【作者】 郝增周；

【导师】 潘德炉； 孙照渤；

【作者基本信息】 南京信息工程大学 ， 气象学， 2007， 博士

【摘要】 海雾发生使海面能见度下降，给船只航行、人类生产和生活带来隐患。海洋上常规地面监测站点稀少，难以实现大范围的同步观测，幸运的是，卫星遥感为海雾研究提供了大面积、实时动态监测的重要技术。本研究利用14轨AVHRR／NOAA17已知黄、渤海海雾的先验数据和辐射传输模式模拟数据，对海雾及其他目标物(水体、中、高云系等)的遥感辐射特性进行分析，发掘海雾的主要辐射特性；在此基础上，提出了一个富有物理意义的海雾监测算法，全面探索了五个海雾微物理辐射特性的监测反演技术；分析2006年1月至5月长时间NOAA17数据的监测结果验证了海雾监测算法的精度，并指出了算法的不足，最后对我国自主气象卫星FY-1D进行了算法典型应用。得到了以下主要结论：1．海雾在可见光、近红外波段上满足关系：Ch1＞Ch3a＞Ch2，甚至出现Ch3a＞Ch1＞Ch2的辐射特性。这一辐射特性主要由雾滴尺度与三波段的相对大小引起。该辐射特性的发现为海雾监测和小粒径的云滴判别提供了重要科学依据。2．近红外通道Ch3a在中、高云上的不同表现：高云，Ch2＞＞Ch3a；中低云，Ch2＞Ch3a，说明Ch3a反射率特性可用于云位相的判定。3．验证了可见光通道Ch1的反射率变化随云雾层光学厚度的变化较大，近红外通道Ch3a的反射率是云雾滴粒子大小的函数，为联合通道Ch1和Ch3a反射率信息，监测反演海雾的微物理特性提供了基础。4．基于辐射特性的分析，设计了一个富有物理意义的黄、渤海海雾监测算法：云地分离、位相判别、粒径判断、图像特征分析、高度分析和修补漏点。通过先验数据和连续5个月数据的监测结果分析，验证了该算法对黄、渤海海雾具有大范围、近实时的监测能力；对其他卫星FY-1D数据的典型应用试验，表明该算法具有普适性。5．分析海雾有效粒径、雾层光学厚度、雾中能见度和液水含量之间的关系，在假定雾滴谱分布模型的前提下，建立了单位体积内雾滴数与能见度、液水含量间的关系，通过事先建立的查询表与实际观测的可见光、近红外反射率信息监测反演了五个海雾微物理特性。6．通过长时间序列的数据验证表明，该算法对沙尘影响海域，高云遮挡下的海雾区，以及部分低云边缘区会产生错判、漏判现象，为今后算法的改进提供了努力方向。本论文的主要创新点：1．对实际卫星观测数据和辐射模拟数据分析不同目标物的遥感辐射特性，发现海雾在可见光、近红外波段的反射率满足关系：Ch1＞Ch3a＞Ch2，甚至Ch3a＞Ch1＞Ch2，为海雾监测，小粒径云滴的判别提供了重要科学依据。2．根据不同目标物的辐射特性，提出了一个富有物理意义、具有普适性的黄、渤海海雾监测算法。经长时间序列卫星数据的监测应用，验证了该算法具有监测海雾事件的能力，并在FY-1D卫星数据上得到了成功典型应用。3．首次在国内对海雾微物理特性进行了探索性监测反演，为填补海雾微物理性质数据的空缺提供了技术支持。更多还原

【Abstract】 The horizontal visibility at the surface of the ocean decreases to less than 1km,when the sea fog occurs. It takes some difficulty to the ship safety, and some troubleto human production, living and wealth. Due to the measurement station is rare, thetraditional in situ measurement is difficult and unfeasible to observe the fogphenomenon in real-time or in large range. Luckily, satellite measurements providethe way and mean to detect the sea fog in near real-time and in large range.In this paper, fourteen AVHRR/NOAA17 satellite datum, which act as aforehandresearch events, and simulates by radiance transfer model (Streamer) are used to theresearch on radiance properties of sea fog and others. By comparing the differentobject observation and simulate radiance, it shows the major radiance characteristic ofsea fog. On the base of those, a sea fog detection algorithm, which is meaning inphysics, is proposed. And the retrieval of five microphysical parameters of sea fog isdiscussed fully. Using the AVHRR/NOAA17, the detection algorithm is run throughfive months (Jan to May, 2006). By comparing and analyzing, the precision of fogdetection is validated and the limitation is pointed. Lastly, the representativeapplication is successfully done on Chinese polar-orbit meteorology satellite FY-1D.Some significant results are revealed as follows:1. The reflectivity of sea fog at visible and near-infrared channels satisfies:Ch1＞Ch3a＞Ch2, or maybe Ch3a＞Ch1＞ch2. This radiance characteristic is mainlyconducted by the relative size of sea fog and three channels. This result providesimportance scientific argument for the sea fog detection and the judge of small sizecloud particles.2. The difference of mid- and high cloud at the near-infrared channel is that highcloud, Ch2＞＞Ch3a; mid cloud, Ch2＞ch3a. It shows that the reflectivity at channel 3aof NOAA17 is used to judge the cloud phase.3. The reflectivity at Ch1, which varies with the cloud optical thickness, and thereflectivity at Ch3a, which is the function of the size of cloud/fog particles, isapproved. Therefore, combining the reflectivity of Ch1 and Ch3a, the retrieval ofmicrophysical parameters of sea fog is possible and feasible. 4. Based on those radiance properties, an algorithm, which is used for sea fogdetection over the Yellow Sea and Bohai, is proposed. The algorithm is meaning inphysics by six steps: (1) separating of cloud and clear region; (2)judging the cloud phase; (3) differentiating of the size of particle; (4)analyzing the picture characteristic;(5)distinguishing the cloud height; and (6)repairing the lost pixels. By analyzing thedetections of the 14 aforehand events and five months continuous satellites, it showsthe algorithm has the ability of sea fog detection in large range and near-real time.The application tests for FY-1D satellites is successful. The wide applicability of thealgorithm is presented5. The relationship between the efficient radius, the fog optical thickness,visibility, fog water content is analyzed and deduced. Assuming the fog sizedistribution, the relation between the number of fog drop in unit volume and visibility,fog water content is created. Then, on the base of fog detection, linking the satellitereflection of Ch1 and Ch3a, the five fog microphysical parameters is derived by LUT.6. The detection of sea fog over the Yellow Sea and Bohai for a long time isanalyzed. The fog detection is weak for: (1) the ocean region affected by Asia dust; (2)the fog region sheltered by high cloud; (3) the edge of some low cloud. It gives thedirection of improving on fog detection in the future.Some innovations is listed as follow:1. After analyzing the radiance characteristic of the different object by satellitemeasurements and model simulations, the reflectivity of sea fog at visible andnear-infrared channels satisfies: Ch1＞Ch3a＞Ch2, or maybe Ch3a＞Ch1＞ch2 is found.Those provides the important science base for the detection of sea fog and the judgeof small size cloud particles.2. Based on those radiance properties, an algorithm, which is meaning in physicsand wide-application for others, is proposed to detect sea fog events over the YellowSea and Bohai. After applying for satellite measurements in a long time, it shows thatthe algorithm has the capability for detecting sea fog event. And it have get successfulapplication test for FY-1D.3. The fog microphysical parameter is first studied exploringly in home. Itgives important technology for filling up microphysical parameter which is difficultto measure over ocean.更多还原