A Monte Carlo code based on the geometric optics approach which applies to spheroids and polyhedral particles composed of one or more materials has been developed. The phase functions of hexagonal solid columns have been integrated with a log-normal particle size distribution and used as an input to the radiation transfer program based on the discrete ordinates method ('rstar' program of Center for Climate System Research, University of Tokyo) to calculate expected measured radiances from space. It was found that a difference of calculated reflected radiances with a case of spherical particles is large. This result indicates that one can not apply the Mie theory, which is valid for retrievals only in the case of spherical particles, to calculate radiative properties of cirrus clouds.

In this paper we show the importance of infrared interferometer data to the understanding of cirrus crystal absorption and retrieval of cirrus microphysical and bulk properties. The sensitivity of interferometer data to cirrus size, shape, and absorption process is shown through simulation using two crystal shapes namely hexagonal columns and planar polycrystals. The ice water path (IWP) and mean crystal dimension is varied between 15 - 44 gm^-2 and 20 - 80 micrometers respectively. The atmospheric radiative transfer modeling is done through line by line calculations in which an absorbing cirrus layer is placed and the optical depth predictions for the two habits and sizes are modelled according to anomalous diffraction theory (ADT) and Mie theory. To show potential retrievals we utilize actual cirrus interferometer data obtained from the Airborne Research Interferometer Evaluation (ARIES) instrument between the wavelengths of 6 micrometer to 15 micrometer. On the 10th June 1997 the Meteorological Office C-130 aircraft underflow a piece of semi-transparent cirrus which occurred off the east coast of England. Using the line by line model and assuming ADT as the absorption theory an IWP of 20 gm^-2 and absorption optical depth of 0.5 are retrieved which gives an r_e of about 40 micrometers. The retrieved r_e is representative of mid-latitude cirrus.

The scattering phase function angular characteristics, including the effects of particle nonsphericity on the phase function and the uncertainties using the Henyey-Greenstein approximation for hexagonal ice crystals are investigated systematically. Some new and significant facts are obtained.

A new method of particle sizing for mixed-phase and ice clouds proposed and tested by numerical simulation in authors' papers is applied here to experimentally measured scattering phase function. The method enables us to identify each component of a bi-component cloud composed of ice crystals and water droplets and to retrieve separately a size distribution for each cloud component. We use mainly available referenced data to test the inversion method with respect to the retrieval of size composition of mixed-phase and ice cloud under both single- and bi-component assumption and try to explain the known fact of the discrepancy between measured scattering phase functions for an ice cloud and those theoretically predicted by the ray tracing treatment, for instance, for convex ice crystals. Applying the inversion method enables us to show that one of effective ways to describe the scattering phase function behavior of mixed-phase and ice clouds is the bi-component assumption. It is rather natural for a mixed-phase cloud because of the existence of water droplets and ice crystals in the cloud simultaneously. On the other hand, one of an important physical reason for the bi-component assumption in an ice cloud lies in the well known fact that, as the cloud is transformed from water phase to ice one a high proportion of the particles can frequently stay as small supercooled water droplets even at very low temperature.

A near-global (50 degree S to 50 degree N) survey of ice crystal sizes for both continental and oceanic areas is presented, along with a description of the methodology for retrieving ice cloud particle size from ISCCP data. To retrieve ice particle sizes, a radiative transfer model is used which includes all major absorbing gases and cloud scattering/absorption to compute synthetic radiances as a function of satellite viewing geometry. The surface reflectance effect is considered according to retrieved monthly average reflectance values at each location. Ice crystal shapes are assumed to be hexagonal columns. The model results have been validated against clear sky observations and are consistent with the observed radiance range under cloudy conditions. We investigated the relationship between cloud temperatures/optical thickness and cirrus ice crystal sizes based on the retrieved cirrus particle size data. We found that for thick clouds ((tau) greater than or equal to 10), most of the regions over the globe show positive relationships between cloud temperature and cirrus ice crystal sizes. However, if relatively thin clouds are included ((tau) greater than or equal to 3), this correlation becomes negative for tropical areas but remains mostly positive for midlatitudes.

For the present investigation a new method for the calculation of the single scattering and absorption properties of hexagonal ice crystals called the Discretized Mie Formalism (DMF) is used together with a radiative transfer model in order to determine the influence of particle size and shape on the brightness temperature difference (BTD) between the two NOAA AVHRR infrared window channels 4 and 5. Comparing the results to those of optically equivalent spheres it is shown that particle shape has an influence on the observed BTD which is decreasing with increasing particle size.

Using a limited data set of spatially and temporally coincident DMSP SSM/T-2 data and NOAA AVHRR imagery, a methodology has been demonstrated to more accurately retrieve water vapor profiles with a non-linear, physical relaxation algorithm that is constrained by cloud top parameters derived from analysis of visible/infrared imagery. In addition, a unique multisensor, data fusion approach was developed for the specification of cloud top phase in the AVHRR imagery which must first be determined, due to the optically properties of thin cirrus clouds, to ensure the accurate specification of other cloud top parameters, including temperature, pressure, and height. Additional research is underway to determine if 1.6 micrometer data will significantly improve the capability to specify cloud top phase in daytime imagery using this data fusion technique. It is postulated that the optimal detection of thin cirrus and specification of cloud top phase requires the use of both 3.7 micrometer and 1.6 micrometer imagery. However, since both are not scheduled for simultaneous transmission in the NOAA-K data stream, different implementation strategies are recommended for use with the transmission of the 3.7 micrometer channel, the 1.6 micrometer data, and both should they become available in the future. The strategy for use with the 3.7 micrometer channel is the topic of this paper.

We have been investigating the potential of high spectral resolution infrared measurements (600 - 2500 cm^-1) such as those made by the HIS aircraft instrument and anticipated from the Infrared Atmospheric Sounding Interferometer instrument (IASI), to be launched on the METOP- 1 satellite by the end of 2001, for the retrieval of cirrus cloud properties. Simulations investigating the sensitivity of optical properties to particle size and shape have been carried out for a number of different particle distributions. Ray tracing for randomly oriented hexagonal columns, Mie theory for spheres and the T-matrix formulation for spheroids of different aspect ratios, have been used. The effect of particle size, shape, and cloud ice water path (IWP) on the exitant radiance has been investigated. Comparisons of the exitant radiance for clouds of equal IWP and differing microphysics are shown. From these studies some spectral regions showing sensitivity to cloud microphysics are identified and spectral difference plots are used to display the different behavior exhibited by different particle distributions. As a comparison, some examples of aircraft data obtained with the HIS instrument are also shown. It is concluded that particle size and shape have distinct spectral signatures and there appears to be some potential for using particular spectral variations to retrieve radiatively important microphysical and physical cloud parameters.

During the aircraft campaign EUCREX 94 different missions with the multispectral sensor OVID were flown inside frontal ice cloud systems. This study present estimated effective radii and cloud optical depths from measurements around 1.05 and 1.6 micrometer under the assumption of different particle shapes. The best agreement with independent measurements of other instruments result from the assumption of an irregular polycrystal. The measured effective radii vary between 18 and 46 micrometer which is compatible with published particle size distributions of moderate ambient temperatures between minus 45 and minus 55 degrees Celsius. An additional consideration of spatial features allow the distinction of cloud layers in different altitudes in the atmosphere and perhaps the estimation of cloud parameters from individual layers. This study show an example of such a recognition and discuss the potential for an operational algorithm.

In preparation for new meteorological infrared satellite sounders (such as the Infrared Atmospheric Sounding Interferometer, IASI) an interferometer has been mounted on the Meteorological Office's research aircraft. Early development of assimilation software will allow the new satellite data to be used at numerical weather prediction centers soon after launch. ARIES, the Airborne Research Interferometer Evaluation System, with a 1 cm^-1 wavenumber resolution over 600 to 3000 cm^-1 wavenumbers (wavelength 16(DOT)7 to 3(DOT)3 micrometer) is described. Preliminary data are compared with results from a line by line radiative transfer model using the atmospheric profile measured by the aircraft, (which can also measure cloud in-situ).

TAFTS is an airborne far-infra-red Fourier Transform Spectrometer (FTS) currently under construction. It is designed to make spectroscopic measurements of high radiometric accuracy in the upper troposphere/lower stratosphere in the band 12 micrometer to 120 micrometer (800 cm^-1 to 80 cm^-1). Its scientific mission is the direct observation of the radiative properties of upper troposphere humidity (UTH) and cirrus clouds, both of which have been shown by modeling studies to have great significance in the global radiation budget. Details of the instrument's design are presented.

It is shown that it is possible to retrieve aerosol properties using polarization measurements from satellite, or aircraft even when the surface polarization is significant and unknown. This extends the domain for which it is possible to intercompare ground and aircraft/satellite estimates of aerosol properties and allows the retrieval of aerosol properties to be made above bare soil surfaces.

The development of microwave heterodyne receiver technologies like Superconductor-Insulator-Superconductor-(SIS) mixer and Hot-Electron-Bolometer (HEB) offers the possibility to detect a large number of minor atmospheric constituents, which have been under the detection limit before. The requirements on the functional performance of spectrometers have increased at the same time. The Chirp Transform Spectrometer (CTS) fulfills these requirements, because of its high resolution, stability, linearity and dynamic range and has a large development potential for future spaceborne applications due to its very low mass and power consumption. The development status of CTS at the Max-Planck-Institut fur Aeronomie (MPAE) will be described and possible spaceborne applications will be discussed.

The Earth Radiation Budget (ERB), the balance between the incoming solar radiation from the sun and the outgoing reflected and scattered solar radiation and the thermal infrared emission from the Earth, provides information on the fundamental energy source of the climate system. To fulfil global coverage and sampling requirements, the ERB measurements have to be made from space. Broad-band measurements are necessary because all spectral regions in both the solar and infrared contribute to the radiative fluxes. Satellite data are used in a wide range of basic studies of the radiative forcing of the climate, such as understanding the effects of variations in trace gases, clouds and the surface. They also provide essential validation for climate models. All such measurements to date have been made from satellites in low earth orbit (LEO). There are strong diurnal variations in the radiation budget, particularly over land, in response to the diurnal variation of solar heating. Four LEO satellites could provide coverage of the diurnal cycle with a temporal resolution of 3 hours. At least hourly measurements are needed to resolve the diurnal cycle of tropical convection properly, and no practicable system of polar orbiting or other LEO satellites can deliver this. From the above, it appears that the only viable solution to the problem of diurnal sampling of the Earth's radiation budget is the inclusion of suitable sensors on the geostationary satellites which would allow for an essentially perfect temporal sampling. Disadvantages include the fact that geostationary satellites are much further from the Earth than polar orbiters, which affects the instrumental design, and each one can only provide a limited coverage of the globe. The Geostationary Earth Radiation Budget instrument (GERB) is a highly accurate visible-infrared radiometer designed to make unique measurements of the outgoing shortwave and longwave components of the Earth's Radiation Budget (ERB) from geostationary orbit. Such measurements have not been achieved previously, and are extremely important, because they will permit a rigorous test of our understanding of the diurnal variations in the ERB: this will enable improved operational weather monitoring and permit further important developments in climate change research. GERB will be launched on the (MSG) geostationary satellite in the year 2000. Both short-wave (0.32 - 4 micrometer) and total (0.32 - 30 micrometer) radiance measurements would be made, with longwave (4 - 30 micrometer) data obtained by subtraction. The accuracy requirements (1% short-wave and 0.5% longwave) are consistent with previous radiation budget measurements. The availability of GERB on MSG will also allow a more accurate calibration of the principal Meteosat Second Generation (MSG) operational sounding instrument, SEVIRI (Spinning, Enhanced Visible and InfraRed Imager).

Observations of low level stratiform clouds made over the Bay of Fundy from 15 August to 26 September 1995, as part of the Radiation, Aerosol, and Cloud Experiment (RACE) are used in this study. Aircraft, LAND Resources SATellite System (LANDSAT), and the Center for Atmospheric Research Experiment (CARE) lidar observations were used to obtain effective radius (rett), droplet number concentration (Nd). and cloud optical thickness (t). Radiation observations with a 28.5 m resolution at six solar reflectance channels from 0.45 μm up to 2.35 μm of the Thematic Mapper (TM) on LANDSAT were used. The 10.4-12.5 μm infrared channel has a field of view of 114 m. The visible extinction coefficient ( crex1) obtained from an aircraft mounted extinction meter was related to both liquid water content (L WC) and Nd. Optical thickness (t) were obtained from LANDSAT observations. Then, Nd, LWC, and reffalong lines of longitude are averaged at about 10 km intervals. The results showed that the relationships obtained between reff and t for stratiform clouds can be used to estimate reff values from satellite derived 't. Key words: Optical thickness, effective radius, extinction coefficient, aircraft and LANDSAT observations

The radioactive gas and aerosol ejections during and after accident at the nuclear power-station are very dangerous source of environmental pollution. These ejections are spread by meteorological air flows above considerable distances (up to several thousands km and more). This radioactive cloud is gradually blown away, changing its form and density. Because of the air heating over the accidental place the radioactive cloud can rise up to 14 km (tropopause). Therefore the dissipation and settling of radioactive nuclides are occurred over very large area. In this paper the optical method of registration of radioactive air ejection by means of particular bands of atmospheric fluorescence with very high threshold of excitation due to Auger effect is presented. There are three emission bands which are quite prominent and lie in the blue range of spectrum (400-480nm). It is important that they are absent in other events of natural and technological air emissions, besides short time lightning. The intensity of these emissions is much higher than background, especially at night. The transformation ratio from y-radiation flux to the visual one (for three bands) is 10-1 - 10-2 and for large accident like in Chernobyl one the optical emission (for radioactive cloud of 106 Ci) could be approximately 1014 - 1015 photons·c-1. For remote sensing of the radioactive air ejection the filter radiometer should be recommended the filters corresponding to the wavelengths of the particular emission bands. In existing aerospace remote sensing experiments these filters still haven't been used. Keywords: nuclear power-station, accident, radioactive cloud, optical fluorescence, Auger effect

The Crau-Camargue airborne experiment took place between the 10th and the 14th of July 1995 on the south coast of France. The purpose of the experiment was to measure microwave for a number of arid and verdant land types as found in the region. The UK meteorological C-130 aircraft participated, taking measurements with its two microwave radiometers, MARSS and Deimos, and with a multi-channel infra-red (Ir) radiometer, SAFIRE. Results are presented on the microwave emissivities spectra of verdant and arid land types.

Retrieval of atmospheric profiles of temperature and chemical components is generally accomplished by inverting the radiance function. Since the problem is non linear with respect to all variables, then a suitable linearization is resorted in order to solve the inverse problem. In this paper we introduce a new quasi-analytical method for estimating the coefficients of the linear problem that relies on a radiative transfer equation code for computing optical depths.

A new monochromatic radiative transfer algorithm based on compressed lookup tables of pre-computed atmospheric molecular absorption coefficients has been developed. These compressed look-up tables are called the kCompressed Database. Our motivation is to compute monochromatic absorption coefficients for any realistic Earth atmospheric situation (pressure, temperature, gas amount) at the same accuracy as a line-by- line code, but faster. In addition, the procedure for producing atmospheric transmittances is extremely simple, and easy to code. Although the kCompressed Database was originally developed to compute layer-to-space transmittances that are needed to produce fast transmittance models for high spectral resolution infrared temperature and humidity sounders, we have now developed a complete (non-scattering) atmospheric radiative transfer code around the kCompressed Database, called kCARTA (for kCompressed Atmospheric Radiative Transfer Algorithm). In addition, Jacobians with respect to gas amount and temperature can be rapidly performed, providing the user insight to the regions to which the measured radiance is most sensitive.

We present the results of radiative transfer calculations in the microwave region focussing on the effects of multiple scattering by nonspherical hydrometeors. The use of microwave frequencies, for example 37 GHz, leads for raindrops, typical with 4 mm diameter, to a size parameter of 1.5 and requires an exact scattering solution rather than the Rayleigh approximation. The distortion of raindrop shapes from the spherical geometry then becomes significant and has to be taken into account when the scattering properties are calculated. The model developed uses the full Stokes vector to include the effects of cross-polarization terms between all four Stokes components which are important when the scattering of nonspherical particles is considered. In this case the third and fourth Stokes component do not decouple from the first and second components. The one-dimensional microwave radiative transfer model is based on the successive order of scattering method and assumes azimuthal symmetry. The shape of the hydrometeors is approximated by rotational symmetric ellipsoids with a size dependent aspect ratio. These particles have a fixed orientation with their rotational axis aligned along the vertical. Results for nonspherical and spherical scattering will be presented. The differences between both methods, showing up to 15 Kelvin change in the polarization difference, depending on rain rate, frequency and viewing angle, will be discussed.

Inversion of the radiative transfer equation to retrieve the vertical profile of temperature from high resolution radiance spectra is an important problem in remote sensing of atmosphere. Because of its non linearity and ill conditioning, regularization techniques have been resorted in order to reduce the error of the retrieval. In this paper Generalized Singular Value Decomposition (GSVD) and Truncated Generalized Singular Value Decomposition (TGSVD) have been used to solve the linear model; the optimal regularization parameter for the proper amount of smoothing have been chosen by the L-curve criterion. A significant test problem has been worked out with reference to the Infrared Atmospheric Sounding Interferometer (IASI). The effectiveness of the methods to reduce variance and bias in the output profile has been addressed. We show that GSVD plus L-curve criterion or TGSVD plus L-curve are really effective in reducing error, variance and bias of the retrieved profile.

The POLarization and directionality of the Earth's Reflectances (POLDER) instrument onboard the Japanese ADEOS satellite offers unique possibilities for the retrieval of aerosol parameters with its polarization and multi-angular capability. In this study we examine a technique that simultaneously retrieves multiple aerosol parameters, namely asymmetry factor, single-scattering albedo, surface albedo, and optical thickness, using simulated multiangular POLDER reflectances. It is found that, for a typical illumination and observation geometry, these parameters can be retrieved rather accurately in the absence of noise. The retrieval accuracy deteriorates considerably, but remains tolerable, when a 0.1% white noise is present. The retrieval of these parameters in a more realistic atmosphere, i.e. with Rayleigh scattering, achieves similar accuracies, provided the assumption of the aerosol profile is valid. Indeed, this technique is sensitive to the aerosol profile assumed in the radiative transfer model.

Polarized reflectance from POLDER measurements are used to infer cloud properties such as cloud phase (water, ice), cloud droplet size, and bidirectional reflectances. The results show that data from simulator of POLDER instrument carried on aircraft can be used to infer cloud phase and droplet sizes. The POLDER data from satellite can be used to detect cloud phase but difficult to infer droplet sizes due to less angular observations around rainbow feature. Initial results of satellite POLDER data shows the potential ability of cloud classification using polarization information combined with reflected intensity. Combining with reflected intensity, the multiangle observations of polarized reflectance can be used to discriminate scene types such as ice cloud, water cloud, thin cirrus over water cloud, aerosol over water cloud and sun glint.

The small amount of water vapor in the upper troposphere plays an important role in the Earth's climate system, e.g. by determining the occurrence and distribution of cirrus clouds. However, only little is known about the vertical distribution of upper tropospheric humidity (UTH). In this paper we propose a method for the retrieval of UTH from measurements taken by the Millimeter wave Atmospheric Sounder (MAS) onboard the space shuttle. MAS is a microwave limb sounder that was originally built to measure trace gases in the stratosphere and mesosphere. One of its channels around 204 Hz is sensitive to the amount of water vapor near the tangent point down to altitudes of about 7 km. Retrievals performed with simulated data have shown promising results. First comparisons with measured data show large discrepancies between actual measurements and the predictions of the widely used Liebe model. Preliminary results suggest that the Liebe model strongly underestimates the dry air continuum in this frequency range for the limb-sounding geometry.

Classical trajectory calculations have been performed to determine the rotational distribution of vibrationally excited nitric oxide from collisions with atomic oxygen. The reaction occurs on two electronic potential energy surfaces which must be considered for a realistic description of the O+NO collision dynamics. The results, which have been statistically averaged over both electronic potential energy surfaces, are in good agreement with the available experimental data for vibrational relaxation of NO(v less than or equal to 9), as well as the temperature dependence of NO(v equals 1). The state-to-state relaxation rate coefficients involve the formation of long-lived collision complexes and indicate statistical behavior in O+NO collisions. The present study confirms earlier analysis that the NO(v equals 1) rotational distributions can indeed by described by a Maxwell-Boltzmann distribution, albeit with a rotational temperature of approximately 75% of the initial translational temperature. Thus, it appears possible to establish a lower bound to, and an estimate of, the nighttime quiescent terrestrial thermosphere by measuring the rotational envelope of the 5.3 micrometer emission from NO.

The Stratospheric Aerosol and Gas Experiment (SAGE) III is planned to be launched in summer 1998. It will provide aerosol extinction coefficients from solar occultation measurements at more wavelengths than SAGE II. This paper describes two methods used for the retrieval of some aerosol characteristics of great interest for the climate modeling and for the study of heterogeneous chemistry: effective radius, effective variance, surface area density and volume density. The first technique consists in a King inversion scheme and the second relies on a least squares fit on the extinction measurements. The two methods are applied to simulated extinction measurements at SAGE III wavelengths to investigate the ability of retrieving the aerosol characteristics in case of unimodal or bimodal log-normal size distributions. The contribution of channel 1.550 micrometer is estimated and uncertainties are also determined. The results derived from the two techniques are consistent in unimodal and bimodal cases for the four quantities: the least squares fit method is much faster but leads to larger uncertainties, in bimodal case the King method allows a better retrieval.

For microwave remote sensing of atmospheric composition or temperature an inversion of the measured primary data is in general necessary to obtain the desired profiles. In addition to the atmospheric quantities some unknown instrumental parameters might also need to be estimated from the data. Especially when designing a new instrument the question arises, whether the suggested data set contains enough information to retrieve the profiles and parameters with the desired accuracy. The singular value decomposition algorithm is shown to be a universal and powerful tool to analyze any linear or moderately nonlinear forward model and quantify the amount of retrievable parameters. In addition the method can be used as a simple and robust inversion technique, thus giving in one step not only an analysis of the relationship between measurement and parameter space, but also a solution of the inverse problem. The application of this method is illustrated using data obtained by ground-based measurements of ozone at 142 GHz.

In this contribution, the procedure of deriving a.n approximate solution for scattering on a finite cylinder from that of an infinite one by application of Huygens' principle has been extended to the case of non-circular cross-sections. To the best of our knowledge, this has been performed for the first time. Within our approximation, we are able to calculate total cross-sections and the albedo consistently, which are important for practical applications. The basic problem of the infinite cylinder we treated with the Discretized Mie Formalism (DMF). We proved that this technique can be successfully applied up to size-parameters of at least 100. Results for circular cylinders by our approximation and by the DMF are presented. Keywords: Hexagonal Ice Crystals, Cirrus, Nonspherical Hydrometeors, Discretized Mie Formalism (DMF), Huygens' principle

An important source of uncertainty in radiative forcing of climate models is their treatment of tropospheric aerosols. Satellite radiometry offers the only means of providing aerosol characterization on a global scale. POLDER (POLarization and Directionality of the Earth's Reflectances) instrument was launched in August 1996 on the ADEOS platform. Aerosol products are derived from combination of multiangular, multispectral and polarized observations in visible and near infrared bands. Validation effort has been planned to test POLDER aerosol products by comparison to independant measurements. The validation is achieved using ground based sunphotometers from a worldwide station network and by performing or participating to specific campaign such as ACE-2 (Aerosol Characterization Experiment) for marine and desert dust aerosol, AMT-4 (Atlantic Meridian Transect) with a ship crossing from Falklands to England (marine aerosol) and to continental campaigns in France (urban and continental aerosol). In the paper, two aspects of the validation will be presented. The first point concerns the analysis of data acquired during the specific campaigns. In that case, sunphotometer and polarimeter measurements will be qualitatively analysed but also used into comparison with POWER model database. In a second step, we shall report systematic studies of the worldwide sunphotometer network measurements (statistic on aerosol loading, Angstrom coefficient related to particle size). First results of these statisical studies will be presented and will be taken into account in the aerosols model database used into the POLDER aerosol algorithm. Keywords: POLDER, aerosols, polarization, validation.

Consideration of the radiative transfer through the atmosphere is essential for the quantitative analysis of the satellite- sensed data. To solve the radiative transfer equation directly, a detailed information about atmospheric radiation- related parameters, such as atmospheric soundings, atmospheric aerosol characteristics are required. This study presents a simple method to extract atmospheric aerosol radiative characteristics for LOWTRAN7 simulation of short wave radiation. Atmospheric aerosol radiative characteristics are obtained mainly from observed column optical depth with the statistical data of Spinhirne's atmospheric model of aerosol scattering coefficient and d'Almeida's atmospheric aerosol radiative characteristics. Also, to examine the application of this method, the results simulated by LOWTRAN7 with atmospheric aerosol radiative data are compared with CAGEX (CERES/ARM/GEWEX Experiment) data as ground-truth radiative measurements. Standard errors of LOWTRAN7's results with respect to CAGEX data were 1.9% (downward direct flux at the surface), 6.9% (downward diffuse flux at the surface). It is shown in the results that a large part of error in LOWTRAN7 flux simulation is found in the diffuse component due to scattering mainly by atmospheric aerosol. In a conclusion, better information about the radiative characteristics of atmospheric aerosols is required for improving the accuracy of radiative transfer simulation by model.

The formulae for cross-sections and extinction efficiency factors are derived for the cases when a polarized optical radiation interacts with a hexagonal plate and a hexagonal column. A hexagonal crystal has a pair of parallel faces. For such a crystal, there always are the refraction beams, leaving the crystal in the direction of incident wave propagation. As a rule, cross-sections of these beams are commensurable with the geometrical shadow of a crystal. Hence, the diffraction and scattered field are comparable to each other. Each of these fields is determined by means of one and the same formalism. Therefore, the diffraction and scattered fields have the identical structure, that permits to sum them taking into account a mutual phase shift. Using the methods of physical optics as a formalism makes it possible to obtain the analytical expressions for characteristics of extinction for a hexagonal plate and for a hexagonal column as well.

The lack of a global cirrus cloud database is a major deficiency in validating Global Circulation Models which are the basis for estimating long term climate change. Cirrus clouds being composed of ice particles act to reflect the incoming solar radiation and to block infrared energy radiation loss from the earth. The net effect can be either warming or cooling. The essential data are global distributions of the Ice Water Path (IWP) and effective particle size. We have modeled the outgoing earth radiance spectrum between 8 and 1000 micrometer as a function of IWP and effective particle size. The results are used to estimate cirrus retrieval accuracies for cirrus from far infrared/submm measurements by Fourier transform spectroscopy. We describe the aircraft-based Far Infrared Sensor for Cirrus (FIRSC) instrument which is currently under construction. We also discuss the potential contribution of far infrared/submm measurements for the validation of cirrus products anticipated in the MTPE MODIS program.

A precise scene identification is important for Earth Radiation Budget Experiments to study in more detail cloud radiative effects and also to develop specific angular dependence models for the conversion of measured radiances into radiative fluxes. For this purpose, the ScaRaB radiometer includes two narrow-band channels in addition to the broad- band channels which retrieve earth emitted LW and solar reflected SW radiation. By applying the ISCCP (International Satellite Cloud Climatology Project) algorithm on the ScaRaB data, clear sky LW fluxes have been improved compared to ERBE (Earth Radiation Budget Experiment), especially in the tropics where atmospheric water vapor is high.

Data from the ScaRaB radiometer flown on board the Meteor-3/7 satellite were first employed for validating and correcting a TOA Earth radiation budget product generated from GOES-7 and the latter was then combined with ground radiation measurements for addressing the effect of clouds on atmospheric absorption of solar radiation. By virtue of comparison between coincident and collocated radiative quantities derived from ScaRaB and GOES sensors, it was found that GOES calibrations for both visible and infrared window channels appear to be adequate, but narrow to broad-band conversion of short-wave measurements suffers systematic errors. After correcting this problem, the cloud radiative forcing at the top of the atmosphere (TOA) and at the surface were derived from space- and ground-based measurements made during the US Atmospheric Radiation Measurement (ARM). The ratio of the two forcing terms is in excellent agreement with that determined by radiative transfer models, in contradiction to the recent claim of cloud absorption anomaly.

Near coincident data from satellite and aircraft overpasses of stratocumulus in the South Atlantic have been analyzed. The satellite data are from the Along Track Scanning Radiometer-2 (ATSR-2) onboard the European Remote Sensing satellite-2 (ERS- 2) and are measurements of reflectance at wavelengths of 0.87 micrometer and 1.6 micrometer. From these, retrieval algorithms estimate the cloud optical depth and effective drop size using well established principles. The aircraft data consist of reflectance measurements made using the Scanning Airborne Filter Radiometer (SAFIRE) from a transit made just above the cloud, and liquid water content (LWC) and effective drop size measurement made using an FSSP probe from a transit made approximately 100 m below cloud top. The aircraft flights were made as part of the South Atlantic Tropical Experiment-2 which took place off the Namibian coast in October 1995 and which was designed to coincide with ERS-2 overpasses where possible. One particular flight, A423, has been studied here as offering the best chance of intercomparison of the respective systems measurements. Two flight legs have been analyzed, one for validation of reflectance measurements and one for validation of cloud effective radius. Poor correlation in initial comparisons based on the geolocation of the two instruments was improved greatly by allowing for, and estimating, the advection of the cloud deck by the local wind. Other adjustments included compensating a small error in the ATSR geolocation and allowing for differences in the respective instrument view angles. Following these adjustments, good agreement is shown for the 11 micrometer brightness temperatures and for the 0.87 micrometer reflectances. Large biases in the 1.6 micrometer reflectances confirm calibration errors that were already suspected for both instruments. Using the same wind advection the cloud effective radius retrievals were compared for the previous flight leg. Agreement is shown to be within 0.5 micrometer for measurements within 5 minutes of the exact collocation time. This is a remarkable result considering the sensitivity of the ATSR retrievals to 1.6 micrometer calibration errors.

The problem of producing polar cloud masks for satellite imagery is an important facet of the research on global warming. For the past three years, our research on this topic has produced a series of classifiers. The first classifier used traditional statistical techniques, and, although the performance was reasonably good, better accuracy and faster classification speeds were desired. Neural network classifiers provided an improvement in both classification speed and accuracy but a single monolithic network proved difficult to train and was computationally expensive. A decomposition of the neural network into a hierarchical structure provided significant reductions in training time and some increase in accuracy. While this technique produced excellent results, to optimize its performance a minimal feature set and a highly accurate and easily computed switching mechanism must be identified. This paper presents recent developments in these two areas. Landsat Thematic Mapper (TM) data from the arctic and antarctic was used to test the network. A minimal feature set, which defines the elements of the network input vector, is desirable for both improving accuracy and reducing computation. A smaller input vector will reduce the number of weights which must be updated during training and concomitantly reduce training and testing times. Small input vectors are also desirable because of the oft-cited 'curse of dimensionality' which states the higher the dimension of the problem to be solved, the more difficult it will be for the network to find an acceptable solution. However, it is also known that if a network has insufficient information, it will not be possible to form an appropriate decision surface. In that case, additional features, and additional dimensions, are required. Finding the proper balance can be difficult. Previously, trial and error was used to find a 'good' selection of features for classification. Features were added individually and those which had no impact on the neural network classification were deleted. A new approach clusters the features and selects the members of the cluster with the highest information content. Features are still removed one at a time but clustering ensures that statistically different features are preserved and allows for parallelization of the feature selection process. In addition, a fuzzy expert system provides a much faster classification accuracy approximation. This technique was able to significantly reduce the size of the feature vectors without sacrificing classification accuracy. The switching mechanism used in this work employs a series of static and adaptive thresholds derived from statistical analysis of polar scenes. This technique is faster than the corresponding neural network switching mechanism and can be easily changed as additional data becomes available. The resulting system, then, uses the adaptive thresholds to select the appropriate neural network from a collection of multilayer perceptron networks each responsible for classifying a subset of the total number of classes. The inputs to these networks are selected by the fuzzy logic algorithm. The difficulty of finding these thresholds, a task performed by a human expert, motivated the use of a genetic algorithm to determine these values. This system was able to achieve 96.45% accuracy on the fundamental problem of distinguishing cloud from non-cloud classes. The time required to classify 468,750 pixels in a satellite image was 50 seconds.3220

The present work empirically deals with the challenging problem of the integration of data obtained from passive and active microwave sources, in order to develop procedures to suitably calibrate and validate satellite-based passive microwave rainfall algorithms by means of multiparameter radar information over midlatitude areas. SSM/I passive microwave radiometer precipitation related parameters have been analysed against multiparameter radar Zh and Zdr three dimensional maps, obtained from the POLAR-55C multiparameter radar set near Florence, Italy. The main objectives of this work are: to try to better analyse the satellite beam-filling problem and its different channel penetration topic; to design and validate an operational procedure in order to integrate SSMfl and the POLAR -55C data. Starting from radar-derived volumetric information and passive microwave multi-frequency data, we have faced several statistical analyses of the obtained data sets. Results report the effectiveness of Montagnana radar and SSM/I data fusion. In particular, it was outlined the· diverse influence that different vertical profile layers have on radiometer channels. Radar measurements versus SSM/I parameters correlation values may be improved by filtering radar data according to several parameters thresholds, in order to tackle beam-filling problem and statistical issues. Eventually, utilised hydrometeor classification schemes seems not to work properly for a whole stack of CAPPis. Keywords: Microwave, Radar, Data Integration, Vertical structure, Statistical Analysis

Using multi-spectral Advanced Very High Resolution Radiometer (A VHRR) data, the spatial distribution of smoke aerosols from biomass burning and dust aerosols are detected. Fires from biomass burning are determined using the spectral information from the A VHRR. The regional radiative impact at the top of atmosphere (TOA) are estimated using coincident Earth Radiation Budget Experiment (ERBE) data and the downward shortwave irradiances are also computed in the presence of biomass aerosols and validated against field measurements. Using empirical relationships, carbon monoxide (CO) concentrations estimated from the A VHRR are compared with space shuttle measurements. Results indicate that biomass burning :frres produce large amounts of smoke aerosols that have a significant regional radiative impact. The instantaneous net radiative forcing is on the order of -24 to -40 W/m2 which indicates a cooling effect at the surface. During April 1994 over tropical Asia, carbon monoxide concentrations range from 160-200 ppbv over active biomass burning areas. Keywords : biomass burning, dust storms, aerosols, radiative forcing, carbon monoxide, A VHRR, MAPS.

The influence of aerosol, described by means of the aerosol optical thickness, on the atmospheric radiance - the sum of the Rayleigh and aerosol components is analysed. Assuming that the Rayleigh radiance is known, the contribution of the aerosol radiance component to the atmospheric one depends on the amplitude of aerosol optical thickness. The amplitude reflects the temporal variability of the aerosol optical thickness caused by appearing of water soluble particles or particles from continental air masses. It was found that the contribution of the aerosol radiance increases with wavelength increases from about 2% (at 400 nm) to about 12% (at 670 nm) and is the most sensitive to the amplitude values of aerosol optical thickness at wavelength 670 nm. Keywords: Baltic Sea, aerosol radiance, Rayleigh radiance, atmospheric radiance, aerosol optical thickness, total solar spectral irradiance, diffuse solar spectral irradiance, empirical orthogonal functions, atmospheric correction, marine aerosols

Using high resolution spectroscopy of the atmospheric oxygen A-band rotational absorption lines (760-780 run}, the probability distribution of optical pathlengths were derived for clear and cloudy skies. A large enhancement of the mean optical pathlength and a systematic change in the form of the probability distribution is found for the cloudy sky. Our observations are compatible with Gamma type distributions. The measurements are compared with a radiative transfer model in order to validate the treatment of the multiple Mie-scattering in clouds. Our measurements show a significant discrepancy compared to a model (DISORT) assuming plane parallel homogenous cloudiness. Larger cloudy sky optical paths may contribute to explain the intensely debated anomalous cloud absorption. These measurements can serve as direct input for atmospheric heating calculations. Keywords: multiple Mie scattering, heterogeneous clouds, oxygen A-band, Gamma distribution

The Global Ozone Monitoring Experiment (GOME), launched on board of ERS-2 in April 1995, is a spectrometer measuring the Earth's reflectivity between 240 and 790 run with a spectral resolution of 0.2-0.4 nm. The spatial resolution of GOME can be varied between 40 x 40 km2 and 40 x 320 km2 . The main geophysical product from GOME is the total ozone column, which is derived from the GOME reflectivity measurements using the Differential Optical Absorption Spectroscopy method. For accurate retrievals of the total ozone column, the presence of clouds has to be taken into account. Therefore, cloud cover fraction is derived from the GOME reflectivity measurements around the 0 2 A-band using the the Initial Cloud Fitting Algorithm (ICFA). The ICFA results are input to the GOME ozone column retrieval algorithm. In this study, the ICFA cloud fraction results are validated in two ways. Firstly, a statistical approach is followed_ by comparing monthly averaged ICFA results with monthly averaged cloud fractions from the International Satellite Cloud Climatology Project (ISCCP). Secondly, a detailed comparison is performed for a limited number of data using cloud fractions derived from Along Traclc Scanning Radiometer-2 (ATSR-2) data and synoptical observations (SYNOP). The main conclusion is that the ICFA cloud fractions generally show a reasonable correlation with ISCCP, ATSR-2 and SYNOP data. However, the difference between cloud fractions over land and sea which shows up in the ISCCP results is not present in the ICFA results. Furthermore, the ICFA cloud fractions are too high over highly reflecting surfaces (e.g. desert and snow). The magnitude of errors in the GOME total ozone column due to errors in cloud fraction and cloud top height has been investigated using a detailed radiative transfer model. From this_ analysis, it is concluded that errors in these cloud parameters may introduce errors in the ozone vertical column of typically 2-33. Keywords: GOME, clouds, ozone, cloud detection

In this paper a robust processing chain for meteorological data handli11g is pmposed. The chain relies on a shape analysis technique, called modal matching, that is applied to the characterization and tracking of meteorological features from satellite images. Results show that a precise correspondence between point pairs in successit1e frames of a sequence can be achiet<ed, enabling to study the complex movements of meteorological structures in greater detail. Moreover, a technique to determine intermediate (not recorded) frames in a sequence is offered, useful for nowcasting and short-term. forecasting issues.