Statistical approaches to cloud classification

For the determination of clouds from satellite data there exist in general more unknown parameters than independent observations. If the bispectral observations are used from the geostationary satellites in the solar (VIS-channel) and in the infrared (IR-channel) range to derive cloud parameters, information is needed whether a pixel radiance is from a cloud free or a cloudy scene. Statistical methods are applied to derive those informations. Various proposed statistical methods are discussed.The histogram analysis developed at the University of Cologne is described in detail: bispectral (two-dimensional) histograms are partitioned into clusters. Cloud cover results are shown. A comparison is given between the results of the histogram analysis, threshold methods (VIS- and IR threshold separately and both combined) and the spatial coherence method developed by Coakly and Bretherton. The cloud cover varies in this example by a factor of two depending on the definition of the threshold between cloud free and cloudy pixels. It is further shown that after a cluster analysis of a two-dimensional histogram the derived cloud cover is not as sensitive to the threshold as for a threshold method. The methods which are discussed here are those proposed for the International Satellite Cloud Climatology Project (ISCCP). The results are from the pilot study of the ISCCP.     

Remote sensing of clouds

The extraction of information on cloud cover from present-day multispectral satellite images poses a challenge to the remote sensing specialist. When approached one pixel at a time, the derived cloud cover parameters are inherently nonunique. More information is needed than is available in the radiances from each channel of an isolated pixel. The required additional information can be obtained for each scene, however, by analyzing the distribution of pixels in the multi-dimensional space of channel radiances. The cluster patterns in this space yield statistical information that points to the most likely solution for that scene. The geostationary and polar orbiting meteorological satellites all have, at a minimum, a solar reflection channel in the visible spectrum and a thermal infrared channel in the 8–12 micron window. With the information from the cluster patterns and application of the equations of radiative transfer, the measurements in those channels will yield cloud cover fraction, optical thickness, and cloud-top temperature for an assumed microphysical model of the cloud layer. Additional channels, such as the 3.7 micron channel on the AVHRR of the polar orbiting meteorological satellites, will will yield information on the microphysical model—e.g., distinguishing small liquid liquid droplets (typical of low level clouds) from large ice particles (typical of cirrus and the tops of cumulonimbus). New channels to be included in future satellite missions will provide information on cloud height, independent of temperature, and on a particle size and thermodynamic phase, independently of each other. A proposed STS mission using lidar will pave the way for the use of active sensors that will provide more precise information on cloud height and probe the structure of thin cirrus and the top layer of of the thicker cloud.     

A method for estimating cross radiance

When imaging the surface from satellites or aircraft, “cross radiance” diminishes the information content of the pictures. In this paper a simple method is presented to estimate the value of cross radiance. This method includes a height-dependent aerosol size distribution model and the calculations refer to the single scattering approximation. The height variation of aerosol size distribution has significant effect on the value of cross radiance, while the areal distribution does not change much in comparison with that of the height-independent aerosol model.     

Clouds and Earth Radiant Energy System (CERES), a review: Past,present and future

The Clouds and Earth Radiant Energy System (CERES) project’s objectives are to measure the reflected solar radiance (shortwave) and Earth-emitted (longwave) radiances and from these measurements to compute the shortwave and longwave radiation fluxes at the top of the atmosphere (TOA) and the surface and radiation divergence within the atmosphere. The fluxes at TOA are to be retrieved to an accuracy of 2%. Improved bidirectional reflectance distribution functions (BRDFs) have been developed to compute the fluxes at TOA from the measured radiances with errors reduced from ERBE by a factor of two or more. Instruments aboard the Terra and Aqua spacecraft provide sampling at four local times. In order to further reduce temporal sampling errors, data are used from the geostationary meteorological satellites to account for changes of scenes between observations by the CERES radiometers.     

Satellite measurements of tropospheric aerosols

The ability to measure tropospheric aerosols over ocean surfaces has been demonstrated using several different satellite sensors. Landsat data originally showed that a linear relationship exists between the upwelling visible radiance and the aerosol optical thickness (about 90% of this thickness is generally in the lowest 3 km of the atmosphere). Similar relationships have also been found for sensors on GOES, NOAA-5 and NOAA-6 satellites. The linear relationship has been shown theoretically to vary with the aerosol properties, such as size distribution and refractive index, although the Landsat data obtained at San Diego showed little variability in the relationship. To investigate the general applicability of the technique to different locations, a global-scale ground-truth experiment was conducted in 1980 with the AVHRR sensor on NOAA-6 to determine the relationship at ten ocean sites around the globe. The data for four sites have been analyzed, and show excellent agreement between the aerosol content measured by the AVHRR and by sunphotometers at San Diego, Sable Island and San Juan, but at Barbados, the AVHRR appears to overestimate the aerosol content. The reason for the different relationship at the Barbados site has not been definitely established, but is most likely related to problems in interpreting the sunphotometer data rather than to a real overestimation by the AVHRR. A preliminary analysis of AVHRR Channel 1 (0.65 μm) and Channel 2 (0.85 μm) radiances suggest that useful information on the aerosol size distribution may also be obtained from satellite observations.     

Mixed surface reflection functions determined from a priori knowledge and satellite radiance

A method to derive mesoscale area means of surface solar flux densities from a priori knowledge and actual cloudfree satellite radiances is presented. It is based on the concept of the mixed reflection function which can be derived from existing data. Herewith and with actual atmospheric data derived from the operational meteorological network the cloudfree radiation field is computed. By comparison of computed and measured satellite radiance the surface albedo of the model is tuned. In a case study this method is applied to an agricultural region called La Mancha, Spain, and comprehensively checked against airborne radiance measurements. The surface albedo can be determined to about ± 0.01.     

A comparison of atmospheric aerosol measurements by various satellite sensors

Investigations to measure the vertical optical thickness of aerosols over ocean surfaces has been conducted using several different satellite sensors. Landsat 1 and Landsat 2 data originally confirmed that a linear relationship exists between the upwelling visible radiance and the aerosol optical thickness (about 90% of this thickness is generally in the lowest 3 km of the atmosphere). Similar relationships have also been found for sensors on GOES-1, SMS-2, NOAA-5, and NOAA-6 satellites. The linear relationship has been shown theoretically to vary with the aerosol properties, such as size distribution and refractive index, although the Landsat data obtained at San Diego showed little variability in the relationship. The differences between the results found for the various satellite sensors are discussed, and are attributed mainly to uncertainties in the calibration of the sensors. To investigate the general applicability of the technique to different locations, a global-scale ground truth experiment was conducted with the AVHRR sensor on NOAA-6 to determine the relationship at eleven ocean sites around the globe. Analysis of the data shows good agreement between the satellite and ground truth values of the aerosol optical thickness, and indicates that the technique has global application. At two of the sites, multispectral radiometric measurements of the Junge aerosol size distribution parameter were made, and showed good agreement with a value inferred from the AVHRR Channels 1 and 2 radiances.     

Nighttime cloud properties retrieval using MODIS and artificial neural networks

In this work a methodology for inferring water cloud macro and microphysical properties from nighttime MODIS imagery is developed. This method is based on the inversion of a theoretical radiative transfer model that simulates the radiances detected in each of the sensor infrared bands. To accomplish this inversion, an operational technique based on Artificial Neural Networks (ANNs) is proposed, whose main characteristic is the ability to retrieve cloud properties much faster than conventional methods. Furthermore, a detailed study of input data is performed to avoid different sources of errors that appear in several MODIS infrared channels. Finally, results of applying the proposed method are compared with in-situ measurements carried out during the DYCOMS-II field experiment.     

Comparison of cloud top heights measured by airborne Lidar and TIROS-N image data

The determination of the cloud top height by means of satelliteborne IR-radiometers requires the conversion of the measured radiance to an equivalent blackbody temperature and the assignment of this temperature to a geometrical height. The latter is associated with errors which add up easily to several kilometers. DFVLR did a case study to compare satellite derived cloud top heights with those from airborne Lidar measurements. The difference of the radiosonde temperature from the standard temperature profile results in a 1.8 km difference in cloud top height. The achievable accuracy using actual radiosonde temperatures is ± 0.4 km for optical thick clouds and much less for optical thin clouds.     

星下目标的光谱反射特性分析及云层的自动识别

本文着重讨论了遥感数据的实时分类和星载相机快门的自动关启问题.首先建立了计算各类目标景象亮度的物理模型,分析了大气等因素对遥感数据的影响及各类自然目标的光谱反射特性,计算了三十四种目标在九种大气条件下的亮度值.进而提出了识别植被、土壤、水、冰雪和云等五类目标的最佳波段和分类函数及分类流程图,得到了很高的分类精度.然后又特别讨论了云层和所有非云目标(包括混合目标)的分类问题,并用密度计测量了十张MSS卫星图片上的180个目标的亮度值.最后提出了云层开关的实现途径和系统设计的基本思想.      

The use of stereoscopic satellite observation in the determination of the emissivity of cirrus

The feasibility of determining cirrus “emissivity” from combined stereoscopic and infrared satellite observations in conjunction with radiosounding data is investigated for a particular case study. Simultaneous visible images obtained during SESAME-1979 from two geosynchronous GOES meteorological satellites were processed on the NASA/Goddard interactive system (AOIPS) and were used to determine the stereo cloud top height Z_C as described by Hasler [1]. Iso-contours of radiances were outlined on the corresponding infrared image. Total brightness temperature T_B and ground surface brightness temperature T_S were inferred from the radiances. The special SESAME network of radiosoundings was used to determine the cloud top temperature T_CLD at the level defined by Z_C. The “effective cirrus emissivity” NE where N is the fractional cirrus cloudiness and E is the emissivity in a GOES infrared picture element of about 10 km × 10 km is then computed from T_B, T_S and T_CLD.     

Seasonal variation of gravity wave sources from satellite observation

Recent high-resolution satellite observations of gravity waves in the middle atmosphere have shown correlations with the strength of the stratospheric jet stream, surface topography, and tropical convection. Seasonal variations of wave-induced stratospheric radiance variances are often the manifestations of modulations of these sources and refractive influences. In this paper, we focus on the seasonal climatology of gravity waves observed by the UARS MLS, while also showing some new results from GPS and AMSU instruments. Our analysis is aided by MWFM modeling of mountain waves at high latitudes and CMAP precipitation indices in the subtropics to provide a clearer picture of global gravity wave dynamics.     

The Geostationary Earth Radiation Budget Edition 1 data processing algorithms

The Geostationary Earth Radiation Budget (GERB) instrument is the first to measure the earth radiation budget from a geostationary orbit. This allows a full sampling of the diurnal cycle of radiation and clouds – which is important for climate studies, as well as detailed process studies, e.g. the lifecycle of clouds or particular aerosol events such as desert storms. GERB data is now for the first time released as Edition 1 data for public scientific use. In this paper we summarise the algorithms used for the Edition 1 GERB data processing and the main validation results. Based on the comparison with the independent CERES instrument, the Edition 1 GERB accuracy is 5% for the reflected solar radiances and 2% for the emitted thermal radiances.     

Satellite monitoring of surface temperatures in semi-desert: Differences between anthropogenically impacted terrain and protected natural vegetation

The northern Sinai is a sandy semi-desert. Severe overgrazing and other anthropogenic pressures contribute to an extremely sparse vegetative cover. A 6×6 km area was fenced off in the summer of 1974, constituting an exclosure from the grazing herds and from harvesting of plants for firewood. The vegetation in this exclosure recovered rapidly. In this study, radiances and surface temperatures of the vegetated exclosure and of the surrounding anthropogenically impacted terrain were monitored for the period March–September 1981, using NOAA-6 satellite. This satellite carries the Advanced Very High Resolution Radiometer (AVHRR) that measures visible and solar infrared radiances and also radiation temperatures at 11 μm band. In the digital images, the exclosure forms an easily recognized square, darker in the visible and solar infrared AVHRR channels than the surroundings. We concentrated on the corner in which there was no anthropogenic activity. Based on the ratio of the radiance over the exclosure to that over the surrounding terrain, the protrusions parameter s (vertical projection of the protrusions per unit area) has been estimated. The average value of s for the various satellite passes is 0.20 as measured in the visible channel and 0.18 as measured in the solar infrared. The radiation temperatures of the exclosure and of the surrounding terrain were analyzed. The radiation temperatures of the vegetated exclosure (sand with protruding bushes) are higher (by up to 2.5°K) than those of the surrounding terrain (that can be approximately regarded as bare sand). It is concluded that in an arid climate, under the semi-dormant conditions of vegetation (which prevail at all times except for the desert-bloom period, after a rain) the evapotranspiration is low, so that its effect on the surface temperatures is very small. Under these conditions, the surface temperatures are controlled by the surface albedo and the air flow at the surface.     

Most suitable conditions for aerosol monitoring from space

Numerical modelling is used to search for the most suitable conditions, with the object of determining the atmospheric turbidity (aerosol optical depth) from upward emerging spectral radiances in cloudfree pixels over water surfaces. The most suitable conditions are those where the influence of the turbidity on the radiance most strongly outweighs that of the other optically acting constituents. Since the actual values of these constituents of atmosphere and surface are usually known only within certain limits, using the most suitable conditions minimizes the uncertainty in the turbidity to be derived from satellite measurements. As a result, favorable zenith angles of the satellite and favorable wavelengths are presented for an atmosphere with maritime aerosols with and without Saharan dust. The results represent an advance on a paper published by the authors in 1981 [1].     

A modern robust approach to remotely estimate chlorophyll in coastal and inland zones

The chlorophyll concentration of a water body is an important proxy for representing the phytoplankton biomass. Its estimation from multi or hyper-spectral remote sensing data in natural waters is generally achieved by using (i) the waveband ratioing in two or more bands in the blue-green or (ii) by using a combination of the radiance peak position and magnitude in the red-near-infrared (NIR) spectrum. The blue-green ratio algorithms have been extensively used with satellite ocean color data to investigate chlorophyll distributions in open ocean and clear waters and the application of red-NIR algorithms is often restricted to turbid productive water bodies. These issues present the greatest obstacles to our ability to formulate a modern robust method suitable for quantitative assessments of the chlorophyll concentration in a diverse range of water types. The present study is focused to investigate the normalized water-leaving radiance spectra in the visible and NIR region and propose a robust algorithm (Generalized ABI, GABI algorithm) for chlorophyll concentration retrieval based on Algal Bloom index (ABI) which separates phytoplankton signals from other constituents in the water column. The GABI algorithm is validated using independent in-situ data from various regional to global waters and its performance is further evaluated by comparison with the blue-green waveband ratios and red-NIR algorithms. The results revealed that GABI yields significantly more accurate chlorophyll concentrations (with uncertainties less than 13.5%) and remains more stable in different waters types when compared with the blue-green waveband ratios and red-NIR algorithms. The performance of GABI is further demonstrated using HICO images from nearshore turbid productive waters and MERIS and MODIS-Aqua images from coastal and offshore waters of the Arabian Sea, Bay of Bengal and East China Sea.     

Estimation of cloud top height and effective cloud cover from infrared satellite soundings

The paper deals with a new method for simultaneous determination of cloud top height and effective cloud cover, using infrared radiance data of satellite-borne instruments. These cloud properties derived from the Selective Chopper Radiometer on the Nimbus 5 satellite are compared with nearly simultaneous observations by radiosondes and with satellite images. Encouraging results for Central-Europe during January, April, July, August and October 1974, as well as numerical simulations indicate that the method proposed here, would be useful also for global application. Another advantage of the described procedure are the small amount of computing time, and that no other data are required than 3 of infrared channel values, for each sounded spot.     

Validation of nimbus-7 temperature-humidity infrared radiometer estimates of cloud type and amount

Estimates of clear and low, middle and high cloud amount in fixed geographical regions approximately (160km)² are being made routinely from 11.5μm radiance measurements of the Nimbus-7 Temperature-Humidity Infrared Radiometer (THIR). The purpose of validation is to determine the accuracy of the THIR cloud estimates. Validation requires that a comparison be made between the THIR estimates of cloudiness and the “true” cloudiness. The validation results reported in this paper use human analysis of concurrent but independent satellite images with surface meteorological and radiosonde observations to approximate the “true” cloudiness. Regression and error analyses are used to estimate the systematic and random errors of THIR derived clear amount.     

Validation of satellite-derived atmospheric temperature and water vapor concentration using radiosonde and rocketsonde measurements

Measurements of the spectral radiance of the earth's atmosphere from satellites can be related to the vertical structures of temperature and humidity. Derived profiles of these quantities are compared with radiosonde and rocketsonde observations, as well as with horizontal and vertical cross-sections of the atmosphere. In some regions of the atmosphere, particularly where large gradients are found, significant differences occur. A method for overcoming these by use of Typical Shape Functions is discussed. Transmittances computed from theory require modifications which are not well defined, and radiances measured from some satellite instruments disagree with computed values in ways which suggest calibration or instrument problems.