本文主要利用来自美国国家大气和海洋管理局(NOAA)全球自动气象站观测网的高时间分辨率风速观测数据，对1994年到2014年间全球海陆风进行了统计分析，着重选取了上海、东京、洛杉矶三个典型城市分别深入探讨了城市化、太阳辐射以及大尺度西风带对海陆风的影响。在完成多个单站点海陆风信号识别分离的基础上，进一步选取了全球43个沿海站点的数据，开发了一种适用于全球不同气候类型区域的海陆风信号识别及提取方法，对局地观测风场进行主模态分类，并基于此量化了在气候态时间尺度下，全球不同气候类型区域海陆风的显著性。其中，选取气候态时间尺度上海陆风显著存在的测站，本文根据海陆风风速长时间变化趋势量化了海风和陆风对于站点地表风速变化的贡献，并分析了造成两者贡献差异的原因。根据不同地区海陆风显著性以及海陆风风速变化趋势的差异，探讨了造成差异的根本影响因子和多种间接影响因子。本文的主要研究内容和结论如下：

1) 以快速城市化的上海为例，选取1994年至2014年夏季观测数据开展研究，发现在全球变暖的气候背景下，上海地区海陆风日数和海陆风风速均存在显著减小的趋势，热力因素和动力因素共同作用导致了海陆风风速减弱。就热力因素而言，大尺度空间范围上，陆地增温快于海洋，促进了海风的增加和陆风的减小，局地城市化带来的增温效应具有同样作用。城市化带来的阻挡作用作为动力因素消除了其他热力因素共同作用带来的风速增加效应，是海陆风风速减弱的决定性因素。配合局地到达地面的太阳辐射多年变化，城市化还是海陆风日数减少的重要因素。

2) 以城市化达到高水准并且水平保持稳定的东京为例，选取相同时间段夏季数据开展研究，发现海陆风日数和海陆风风速均未有明显的减小或增加趋势，且海陆风日数和海陆风风速间没有对应关系。若大尺度背景风场风向和海陆风风向同向，则有利于海陆风日的增加，反之亦然。海陆风是局地太阳辐射影响能见度的重要媒介。高太阳辐射情景下，海陆风循环的强度大于中低辐射情景，在城市热岛环流作用的配合下，对市区大气污染物的向外部或向上空输送能力增加。

3) 以属于地中海气候类型、城市化达到高水准并且保持稳定的洛杉矶为例，选取相同时间段四季数据开展研究，发现海陆风风速年均值未有明显的减小或增加趋势，但海陆风具有显著的季节性变化差异。表现为海风夏季强冬季弱，而陆风没有明显且统一的季节性变化规律。地中海气候的雨热不同期导致局地太阳辐射的季节性变化强，是海风风速具有明显季节性变化的决定性因素, 而日最低气温季节性变化弱，是导致陆风风速季节性变化小的次要原因，主要原因是地中海气候地区上层大气中的西风带在冬季明显强于夏季，促进陆风环流增强，削弱了夏冬之间的环流风速差异。

4) 选取全球43个分属不同气候类型区域的沿海站点为例，发现局地观测风场主模态可分为三类：1) 仅海陆风为风场主模态构成因子(主模态为“仅海陆风”); 2) 海陆风和系统风同时为风场主模态构成因子(主模态为“海陆风和系统风”)，其中系统风指的是任何来自天气尺度系统或行星尺度系统的风; 3) 海陆风不是风场主模态构成因子，没有显著的海陆风(主模态为“无明显海陆风”)。其中，前两类主模态表示海陆风显著存在，即海陆风的显著存在性(OE-SLB)可以被验证。可用OE-SLB被验证测站数占所有同气候类型测站数的比例来表征海陆风在某种气候类型下的显著性。气候类型是造成海陆风显著性差异的决定性因素。具体而言：海陆风在地中海气候(72.7%)，热带季风气候(75%)，热带草原气候(66.7%)和热带沙漠气候(75%)下显著性较高。在亚热带季风气候(22.2%)，热带雨林气候(33.3%)，海洋性气候(0%)以及大陆性气候(0%)下显著性较低。全球海陆风风速没有统一的变化规律，海风风速一般大于陆风。其中地中海气候站点的海陆风风速最稳定，热带沙漠气候站点具有海陆风风速反转现象，陆风风速大于海风风速。在海陆风显著性较高的气候类型下，洋流、特殊的地形、特殊的水面分布、水体面积等仍会造成OE-SLB无法被验证，从而形成一些特例站点。

5) 在上述43站中挑选出18个海陆风显著存在站点进一步研究发现，在主模态为海陆风和系统风的情况下，两者均为地表风速变化的主要影响因子，但季风气候区域除外，系统风仍然占据了影响风速变化的主导地位。在主模态为仅海陆风的情况下，海陆风贡献了82%的地表风速变化，在这种情况下，系统风的贡献值在2.65%至17.16%之间，具体数值和局地云量呈现良好的二次关系，随着云量的上升，系统风的贡献快速增加。

This paper mainly analyzes the changes of global sea land breezes (SLBs) from 1994 to 2014 by using the high-time resolution wind speed observation from the global automatic meteorological station observation network of the National Oceanic and Atmospheric Administration (NOAA). The paper focuses on the selection of three typical urban cities, which are Shanghai, Tokyo and Los Angeles, respectively, to discuss in detail how the urbanization, in situ solar radiation and large-scale westerlies influence SLBs. On the basis of the SLB signal being extracted at these single sites, the observation at 43 coastal sites around the world is selected further, and a method of extracting SLB signals under different climate regimes is developed. The significance of SLB under different climatic regimes around the world is described quantitatively using this method. Furthermore, according to the climatological trends of SLB, its quantified contribution to the global surface wind speed (SWS) variation in the form of both sea wind (SW) and land wind (LW) is given and the causes of differences between the contribution of SW and that of LW are analyzed at sites where the SLB’s being significant climatologically is verified. According to the differences among the SLB’s significance under different climate regimes and the trends of wind speed, the basic influential factors and many other indirect influential factors are discussed. The main contents of this study are as follows: 

1) Taking Shanghai which has been undergoing fast urbanization as an example, summertime data from 1994 to 2014 is selected for the study: in the context of global warming, the number of SLB days and SLB wind speed are decreasing significantly, which are caused by the combined effects of dynamic and thermodynamic factors. As far as thermodynamic factors are concerned, land temperature increases faster than ocean temperature at a large scale, which promotes the increase of SW and the decrease of LW. The warming effect of regional urbanization has the same effect. The blocking effect of urbanization, as a dynamic factor, eliminates the wind speed increase effect caused by the combined effect of other thermodynamic factors, and is the decisive factor of the continuous decrease of wind speed. With the help of multi-year variations of in-situ radiation, urbanization also serves as an important factor of the decrease of SLB day number. 

2) Taking Tokyo as an example, which has reached a high level of and stable urbanization rate, the summertime data during the same period as Shanghai is selected to study: there are no obvious decreasing or increasing trends of SLB speed and SLB day number. Also, there is no corresponding relationship between the SLB day number and the SLB speed. If the wind direction of large-scale background wind field is the same as the direction of SLB, it is beneficial to the increase of SLB day number, and vice versa. SLB is an important medium between in situ solar radiation and visibility. The daily circulation intensity of SLB is greater under high solar radiation scenario than medium and low scenarios, with the help of urban heat island circulation, its outward and upward transport capacity of atmospheric pollutants in urban areas is increased. 

3) Taking Los Angeles as an example, which belongs to Mediterranean climate regime, and has maintained a high level of urbanization. The data of four seasons during the same period is selected to study: the annual SLB speed has no obvious variation trend, such as being decresing or increasing continuously, but the SW and LW speed have significant seasonal variation differences, which shows that the SW is strong in summer and weak in winter, and the LW has no obvious and unified seasonal variation. The seasonal variation of in situ solar radiation is strong due to the different periods of rain seasons and hot seasons over Mediterranean areas, which is the decisive factor of seasonal variation of SW speed. The annual variation of the lowest daily air temperature (DLT) is weak, which is the secondary cause of the small seasonal variation of LW wind speed. The main reason is that the upper layer westerlies under Mediterranean climate regime are much stronger in winter, which promote the enhancement of LW circulation, and narrow the gap between wind speed between summer and winter. 

4) Taking 43 sites around the world which belong to different climate regimes as examples: it has been found that the regional wind fields can be classified into 3 master modes. 1) Sites where the SLB is the only component of the master mode of local wind field (the master mode of local wind field is ‘SLB only’); 2) Sites where both the SLB and system wind are components of the master mode of local wind field (the master mode of local wind field is ‘SLB and system wind’). Note that the system wind represents any wind from synoptic scale systems or planetary scale systems; 3) Sites where SLB is not a component of the master mode of local wind field and there is no significant SLB (the master mode of local wind field is ‘No obvious SLB’). Among these three modes, the first two types represent that the SLB is significant, or in other words, the obvious existence of SLB (OE-SLB) can be verified. The SLB’s significane under certain climate regime can be represented by the ratio of the number of the sites where OE-SLB is verified to the number of all sites under the same climate regime. Climate regime is the decisive factor of the differences among the SLB’s significance. Specifically, SLB’s significance is relatively high under Mediterranean climate (72.7%), tropical monsoon climate (75%), savanna climate (66.7%) and tropical desert climate (75%) but low under subtropical monsoon climate (22.2%), rainforest climate (33.3%), marine climate (0%) and continental climate (0%). There is no uniform trend of global SLB speed, and the SW speed is generally greater than LW speed. The SLB speed over Mediterranean climate areas is the most stable, and the tropical desert climate site has the reversed SW and LW speeds, which means that the LW speed is greater than the SW speed. Under the climate regimes where SLB’s significance is at a high level, ocean current, special topography, special water surface distribution and water surface area will hamper the verification of OE-SLB. Under these special circumstances, some abnormal sites can be found. 

5) Selecting 18 sites from the above 43 sites where SLB exists significantly and taking them as examples: when both SLB and system wind are components of the master mode of local wind field, both of them are the main causes for SWS variation. Except for monsoon climate, system wind still plays the dominant role of SWS variation. At sites where SLB is the sole component of master mode, the SLB contributes 82% of the SWS variation. In this case, the contribution of the system wind is between 2.65% and 17.16%. The specific value and the local cloud fraction show a good quadratic relationship with each other. The contribution of system wind increases fast as the local cloud fraction increases.