利用MODIS云检测产品客观确定AIRS云检测

云检测是用卫星资料研究云对气候系统作用至关重要的第一步,采用Nagle的基于扫描几何特性的共同空间匹配算法,用MODIS云检测产品客观确定了AIRS云检测,是将高空间分辨率成像仪产品与高光谱分辨率传感器观测资料结合使用的一次实际应用。合成使用高空间分辨率成像仪与高光谱分辨率传感器观测资料的关键步骤在于:①有效而精确的时间和空间上的匹配;②传感器像素内图像仪云检测产品的断定。不仅是云检测,利用AIRS视场内空间匹配的MODIS云产品,还可以判断AIRS的云相态、有效云量及云光学厚度等。     

基于MODIS云产品的AIRS云相态的识别和云量产品的检验

选用中国中东部(95～130 °E, 20～60 °N)午后同一时刻相对应的AIRS与MODIS各15景样本数据,将MODIS在红外波段上四种基于亮温的云识别算法对AIRS晴空像元和云像元中不同的云相态进行区分,分别统计出AIRS不同类型像元中这四种云识别算法的亮温差异,根据AIRS在晴空、水云和冰云三种不同相态下亮温的差异,确定出其阈值,实现AIRS晴空和云相态识别。在此基础上,对AIRS不同云相态下反演的有效云量进行检验和改进。结果表明:运用亮温阈值法对AIRS进行云相态识别能够较好地反映AIRS的晴空和云相态特征,云相态分布与MODIS云相态产品对应效果较好,特别在云边缘区域。对AIRS不同云相态下反演的有效云量进行对比和误差分析后发现:当水云有效云量较高、冰云有效云量较低时,AIRS反演的有效云量误差较大。在误差分析的基础上,提出了AIRS有效云量的偏差订正算法,对AIRS反演的有效云量具有一定改进,为AIRS云产品的应用提供参考依据。      

云检测在高光谱大气红外探测器 辐射率直接同化中的应用

对于高光谱大气红外探测器AIRS(Atmospheric Infrared Sounder)辐射率的同化,云检测技术是关键的一步.本文借鉴G01dberg的云检测的思想,通过AIRS通道和相应微波通道的经验组合来进行云检测,此云检测方案不需要对通道进行偏差订正,并且除背景场海表温度,不依赖于大气的先验信息,是一个快速简...     

高光谱大气红外探测器AIRS资料云检测及晴空通道应用技术初步研究

目前GRAPES-3DVar(全球\区域同化预报系统)使用的AIRS云检测方案只能够检测出完全晴空时的视场资料,对于受到云污染的视场,所有通道资料全部被剔除。研究表明有云视场当中的云顶以上资料对数值预报的影响更加重要,研究也更加有意义。因此,文章借鉴McNally和Watts云检测方案,结合GRAPES-3DVar系统和仪器特征,建立了一种适合于GRAPES-3DVar模式的晴空通道检测方案。该方案不但可以检测出完全晴空的视场资料,同时也可以检测出有云视场当中不受云影响的云顶以上通道资料;另外,还可以求出云顶高度,判断出高、中和低云。然后,将AIRS观测资料分别用晴空视场检测方案和晴空通道检测方案进行云检测,从结果分析看,晴空视场检测方案可以检测出的晴空资料占总资料的9.14%,而晴空通道检测方案检测出的晴空资料达到了34.86%,是传统云检测方案的3.8倍,大大增加了资料的使用度。     

两种红外晴空辐射订正方法及其反演大气参数的比较

比较了目前开发的两种红外晴空辐射订正算法：1）匹配的红外和微波探测值融合的算法,如AIRS／AMSU晴空订正算法（大气红外探测器AIRS,先进的微波探测器AMSU）;2）匹配的红外成像与探测值融合的算法,如AIRS／MODIS晴空订正算法（中分辨率光谱成像仪MODIS）。结果表明,相对于MODIS的晴空观测值而言,AIRS／MODIS晴空订正算法优于AIRS／AMSU晴空订正算法,而且AIRS／MODIS晴空订正辐射值反演的大气温湿廓线与欧洲中期天气预报中心（ECMWF）的客观分析场更接近;尽管这两种晴空订正算法本质上相当不同,但都是有效且互补的,当微波探测资料不能获得时,融合红外成像仪和探测器资料就成为唯一可靠的、间接使用有云影响的红外观测值的办法。     

基于带约束项广义变分同化AIRS云影响亮温研究

经典变分同化基于误差服从高斯分布理论,在同化受云影响的红外探测器通道亮温时,需进行云检测或只同化权重函数峰值位于云顶之上的通道亮温。在云检测过程中需对亮温进行严格的质量控制以剔除"离群值",导致丢失大量有用数据。文中基于带约束项非高斯模型的广义变分对高光谱大气红外探测器(Atmospheric Infra-Red Sounder,AIRS)受云影响亮温进行了初步的同化研究。在执行过程中,首先,动态选择AIRS通道形成通道子集。其次,把云参数(有效云量和有效云顶气压)作为辐射传输模式输入变量参与变分同化极小化迭代并用于通道子集亮温模拟。同化试验结果表明,对于高云,基于Cauchy-估计的变分同化反演结果最好,而对于中云和低云,基于Huber-估计得到了较好的同化反演结果。然而,在反演模式高层温度时,基于Fair-估计反而得到了较差的结果,但其对于湿度反演效果较为理想,其结果可能与Fair-估计分布的固有特点有关。带约束项广义变分同化方法对受云影响亮温的同化效果比经典变分方法的好,但依赖于M-估计的选取。     

基于光谱分析和综合阈值的FY-3/MERSI云检测

云检测是卫星遥感影像应用中的重要环节。为了获取FY-3中分辨率光谱成像仪(Medium Resolution Spectral Imager,MERSI)传感器精确的云掩膜数据,在分析云与植被、水体、裸露地表等地物在不同波段光谱差异的基础上,结合FY-3/MERSI数据的通道特性,构建基于多光谱和综合阈值的云检测算法。采用热红外通道亮温(CH_5)检测出高云,使用云检测指数(Normalized Cloud Detection Index,I_(NCD))和可见光波段反射率(CH_3)检测中低云。利用江西地区的FY-3/MERSI数据进行应用,将检测结果与其他算法、人机交互解译结果进行了对比分析。结果表明,该算法对FY-3/MERSI影像上的云区具有较好的检测效果,同时利用地面气象观测资料进行验证,检测精度均高于88%,能够满足当前云检测精度需求。     

基于深度学习的FY3D/MERSI和EOS/MODIS云检测模型研究

针对FY3D/MERSI和EOS/MODIS的云检测问题,提出了一种基于深度学习技术的全自动云检测算法,首次将深度学习引入到卫星影像云检测领域。本算法使用深度全卷积神经网络(Deep Convolutional Neural Networks)作为核心结构,基于EOS/MODIS基本云检测原理选择合适的通道作为特性向量参数,针对不同的场景进行分类和网络模型的训练,最终得到基于深度学习的云检测模型。经过EOS/MODIS数据和FY3D/MERSI数据的测试,云检测的精度达到98%以上,可以看出基于深度学习的云检测算法能够用于云检测,该算法具有效率高、精度高等特点,云检测效果理想。     

MODIS图像的薄云透明云检测

基于云与土壤、植被、冰雪、水体等目标的光谱差异,本文结合MODIS影像的波段范围,提出了一种对不同下垫面通用的多光谱云检测算法。同时,本文还加入了大气可降水量指标,用以对基本云区进一步检测,最终实现薄云、透明云的检测。试验证明:通过水汽反演结果的不同,薄云、透明云能够被有效检测。     

中国区域云特性分析及在FY-2云检测中的应用

云检测中所使用的云检测阈值正确与否是关系到云检测精度的重要因素。该文利用1983年7月-2007年6月ISCCP数据对覆盖我国及周边区域不同云类云顶温度的年、日变化特征进行分析, 得到云顶温度的分布特征; 对晴空下垫面与最暖云云顶温度差随纬度分布特征、不同区域晴空下垫面及云顶温度与晴空地表温度差日变化特征进行分析, 这些特征及每3 h的晴空下垫面24年平均亮温作为云检测算法的背景场, 用以判识实时动态提取云检测阈值的合理性。个例分析表明:利用多年平均晴空下垫面温度及最暖云云顶温度与晴空下垫面温度之差, 可有效识别云检测阈值的合理性, 合理的阈值有助于提高连续多日被云覆盖及冰雪下垫面条件下的云检测精度。     

基于加密探空资料的上海不同云天下AIRS/Aqua相对湿度廓线反演精度研究

目前云对卫星相对湿度廓线反演精度的影响研究大多是针对云量,对其他云属性的影响研究尚少,云高也是影响卫星相对湿度廓线反演精度的重要因素。利用上海宝山站L波段(1型)加密探空资料,分析了上海地区7—9月不同质量控制标识、云量和云顶高度条件下大气红外探测器AIRS/Aqua (Atmospheric Infrared Sounder) 相对湿度廓线的反演精度,以期为今后开展AIRS等卫星资料的同化研究提供科学依据。结果表明:(1)AIRS相对湿度廓线反演误差随着云量的增加而逐渐增大,并且随着气压值的升高,少云与多云时的均方根误差(Root Mean Squared Error, RMSE)之差有逐渐增大的趋势;(2)云顶高度越高,AIRS相对湿度廓线反演精度越差,云顶以上湿度廓线反演精度更高,而云顶以下高度的反演误差较大;(3)高云且多云时,AIRS相对湿度廓线的反演精度最差,850 hPa处,AIRS相对湿度反演数据与探空资料绝对误差的下限达到了[-63.51%];(4)虽然质量控制标识为0时,AIRS湿度廓线在对流层范围内的反演精度仍达不到无线电探空的水平,但是相对于质量控制标识1时,反演精度明显提高。      

STUDY ON CLOUD CLASSIFICATIONS BY USING AVHRR,GMS-5 AND TERRA/MODIS SATELLITE DATA

This paper presents the automated pixel-scale neural network classification methods beingdeveloped at National Satellite Meteorological Center(NSMC)of China to classify clouds by usingNOAA/AVHRR and GMS-5 satellite imageries.By using Terra satellite MODIS imageries,anautomated pixel-scale threshold technique has been developed to detect and classify clouds.Thestudy focuses on applications of these cloud classification techniques to the Huaihe River and theChangjiang(Yangtze)River drainage basin.The different types of clouds show more clearly onthis cloud classification image than single band image.The results of the cloud classifications arethe basis of studying cloud amount,cloud top height and cloud top pressure.Cloud mask methodsare widely used in SST,LST,and TPW retrieval schemes.Some case studies about cloud maskand cloud classification in satellite imageries,which are related with the study of Global Energyand Water Cycle Experiment(GEWEX)in the Huaihe River and the Changjiang River drainagebasin are illustrated.     

红外高光谱资料模拟大气廓线反演对云的敏感性

用不同云顶高度和不同有效云量时的星载红外高光谱观测值,模拟大气温湿垂直廓线反演对云参数的敏感性。用特征向量统计反演法按照云顶高度和有效云量分类反演了大气温湿廓线,统计了不同云高（200,300,500,700和850hPa）及云高有50hPa误差时温度、水汽混合比反演的均方根误差随有效云量的变化。结果表明,随着有效云量的增大,云顶以下各高度层上的温度、水汽混合比反演误差都明显增大。云高有50hPa误差较准确已知而言,温度和水汽混合比的反演误差增大,但温度反演对云高误差的敏感性比水汽反演要高。     

Ozone Profile Retrieval from Satellite Observation Using High Spectral Resolution Infrared Sounding Instrument

I. IntroductionOzone p1ays a very 1mportant ro1e in globa1 climate change. This is particularly true inthe stratosphere, where ultra--v1olet solar radiation is strongly absorbed by ozone, leading tosubstantlal change in the earths atmospheric thermal, physical and chemical structure. Al-though the troposphere contains only about l0% of the total atmospheric ozone, the varia-tlon of tropospheric ozone may have more significant cllmatic effect than stratospheric ozoneon the earth's surface temP…     

STUDY ON CLOUD CLASSIFICATIONS BY USING AVHRR,GMS-5 AND TERRA/MODIS SATELLITE DATA*

This paper presents the automated pixel-scale neural network classification methods being developed at National Satellite Meteorological Center(NSMC) of China to classify clouds by using NOAA/AVHRR and GMS-5 satellite imageries.By using Terra satellite MODIS imageries,an automated pixel-scale threshold technique has been developed to detect and classify clouds.The study focuses on applications of these cloud classification techniques to the Huaihe River and the Changjiang(Yangtze) River drainage basin.The different types of clouds show more clearly on this cloud classification image than single band image.The results of the cloud classifications are the basis of studying cloud amount,cloud top height and cloud top pressure.Cloud mask method sare widely used in SST,LST,and TPW retrieval schemes.Some case studies about cloud mask and cloud classification in satellite imageries,which are related with the study of Global Energy and Water Cycle Experiment(GEWEX) in the Huaihe River and the Changjiang River drainage basin are illustrated.     

Validation of the Atmospheric Infrared Sounder Retrieval Products over China and Their Application in Numerical Mode

The atmospheric infrared sounder (AIRS) instrument onboard Aqua Satellite is a high spectral resolution infrared sounder. In recent years, AIRS has gradually become the primary method of atmospheric vertical observations. To examine the validation of AIRS retrieval products (V3.0) over China, the AIRS surface air temperature retrievals were compared with the ground observations obtained from 540 meteorological stations in July 2004 and January 2005, respectively. The sources of errors were considerably discussed. Based on the error analysis, the AIRS retrieved surface air temperature products were systemically corrected. Moreover, the AIRS temperature and humidity profile retrievals were compared with T213 numerical forecasting products. Because T213 forecasting products are not the actual atmospheric states,to further verify the validation, the AIRS temperature and humidity profile products were assimilated into the MM5 model through the analysis nudging. In this paper, the case on February 14, 2005 in North China was simulated in detail. Then, we investigated the effects of AIRS retrievals on snowfall, humidity field,vertical velocity field, divergence field, and cloud microphysical processes. The major results are: (1) the errors of AIRS retrieved surface air temperature products are largely systematic deviations, for which the influences of terrain altitude and surface types are the major reasons; (2) the differences between the AIRS atmospheric profile retrievals and T213 numerical prediction products in temperature are generally less than 2 K, the differences in relative humidity are generally less than 25%; and (3) the AIRS temperature and humidity retrieval products can adjust the model initial field, and thus can improve the capacity of snowfall simulation to some extent.     

Validation of the Atmospheric Infrared Sounder Retrieval Products over China and Their Application in Numerical Model

The atmospheric infrared sounder (AIRS) instrument onboard Aqua Satellite is a high spectral resolution infrared sounder. In recent years, AIRS has gradually become the primary method of atmospheric vertical observations. To examine the validation of AIRS retrieval products (V3.0) over China, the AIRS surface air temperature retrievals were compared with the ground observations obtained from 540 meteorological stations in July 2004 and January 2005, respectively. The sources of errors were considerably discussed. Based on the error analysis, the AIRS retrieved surface air temperature products were systemi-cally corrected. Moreover, the AIRS temperature and humidity profile retrievals were compared with T213 numerical forecasting products. Because T213 forecasting products are not the actual atmospheric states, to further verify the validation, the AIRS temperature and humidity profile products were assimilated into the MM5 model through the analysis nudging. In this paper, the case on February 14, 2005 in North China was simulated in detail. Then, we investigated the effects of AIRS retrievals on snowfall, humidity field, vertical velocity field, divergence field, and cloud microphysical processes. The major results are: (1) the errors of AIRS retrieved surface air temperature products are largely systematic deviations, for which the influences of terrain altitude and surface types are the major reasons; (2) the differences between the AIRS atmospheric profile retrievals and T213 numerical prediction products in temperature are generally less than 2 K, the differences in relative humidity are generally less than 25%; and (3) the AIRS temperature and humidity retrieval products can adjust the model initial field, and thus can improve the capacity of snowfall simulation to some extent.     

寿县地区云中液态水含量的微波遥感

该文利用地基双频（22.235GHz（1.35cm）和35.3GHz（8mm））微波辐射计和热带降雨测量卫星微波成像仪（TRMM/TMI）两种被动微波遥感资料分别反演安徽寿县地区云中液态水含量，运用统计回归方法得到该地区云水含量的反演公式，揭示了该地区云中液态水含量存在一个阈值0.4mm，当云中液态水含量超过这一阈值时，该地区一般就会出现降水。在降水发生之前，水汽和云中液态水含量均会出现一个显著增加的过程，而在降水结束之后则会出现一个显著减少的过程。     

太赫兹频段雷达探测的冰云微物理参数反演算法模拟研究

在假设冰云粒子呈球形以及粒子谱服从对数正态分布的条件下,利用离散偶极子近似法(DDA),计算出太赫兹频段(220 GHz)冰云粒子的雷达反射率因子,及其与瑞利假设下雷达反射率因子的比值。忽略衰减和多次散射的影响,根据太赫兹波段冰云的雷达反射率因子,基于最优估计理论反演冰云的微物理参数,并验证该算法的可靠性。反演结果表明,当冰云粒子大小在设定的尺度范围内时,有效粒子半径(r_e)的反演误差小于4%,粒子谱宽(σ)的误差小于2.5%、粒子数密度(N_T)的误差小于1%,冰水含量(IWC)的误差小于5%。还分析了当N_T和σ为定值时,反演结果随粒子尺寸的变化情况,当冰云粒子尺寸在模拟计算设定的范围内时,r_e的反演误差小于0.04%,σ的反演误差小于0.02%,N_T的反演误差小于0.50%,IWC的反演误差小于0.08%,如果冰云粒子大小超出模拟计算设置的范围,反演误差随着r_e增加而增大。该结果证明了基于最优估计理论反演得到的冰云微物理参数与模拟设定值有良好的一致性,说明该方法可应用于太赫兹频段云雷达的冰云观测及云微物理参数的反演和研究。