RETRIEVING OVER-OCEAN PRECIPITABLE WATER FROM THE SSM/I MEASUREMENTS WITH SEVERAL REGRESSION ALGORITHMS*

A collocated SSM/I and radiosonde measurement data set provided by the NASDA(Japan) was used to retrieve the total precipitable water(PW) over oceans.The retrieval results obtained with several regression algorithms were compared against the radiosonde measurements.It is shown that:(a) the routinely operational algorithm of Alishouse et al.(1990) yields significant underestimation in high PW regime and overestimation in low PW regime;(b) a cubic correction by Colton and Poe(1994) is not sufficient and globally improves slightly the retrieval results;and(c) the regression algorithm with the form of brightness temperature(T_b) function In(280-T_b) gives a little largely scattered retrievals in whole PW range but without considerable over-and underestimates in low and high PW regimes.To improve the estimation of the oceanic precipitable water from the SSM/I measurements,a composite algorithm with different forms of T_b function in low.medium and high PW regimes is proposed and tested.     

青藏高原探空大气水汽偏差及订正方法研究

水汽是大气的主要成分和降水的主要物质来源.青藏高原大气水汽分布对区域天气和气候有很大影响,为了探讨探空观测的大气水汽总量（R）资料的可靠性,本文以地基GPS遥感的大气水汽总量（G）为参照标准,对拉萨（1999~2010年）和那曲（2003年）的R进行对比分析和偏差（R-G）订正.结果表明:近10多年拉萨站R比G明显偏小,偏小程度随使用不同的探空仪而异.GZZ-2型机械探空仪和GTS-1型电子探空仪多年平均的PW偏差分别为-8.8%和-3.9%,随机误差分别为17.6%和13.6%.近10多年PW偏差变化呈减少趋势,这与探空仪性能改进有关.分析发现,青藏高原PW偏差具有明显季节变化和日变化特征,夏季比冬季明显,1200 UTC比0000 UTC明显.拉萨站GZZ-2型和GTS-1型探空仪在1200 UTC多年平均的PW偏差分别为-15.8%和-7.3%,在0000 UTC分别为-1.6%和-0.4%.那曲站GZZ-2型探空仪在1200 UTC和0000 UTC的PW偏差分别为-12.4%和-0.3%.分析还表明,太阳辐射加热与气温的日变化和季节变化是造成高原PW偏差日变化和季节变化的重要原因.据此,提出了高原PW偏差的订正方法,并以拉萨和那曲站为例进行PW偏差订正,订正后的PW系统偏差显著减少,随机误差也相应得到了改善.     

空基微波辐射计遥感晴天大气可降水量：不同通道组合和亮温函数形式的效果的比较分析

为了监测全球海洋上空的大气可降水量，已发射上天多种星载微波波谱仪和辐射计，相应地发展建立了多种反演算式。我们利用逐步回归分析，研究比较了反演晴天大气可降水量（PW）的多种通道组合以及算式中不同的亮温（Ta）函数形式的效果。主要结果有：（1）对应于高中低PW值有不同的最佳通道组合；（2）常选的水汽吸收中心线22.235 GHz不太适合于反演高中且变化范围大的PW值；（3）有必要建立分气候区域和分季节的反演算式；（4）在反演算式中采用ln(T0-TB)一般比线性形式有更好的回归和反演效果，但在有22.235 GHz通道时则不然；（5）反演中低或变化范围大的PW时，22.235 GHz亮温的平方项能显著改善回归和反演结果。     

Remote sensing of cloud liquid water

Summary A method is presented to infer cloud liquid water path (LWP in kg/m²) over the ocean from passive microwave measurements of SSM/I. The algorithm to retrieve LWP is based on simulated satellite observations. They are calculated with a radiative transfer model applied to about 3000 radiosonde ascents over the Atlantic Ocean. Since radiosonde observations do not contain direct information about cloud water and ice, these parameters are parameterized based on relative humidity and temperature using modified adiabatic liquid water density profiles. A multiple linear regression is applied to the simulated radiances and the calculated LWP to derive the algorithm. The retrieval accuracy based on the regression analysis including instrumental noise is 0.03 kg/m². Validation of the LWP-algorithm was pursued through a comparison with measurements of a ground-based 33 GHzmicrowave radiometer on board of R.V. Poseidon during the International Cirrus Experiment 1989 at the North Sea (ICE'89). The LWP values agree within the range of uncertainty caused by the different sampling characteristics of the observing systems. The retrieval accuracy for clear-sky cases determined using colocated METEOSAT data over the North Sea is 0.037 kg/m² and confirms the accuracy estimated from regression analysis for the low liquid water cases.The algorithm was used to derive maps of monthly mean LWP over the Atlantic Ocean. As an example the Octobers of the 5 years 1987–1991 were selected to demonstrate the interannual variability of LWP. The results were compared with the cloud water content produced by the climate model ECHAM-T2 from the Max-Planck-Institut Hamburg.Observations during ICE'89 were used to check the accuracy of the applied radiative transfer model. Brightness temperatures were calculated from radiosonde ascents launched during the overpass of DMSP-F8 in cloud-free situations. The channel-dependent differences range from about –2 to 3 K.The possibility to identify different cloud types using microwave and infrared observations was examined. The main conclusion is that simultaneous microwave and infrared measurements enable the separation of dense cirrus and cirrus with underlying water clouds. A classification of clouds with respect to their top heights and LWP was carried out using a combination of SSM/I derived LWP and simultaneously recorded Meteosat IR-data during ICE'89.With 11 Figures     

两种探空仪观测湿度垂直分布及其应用比较

对2010年8月在云南腾冲利用芬兰Vaisala RS80和低温霜点仪 (Cryogenic Frostpoint Hygrometer,CFH) 两种探空仪测量大气湿度的垂直分布进行对比分析,同时比较它们白天和夜间测量误差的差别,并对国产GTS1,RS80和CFH共3种探空仪测量水汽总量与地基GPS遥测结果进行比较。结果表明:RS80湿度测值在整个对流层比CFH测值偏干 (23.7±18.5)%;因太阳辐射白天RS80偏干较夜间更明显,比夜间偏干 (13.5±14.8)%。而在对流层上层向平流层过渡区域内RS80湿度数据基本无效。CFH在低温、低湿环境下对湿度能有效测量,但在湿度较高的对流层低层测值偏高,导致比较中CFH水汽总量平均比GPS遥测的水汽总量偏高 (4.3±2.0) mm (样本数为11),而RS80,GTS1与GPS的水汽总量差别分别是 (0.2±1.4) mm (样本数为12), (-0.2±2.2) mm (样本数为43)。地基GPS遥测的水汽总量对对流层上层至平流层的水汽变化不敏感。由于RS80测量相对湿度在高空偏低,通过RS80相对湿度测值来确定中、高云结果是偏低的,特别是对6000 m以上的高云判别上,RS80相对湿度的探测几乎很难甄别到云的存在。     

青藏高原大气可降水量单站观测对比分析

为了探讨青藏高原大气可降水量观测资料的可靠性,对2015年6-9月西藏申扎、改则和那曲3站地基GNSS遥感的大气可降水量、同址探空观测的大气可降水量、风云三号可见光红外扫描辐射计反演的晴空大气可降水量、MODIS大气可降水量和NCEP可降水量进行对比分析。结果表明:探空可降水量和地基GNSS可降水量的偏差较小,均低于2.5 mm。风云三号可降水量明显偏低,与其他观测结果的偏差超过6 mm。全自动探空可降水量离散程度较L波段探空大,均方根误差超过4 mm。      

中国区域大气加权平均温度的时空变化及模型

大气加权平均温度T_m是地基GPS水汽遥感的关键参数,决定了水汽反演的精度。利用2008—2011年全国123个探空站点资料,分析了T_m与其影响要素纬度、海拔、地面气温、水汽压及大气压之间的关系,结果表明:T_m与纬度和海拔随季节变化呈周期性负相关,与地面温度和水汽压的自然对数呈正相关,与大气压呈负相关;T_m的空间变化具有纬度地带性和明显的气候分布特征,其变异程度在空间分布上有显著差别,不同地理位置的T_m受季节性影响程度不一,T_m也具有明显的年际周期性变化,年内T_m的每日变化符合二次函数分布规律。按照全国、气候分区和季节分区方法,分别建立了T_m单因子和多因子回归模型,并利用2012年1—5月数据对所建模型进行验证,T_m的估算误差能满足GPS水汽遥感2%的精度,模型普遍适用于我国地基GPS水汽遥感中T_m的估算。     

Precipitable water as a predictor of LCL height

Based on the precipitable water observations easily available from in situ and remote sensing sensors, a simple approach to define the lifting condensation level (LCL) is proposed in this study. High-resolution radiosonde and microwave radiometer observations over peninsular Indian region during the Cloud Aerosol Interaction and Precipitation Enhancement Experiment Integrated Ground Observational Campaign (CAIPEEX-IGOC) during the monsoon season of 2011 are used to illustrate the unique relationship. The inferences illustrate a linear relationship between the precipitable water (PW) and the LCL temperature. This relationship is especially valuable because PW is easily available as a derived parameter from various remote sensing and ground-based observations. Thus, it could be used to estimate the LCL height and perhaps also the boundary layer height. LCL height and PW correlations are established from historical radiosonde data (1984–2012). This finding could be used to illustrate the boundary layer-cloud interactions during the monsoon and is important for parameterization of boundary layer clouds in numerical models. The relationships are illustrated to be robust and seem promising to get reasonable estimates of the LCL height over other locations as well using satellite observations of PW.

     

PRECIPITABLE WATER MEASUREMENTS WITH SUN-PHOTOMETER

In this paper a method is described of retrieving precipitable water from sun-photometermeasurements.The quantitative relationship between water vapor transmission and precipitablewater is established by means of LOWTRAN 7 model.Calibration of the water vapor absorptionchannel is made through a modified Langley method.The good agreement between the sun-photometer and radiosonde water vapor retrieval indicates that this method is feasible.The sun-photometer is operated at Hefei to monitor the precipitable water within one yearperiod.Characteristics of both diurnal evolution and within-one-year variation of the precipitablewater and their relation with synoptic system as well as surface dew-point temperature arepresented and analyzed.Errors in the retrieved precipitable water from the sun-photometermeasurements are also calculated and discussed.     

地基GPS不同水汽反演方法的误差分析

利用湖北宜昌2007年观测的GPS对流层天顶延迟数据,对采用不同水汽反演方法计算的对流层可降水量PW的正确度和精密度进行对比分析。结果显示：不同天顶干延迟计算模型对GPSPW的精密度影响不大,但对其正确度有明显的影响,与探空PW相比,Hopfield模型计算的GPs册的平均偏差最小,Saastamoinen模型的平均偏差次之,而Black模型的平均偏差最大;大气加权平均温度对GPS PW的正确度有重要影响,对其进行本地化订正可以明显减小GPS PW与探空尸形的偏差,但对GPSPW的精密度影响不大;GPS PW与探空PW的相关性受大气水汽含量的影响,当大气水汽含量较低（PW≤65mm）时,两者的相关系数可达0．92,两者的平均偏差为3．8mm,偏差的均方差为6．4mm,而当大气水汽含量较大时,GPS PW与探空PW的偏差会增大,两者的相关系数会变小,这可能与GPS水汽反演方法有关;GPS PW比探空PW偏小,这可能是由两种探测方法的不同所造成的系统偏差。     

地基GPS遥测大气可降水量应用精度和范围

利用2005年6～8河北省石家庄、张家口两个地基GPS站反演得到的大气可降水量资料,对探空资料计算的大气可降水量和GPS反演的大气可降水量进行比较,并通过2007年7月17～19日河北中南部一次强降水过程的个例分析,发现GPS资料反演的大气可降水量略高于探空站资料的计算结果,但两种资料的可降水量计算结果变化趋势一致,GPS反演的大气可降水量具有较高的使用价值,但单站GPS反演的大气可降水量有效半径距离具有一定限度.     

香港市域大气加权平均温度模型构建及其应用

为提高地基反演大气可降水量中加权平均温度的计算精度和效率,以香港市域为例,根据2017年香港无线电探空资料,设计了一种以地面气压为基础的大气加权平均温度模型,并通过2014—2016年探空数据对该模型进行验证,分析表明该模型与探空数据得到的加权平均温度有良好的一致性,具有较高的精度。基于气压回归模型和气温回归模型对2017年7月香港地区进行地基反演大气可降水量,验证新建模型的水汽反演精度。结果表明,该模型能很好的满足地基反演大气可降水量的精度要求,相比于气温回归模型反演精度有了较好的提升。     

Linear retrieval and global measurement of precipitable water from the SEASAT SMMR data

Summary The Seasat Scanning Multichannel Microwave Radiometer (SMMR) measurements in the 18.0, 21.0 and 37.0 GHz channels, both horizontal and vertical polarizations, are primarily used for precipitable water, cloud liquid water content and rainfall rate determination. Linear regressions using a leaps and bounds procedure are used for the retrieval of precipitable water. The radiation simulated for all the ten SMMR channels with varied global environmental parameters were used for subset selection for water vapour retrieval. Only subsets with channels having uniform grid size (18, 21 and 37 GHz) were used for the analysis. A total of eight subsets using two to five frequencies of the SMMR are examined to determine their potential in the retrieval of atmospheric water vapour content. Our analysis indicates that the information concerning the 18 and 21 GHz channels are optimum for the water vapour retrieval. An attempt to use all the SMMR channels simultaneously gives no significant improvement. A comparison with the radiosonde observations gave an rms accuracy of 0.4 g/cm². The rms accuracy of retrieved precipitable water using different subsets was within 10 percent.Global maps of precipitable water over oceans using two and five channels retrieval are given. These maps are generated on a 10 day average basis as well as on monthly basis for the period 7 July to 6 August 1978. An analysis of these Global maps reveals the possibility of global moisture distribution associated with oceanic currents and large scale general circulation in the atmosphere. A stable feature of the large scale circulation is noticed. The precipitable water is maximum over the Bay of Bengal and in the North Pacific over the Kuroshio current and shows general latitudinal pattern.With 5 Figures     

A comparison of SSM/I and TOVS column water vapor data over the global oceans

Summary This paper presents a comparison of column water vapor (CWV) information derived from both infrared measurements as part of the TIROS-N Operational Vertical Sounder (TOVS) and Special Sensor Microwave/Imager (SSM/I) in an attempt to assess the relative merits of each kind of data. From the analyses presented in this paper, it appears that both types of satellite data closely reproduce the bulk climatological relationships introduced in earlier studies using different data. This includes both the bulk relationship between CWV and the sea surface temperature and the annual variation of CWV over the world's oceans. The TOVS water vapor data tends to be systematically smaller than the SSM/I data and when averaged over the ocean covered regions of the globe this difference is between 2–3 kgm^–2. Using a cloud liquid water threshold technique to establish clear sky values of SSM/I water vapor, we conclude that the differences between TOVS and SSM/I are largely a result of the clear sky bias in TOVS sampling except in the subsidence regions of the subtropics. The clear sky bias is considerably smaller than previously reported and we attribute this improvement to the new physical retrieval scheme implemented by NOAA NESDIS. While there is considerable agreement between the two types of satellite data, there are also important differences. In regions where there is drying associated with large scale subsidence of the atmosphere, the TOVS CWV's are too moist relative to both radiosonde and SSM/I data and this difference may exceed 10 kgm^–2. The explanation for this difference lies in the limitations of infrared radiative transfer. By contrast, in regions of deep convection, such as in the ITCZ, TOVS CWV is systematically lower than the SSM/I CWV. Both TOVS and SSM/I data demonstrate similar kinds of gross effects of large scale circulation on the water vapor except in these subsidence regions where TOVS data leads to an under-prediction of the effects of subsidence drying.With 11 Figures     

GMS-5 估计可降水量的研究

文章证明了由静止气象卫星GMS-5的分裂窗通道和水汽通道亮温反演可降水量的可行性,探讨了GMS-5红外通道亮温与可降水量的关系,建立了由3个通道亮温反演可降水量的经验公式。用60组大气平均廓线,对公式模拟检验误差为0.18 g/cm2,而用实际124组探空和对应的GMS-5亮温资料进行检验,误差0.40 g/cm2。用得到的经验公式可反演大范围的晴空可降水量分布。     

Rain rate estimation from a synergetic use of SSM/I, AVHRR and meso-scale numerical model data

Summary In this paper a retrieval technique for estimating rainfall rates is introduced. The novel feature of this technique is the combination of two satellite radiometers — the Special Sensor Microwave/Imager (SSM/I) and the Advanced Very High-Resolution Radiometer (AVHRR) — with mesoscale weather prediction model data. This offers an adjustment of the model atmospheres to reality which is necessary for calculating brightness temperatures that can be compared with microwave satellite measurements.In sensitivity studies it was found that the estimation of precipitation is determined to a high degree by the particle size distribution of rain and snow, especially by the size distribution of solid hydrometeors. These studies also reveal the influence of the knowledge of the correct cloud coverage inside a SSM/I pixel and the importance of using a realistic temperature profile instead of using standard atmospheres.The retrieval technique is based on radiative transfer calculations using the model of Kummerow et al. (1989). The algorithm consists of two parts: First Guess (FG) brightness temperatures for the SSM/I frequencies are generated as a function of the cloud top height and the cloud coverage, derived from AVHRR data and predictions from a meso-scale model. The rainfall rate of different types of clouds containing raindrops, ice particles and coexisting ice and water hydrometeors is then calculated as a function of the cloud top height. As an example, a strong convective rain event over the western part of Europe and over the Alps is taken to evaluate the performance of this technique. Good agreement with radar data from the German Weather Service was achieved. Compared to statistical rainfall algorithms, the current algorithm shows a better performance of detecting rainfall areas.With 12 Figures     

Physical retrievals of over-ocean rain rate from multichannel microwave imagery. Part II: Algorithm implementation

Summary A new physical inversion-based algorithm for retrieving rain rate over the ocean with the Special Sensor Microwave Imager (SSM/I) is described. In a departure from other rain rate retrieval algorithms, the satellite observables inverted in the present algorithm are not the raw brightness temperatures but rather normalized polarizations for 19.35, 37.0, and 85.5 GHz, plus an 85.5 GHz scattering index which is sensitive primarily to ice particles aloft. The normalized polarizations are interpreted as footprint-averages of theoretically derived analytic functions of the column optical depth associated primarily with liquid water. The effective vertical depth of the rain layer is specified as a function of the SSM/I estimated column water vapor.The retrieval algorithm performs an iterative search for a high resolution (12.5 km) rain field which is simultaneously consistent with the 19.35 and 37.0 GHz normalized polarizations. The first-guess rain rate field is supplied by the 85.5 GHz scattering index. At gridpoints for which the rain column optical depth exceeds the dynamic range of the attenuation-based indices, the first-guess field is left essentially unmodified; elsewhere, the required consistency with the 19 and 37 GHz indices usually results in significant modification of the scattering-based rain rate estimates.The algorithm as described here is a prototype implementation which was developed with reference only to idealized theoretical models; empirical improvements to the numerical scheme and the model coefficients will be made in the future as results from the first Precipitation [algorithm] Intercomparison Project 1 (PIP-1) and the second phase of the GPCP (Global Precipitation Climatology Project) algorithm Intercomparison Project (AIP/2) are analyzed, as well as data from individual validation efforts. Although the algorithm is physically based and uses all SSM/I channels, it is computationally much less demanding than cloud/radiative transfer model-based inversion algorithms published else-where.With 9 Figures     

GPS探测与FY-2反演大气可降水量对比分析

为深入了解FY-2卫星大气可降水量(PW)的反演质量,文章选取2012和2015年地基GPS水汽观测数据,与FY-2的PW反演产品进行了对比分析。结果表明:(1)北京、武汉和海口三站GPS/PW(PW_(GPS))与FY-2/PW(PW_(FY-2))在夏季存在显著正相关,三站的相关系数都达到0.67以上,夏季PW的均方根误差值、月平均偏差绝对值均小于冬季。北京与武汉站PW平均偏差和均方根误差在四季均具有明显日变化特征;(2)当PW_(GPS)20 mm时,北京、武汉、海口和拉萨站FY-2/PW与GPS/PW比较一致,PW偏差均值的绝对值和均方根误差较小,当PW_(GPS)20 mm时,PW偏差均值绝对值和均方根误差随PW_(GPS)值减小而迅速变大。FY-2的PW产品在夏季可以为大部分区域提供高时空分辨率、高精度的大气可降水量,在大气湿度非常低、冬季和夜间条件,反演结果精度有待提高。     

地基GPS遥感观测北京地区水汽变化特征

利用2004—2007年SA34（北京大学）站的GPS观测数据,运用GAMIT软件解算反演了间隔30min的连续变化大气水汽总量（PW）。与北京南郊观测场得到的探空结果作比较,均方根误差（RMSE）在2~3mm之间。通过对大气水汽作月平均,得到每月的大气水汽总量口变化曲线,并初步分析了夏季水汽日变化与地面比湿、降水、地面气温以及地面风矢量的关系。结果表明:北京地区夏季7月大气水汽总量最小值出现在08:00（北京时）左右,8月大气水汽总量最小值出现在08:00到12:00左右（各年表现出一定的差异）,夏季大气水汽总量的最大值出现在01:00到03:00;7月和8月的日变化在夜间变化趋势有所不同;大气水汽总量最大值出现时刻与地面小时降水有一定相关性,且大气水汽总量的日变化明显受风矢量日变化的影响。通过对大气水汽总量的时间序列进行小波分析,得到1年大部分时间里,水汽变化存在大约12d的周期。采用前期的大气水汽总量平均值和短时大气水汽总量增量两个条件进行降水的判断,认为夏季降水的出现时刻与差值的高值区有比较好的对应。     

An Operational Precipitable Water Vapor Retrieval Algorithm for Fengyun-2F/VISSR Using a Modified Three-Band Physical Split-Window Method

The Visible and Infrared Spin-Scan Radiometer(VISSR) onboard the Fengyun-2(FY-2) satellite can provide valuable thermal infrared observations to help create a precipitable water vapor(PWV) product with high spatial and temporal resolutions. The current FY-2/VISSR PWV product in operation is produced by using a traditional two-band physical split-window(PSW) method, which produces low quality results under dry atmospheric conditions. Based on the sensitivity characteristics of FY-2 F/VISSR water vapor channel and two split-window channels to atmospheric water vapor, this study developed a new, robust operational PWV retrieval algorithm for FY-2 F to improve the operational precision of the current PWV product. The algorithm uses a modified three-band PSW method, which adds a scale for the water vapor channel in the improved three-band PSW method. Integrated PWV products from the radiosonde data in 2016 are used here to validate the precision of the PWV retrieved by the modified three-band and traditional two-band PSW methods. The mean bias, root mean square error(RMSE), and correlation coefficient of the PWV retrieved by the modified three-band PSW method are 0.28 mm, 4.53 mm, and 0.969, respectively. The accuracy is much better than the PWV retrieved by the two-band method, whose mean bias, RMSE, and correlation coefficient are 12.67 mm, 29.35 mm, and 0.23. Especially, in mid-or high-latitude regions, the RMSE of the PWV is improved from 10 to 2 mm by changing the inversion in the two-band method to the modified three-band PSW method. Furthermore, the modified three-band PSW results show a better consistency with the radiosonde PWV at any zonal belt and season than the two-band PSW results. This new algorithm could significantly improve the quality of the current FY-2 F/VISSR PWV product, especially at sites where the actual PWV are lower than 15 mm.