尔湖以南天气变率显著增强，容易引起华北沙尘暴发生。 全球尺度气候背景与东亚天气扰动存在一定关联。北极涛动(AO)与华北沙尘暴频次年际变率为负相关关系，冬季AO指数偏高时，对流层高层副热带西风急流核偏东，高纬度西风路径偏北，高纬度风暴轴偏北；对流层中层阻塞高压活动多位于贝加尔湖地区，东亚大槽位置偏东，大气稳定，不易形成锋面气旋等天气系统，这种配置使得我国北方天气扰动减少，不利于华北沙尘暴发生。El Niño发生时，春季华北沙尘暴频次偏少，La Niña发生时，春季华北沙尘暴频次偏多。 El Niño发生时，春季东亚中高纬度西风位置偏北，与中高纬度西风伴随的反向环流中的下沉支强度减弱，位置偏北，造成高层动量下传减弱，同时伴随东亚大槽偏浅，东亚冬季风较弱，华北沙尘暴天气偏少。La Niña发生时，春季环流背景与El Niño发生时基本相反，有利于华北沙尘暴天气发生。全球变暖伴随东亚15天以下近地面天气变率增强，主要表现为近地面北非上空西南-东北向正异常，到达西西伯利亚后为西北-东南方向正异常，到日本南部后表现为西南-东北方向正异常；同时产生贝加尔湖西南方向正位势高度异常和鄂霍次克海负位势高度异常，导致我国东部地区存在明显北风异常，有利于沙尘天气增多。 利用CEMSYS5沙尘数值预报模式，采用NCEP再分析资料作为边界场，模拟1982-2006年我国北方2-5月沙尘天气。模式输出结果表明：3月温差指数偏高时，3月华北强沙尘天气和弱沙尘天气日数偏多，西北地区东部、蒙古南部、我国内蒙中西部以及华北等地，起沙量偏多；西伯利亚南部4-5月地表温度正异常伴随4-5月华北弱沙尘天气日数偏少，4月我国北方气柱沙尘含量和沙尘浓度负异常，5月从蒙古南部沿着中蒙边界到我国东北地区，沙尘浓度为负异常，同时我国北方地区沿着40N整体表现为正异常；AO对华北强沙尘天气频次和弱沙尘天气频次存在显著影响，当冬季AO指数偏高时，春季我国北方大部分地区以及华南等地气柱沙尘含量和沙尘浓度偏低，而且从地面到高空沙尘浓度显著减少；El Niño(La Niña)发生后，华北强、弱沙尘天气日数偏少(多)，但ENSO体现为对3-4月沙尘天气的影响，El Niño(La Niña)发生后，3月华北沙尘浓度偏低(高)，内蒙中西部以及蒙古南部起沙量减少(增加)，而4月我国北方大部分地区沙尘浓度偏高(低)；全球变暖对华北强沙尘天气频次的影响更为突出，全球变暖，华北强沙尘天气频次显著增加，3-4月我国北方地区春季沙尘浓度以及气柱沙尘含量显著偏高，5月我国北方大多数地区沙尘浓度和气柱沙尘含量偏低，塔里木盆地偏低幅度最大，但是辽宁、吉林东部、朝鲜半岛以及东部沿海地带气柱沙尘含量以及沙尘浓度偏高。

In the past 50 years the dust storm frequency in northern China experienced the long-term decreasing trend, but in the latest 10 years it showed the increasing trend. During these periods the annual variation of dust storm frequency was evident, for instance, in 2001 the dust storms happened 18 times, but in 2003 only 2 dust storm events in northern China were reported, so the cause of long-term change and evident annual variation of dust storms in Northern China needs more attention. Much progress has been made in the analyzing the relationship of dust storms with more climatic factors, but it is worth pointing out that most works mainly reveals the impact of climatic factors on dust storm frequency on inter-decadal time scales because the time series of dust storm frequency has the long-term trend which affects the correlation results, and less works had been done on the inter-annual time scales. Of particular importance is to analyze the main cause resulting in the evident annual variation of dust storm frequency, for results on inter-annual time scales maybe are better compared to that on the inter-decadal time scales to understand the mechanisms of the outbreak of dust storms in northern China. In this thesis, the main cause of annual variation of dust storms in northern China is analyzed, and thereby diagnose the possible association between dust storms and some other climatic factors, finally the possible mechanisms is validate by the results from numerical simulation. Spring synoptic variance in East Asia maybe is the main cause leading to the annual variation of dust storm frequency, and the long-term decreasing trend of synoptic variance in East Asia is the prominent reason causing the decline of dust storm frequency in northern China, which explains a large portion of the decreasing in dust storm frequency.When the synoptic variances in East Asia at the mid-high latitude over the top of troposphere become stronger, the surface synoptic variances in East Asia concurrently increase which is affected by the synoptic disturbance from the upper troposphere and helpful for the outbreak of dust storms in northern China, and this synoptic disturbance in East Asia in the upper troposphere is related with synoptic variations in the upper troposphere over the north part of Europe and transported by zonal wind in the mid-high latitude. When spring East Asia synoptic disturbances become stronger the circulation pattern shows that the subtropical jet stream over the southern China in the upper troposphere intensifies evidently and the zonal wind over the mid Asia and the north Pacific also become stronger, at the 300 hPa level the north wind anomaly increase over the west of Baikal and south wind anomaly over the east of Baikal increases, in the middle of troposphere the east Asia trough intensifies, and in the lower of troposphere there occurs cyclone pattern height anomaly, so the Mongolia cyclones and Northeast Low influence the northern China where the northwest wind prevails. The anomaly of the surface temperature in March over Eurasia is related with spring surface synoptic disturbances in East Asia. When the meridian temperature difference between the middle and high latitude increases in Eurasia, it is in association with the intensified zonal wind in mid-high latitude over the west part of Eurasia that results in the stronger zonal wind in the mid latitude over East Asia, more momentum transferring to the lower troposphere and intensified cold front, and thereby maybe easily produce strong wind and dust storm events. The negative temperature anomaly in Apr and May south of Baikal is closely related with the stronger subtropical jet stream northward, the zonal wind southward in the middle and upper latitude, and amplified synoptic variances to the south of Baikal, all of which are helpful for the outbreak of dust storm events in northern China. The synoptic disturbance in East Asia is linked to the background circulation. The Arctic Oscillation (AO) shows negative correlation with annual variations of dust storm frequency in northern China. During the high-AO winter the core of subtropical jet stream moves eastward, the zonal wind northward in high latitude and thereby the storm track northward in high latitude. In the middle of troposphere the blocking activities occur more frequently over the Baikal and then the East Asia trough moves eastward that are related to the stable circulation and may not easily produce the cyclone cold fronts and other synoptic systems. This pattern leads to the drop of synoptic disturbances and thereby reduces the outbreak of dust storm in northern China. When El Niño occurs the dust storms frequency increases and the La Niña is opposite to the El Niño. The occurrence of El Niño is closely related with the zonal wind northward, intensified zonal wind and amplified synoptic variances in middle and high latitude, meanwhile the sink branch of the anti-Hadley circulation combined with the zonal wind in the middle and high latitude weaken and then the momentum transporting decreases, and both the East Asian trough and the winter monsoon get weaker. When La Niña takes place, the spring circulation is opposite to that in El Niño that is helpful for the production of dust storms. Global warming is closely correlated with the anomaly of synoptic variances, which mainly includes the positive anomaly northeastward over the northern Africa, positive anomaly southeastward over the West Siberia and positive anomaly northeastward over the south of Japan. Meanwhile global warming also is linked with the positive geopotential height anomaly over the southwest of Baikal and negative geopotential height anomaly over Okhotsk, which leads to the north wind anomaly prevailing in the east part of China that is helpful for the occurrence of dust storm in northern China. The Computational Erosion Model System (CEMSYS5) is used in this study to simulate the spring dust storms process during the past 25 years nudged by the NCEP reanalysis datasets. Numerical simulation is carried out from Feb.1st to May 31st in each year. The output results show that when the temperature difference in March over the Eurasia increases, the strong and weak dust storm frequency increase and the dust emission also increases over the east part of northwest in China, south of Mongolia, the mid and west part of Inner-Mongolia and northern China. The positive temperature anomaly in April and May is concurrently linked to the decrease of weak dust storm frequency in Northern China. In Apr the dust load and dust concentration over the north part of china decrease. Meanwhile in May the dust concentration declines from the south of Mongolia to the northeast of China along the boundary between China and Mongolia, and positive dust concentration anomaly occurs along 40°N in China. AO makes remarkable influence on the strong and weak dust storm frequency. During the high-AO years the dust load and dust concentration in spring over the most part of North China and the south of China decreases, and evidently weaken from the surface to the lower troposphere. When El Niño (La Niña) occurs, in spring the strong and weak dust storm frequency decreases (increases), but El Niño/Southern Oscillation (ENSO) mostly exerts more influence on the dust weather in March and Apr. When El Niño (La Niña) breaks out, in March the dust concentration weakens and the emission of dust over the mid and west of Inner-Mongolia and south of Mongolia decreases, but in Apr the dust concentration increases in the north China.Global warming maybe makes a great role in influencing strong dust storm frequency. The global warming is linked to the increase of dust storm frequency in northern China, especially for strong dust storms, and the dust load and dust concentration intensifies during the periods from March to April, but in May the dust concentration and dust load evidently decreases, especially in Tarim, on the opposite, dust load and dust concentration increases in the Liaoning, east of Jilin, Korean Peninsular and the vicinity of sea in eastern China.