An aircraft version of the Polder instrument flew over the Straits of Dover to observe coastal waters. Atmospheric corrections are a critical issue for ocean color. The measurements reported here clearly indicated that the usual atmospheric inversion schemes are not applicable. The encountered aerosol models did not correspond to the usual standard models. BRDF POLDER measurements indicated that a local and monodispersed aerosol mode was predominant. Ground-based measurements of the sky radiances and of the degrees of polarization confirmed the POLDER observations. POLDER polarized radiance analysis is expected to improve the aerosol model knowledge in order to correctly retrieve the water leaving radiance.

Bidirectional measurements of the downward radiance, degree and plane of polarization were performed with the high resolution spectrometers OVID and HiRES in the spectral range of 650 nm - 830 nm on two different days with meteorologically different conditions. The measurements have shown that the spectral and bidirectional behavior of the degree of polarization is different for these two days, especially in the 02A band. It can be assumed that this behavior is caused either by different aerosol types on these two days, or by different height distributions of the aerosol types, or by a combination of these two effects. To simulate the measured Stokes vector, first of all aerosol optical thickness and its height distribution are varied and different calculations are performed.

In this work we study the complete evolution of an episode of Saharan dust invasion over Canary Island. To describe adequately this phenomena two magnitudes are calculated from NOAA-AVHRR satellite data: aerosols optical depth (AOD) in the channel 1 and the ratio channel 1-channel 2 (R12) which gives us rough information about the aerosols size distribution. The aerosols optical depth, calculated to 500 nm with a ground based sunphotometer, situated above the inversion layer is used too.

This paper reports the main results concerning polarization by clouds derived from POLDER (polarization and directionality of earth's reflectances) airborne version. These results tend to confirm the high information content in the polarization (phase, altimetry). The preliminary results of EUCREX'94 (European Cloud Radiation Experiment) evidenced the drastically different polarized signatures for ice crystals and water droplets. Here we report systematic and statistically significative observations over the whole EUCREX data set. The results show that the cirrus exhibit their own signature. Preliminary observations performed during CLEOPATRA'91 (Cloud Experiment Ober Pfaffenhofen And Transport) and EUCREX'94 campaigns have shown the feasibility of cloud altimetry using spectral information (443 nm and 865 nm) of the polarized light over liquid water droplets clouds. Altimetry technique has been generalized on ASTEX-SOFIA'92 and EUCREX'94 data sets. All these results are presented and discussed in this paper.

The reflection and transmission properties of inhomogeneous clouds (or ensembles of clouds) are simulated, which is relevant for e.g., satellite remote sensing and climate modeling. One of the main cloud properties is optical thickness. Here cloud inhomogeneity is assumed to arise from cloud optical thickness variability over a certain area. A simple triangular distribution function of cloud optical thickness is used, based on observations of cloud liquid water path distribution. An accurate plane-parallel radiative transfer model is employed to study the effect of cloud optical thickness variability on the reflection and transmission properties of inhomogeneous clouds. The effect of cloud sides is neglected. Results of the reflectivity, albedo, and transmission of inhomogeneous clouds are shown. Inhomogeneous clouds have lower reflectivity and albedo, and higher transmission than homogeneous clouds with the same number of cloud particles. The ratio of average cloud optical thickness to effective cloud optical thickness, which is the optical thickness of a homogeneous cloud that has the same reflection or transmission properties, is about 1.3 for thin clouds (average optical thickness 5), and about 1.6 for thick clouds (average optical thickness 50).

Among the problems that have received increased attention in recent years are the changes in cloud radiative effects due to greenhouse warming and increase in anthropogenic aerosols. Before the end of the century, a series of Earth-orbiting satellites will carry several advanced, well-calibrated, instruments designed to provide global observations of the Earth-atmosphere system. Among them, the polarization and directionality of the Earth's reflectances instrument (POLDER) is a new instrument devoted to the observation of the polarization and directionality of reflected solar radiation. The instrument concept has been accepted on the Japanese Advanced Earth Observing Satellite (ADEOS) platform scheduled for launch in 1996. The aim of this paper is both to show how POLDER may contribute to improve our understanding of the role of clouds and radiation in climate and to highlight the original contributions of this new instrument regarding the cloud-radiation-climate topic.

Presented in this report are mesoscale 2DT and 3DT cloud models with explicit microphysics developed by the author and suitable for simulation of various cloud types (Khvorostyanov, 1995). The models are based on two kinetic equations for droplet and crystal size distribution functions with division of droplet and crystal size spectra into 30 bins from 1 micrometer to 3.5 mm and with detailed account for CCN and ice nuclei activation, condensation/deposition and coalescence/accretion growth. These equations are being solved along with supersaturation equation and with radiative transfer equations. Simple but accurate analytical expressions are derived for scattering and absorption coefficients and single-scattering albedo. This allows us to make detailed calculations of the optical and radiative characteristics of clouds (fluxes, divergence, albedo) at each gridpoint of the computational domain with account for their time evolution along with evolving cloud microstructure and phase state. Two examples are considered here: (1) development of Ci clouds (2D model); (2) seeding from a contrails into underlying liquid As cloud (3D model). Cloud microphysics and interaction with longwave and solar radiation are described along with characteristics which can be measured by means of remote sensing: radiative fluxes, albedo and emissivity for various satellite channels, radar reflectivity. Such information can be used for planning and analysis of field experiments.

The Stratospheric Aerosol and Gas Experiment (SAGE III) is the fourth generation of solar occultation instruments designed to measure aerosols and trace gas species in the stratosphere and upper troposphere. It is scheduled to be launched aboard a Meteor-3M platform in the summer of 1998 and the International Space Station Alpha in 2001. SAGE III preserves the robust characteristics of the SAGE series, including self-calibration and high vertical resolution, and adds new capabilities including a lunar occultation mode. This paper describes the SAGE III instrument and outlines its potential contribution to global change research.

The tomography of the Earth's atmosphere by the solar occultation method leads to a highly non-linear inverse problem if the full solar disc is used as the light source. Well known heuristic methods like Chahine's algorithm or onion peeling fail to solve the inversion. We present a new method referred to as NOPE (for natural orthogonal polynomial expansion) that addresses this class of inverse problems by focusing on the morphological content of the unknown profile and allowing also a fine tuning of the a priori information.

The possibility of utilizing limb scattering to monitor stratospheric ozone from space has been recognized as a possible remote sensing technique for many years. Unfortunately, due to the complexity of the radiative transfer problem associated with the spherical shell geometry and associated multiple scattering, the problem has received relatively little attention. While the complexity of the problem remains, the development of newer codes allows for the problem to now be tractable. Utilizing one such recently developed code, the authors have made a series of sensitivity studies relating changes in the limb radiances to changes in the ozone profiles. The calculations allow for the spherical geometry, include all orders of multiple scattering, and may include aerosols and other absorbing gases when relevant. Calculations of the changes are presented for a series of standard ozone profiles and for perturbed profiles. The results of these calculations indicate that measurement accuracies on the order of plus or minus 1% should provide adequate sensitivity to determine the significant structure of the ozone profile. Due to the complexity of the spherical, multiple scattering problem, direct inversion of the measured radiances is probably not feasible at this time. However, table look-ups provide a reasonable alternative, much as is currently done in the TOMS/SBUV retrieval algorithms. Results of the sensitivity studies for a range of conditions are presented.

A balloon borne instrument BALLAD (balloon limb aerosol detection) has been developed at the LOA (Laboratoire d'Optique Atmospherique). It scans the Earth's limb at three wavelengths (450, 600, and 850 nm) from the float altitude between 30 - 35 km when the sun is low above the horizon; polarization is also measured in the 850 nm channel. A flight has been performed from the southwest of France on October 13, 1994 during the phase II of the SESAME (Second European Stratospheric Arctic and Mid-latitude Experiment) campaign. An analysis of the reflectances at 850 and 450 nm without polarization is presented.

A method based on the absorption of Chappuis band in the visible region has been developed in order to retrieve column ozone content together with aerosol optical depth by means of direct spectral irradiance measurements. Due to the great sensitivity in this spectral region of ozone to aerosol contribution and vice versa a careful supervised fitting using adequate windows is carried out. The fit is according to the Angstrom formula. Taking either the modeled aerosol optical thickness or the experimental one the aerosol size distribution is obtained by an inversion method. Preliminary results from March to July of 1995 are shown in the area of Valladolid, located in the center of the Castilla-Leon region, in Spain. The retrieved values of ozone content are according to the expected ones, but variations during this period are clearly manifested. Also aerosol optical depths for this period are monitored together with the aerosol characteristics.

A brief summary of aerosol optical depth measurements in a maritime atmosphere during the last three decades is presented. The results of more than fifty publications have been analyzed and are summarized in a single table. New results of spectral aerosol optical depth measurements (from 440 to 1030 nm) in the Mediterranean Sea and Atlantic Ocean made from aboard a research vessel are also presented. Comparison of aerosol optical depths obtained over the Mediterranean Sea in the winter 1989-1990 with other Mediterranean data indicate substantial seasonal difference. The angstrom parameter values for the central and western Atlantic indicate good agreement with the results obtained for the north Atlantic. The measurements in the subtropical Atlantic region show significant variations. The pure atmosphere in the winter 1989-1990 evolved in the fall of 1991 into very turbid conditions which were probably associated with Saharan dust.

The results of applying empirical orthogonal functions (EOF) to the decomposition and approximation of the marine and land aerosol size distribution profiles are presented. The calculated empirical orthogonal functions were used to analyze the spatial and temporal variability of aerosol size distributions. The marine aerosol data were collected during cruises on the southern Baltic Sea in April and May 1994, the land one -- during the coastal experiment in Lubiatowo Station in September 1994.

This work is mainly focused on the application of the line-by-line (LBL) radiative parameters databases in processing satellite low-resolution measurements of outgoing solar and IR radiation associated with two space instruments such as the scanner of radiation budget (ScaRab on Meteor-3/7 experiment) and ISTOK-1 (PRIRODA-MIR mission). Details of the efficient computer techniques for routine but accurate calculations of solar and IR radiative parameters in order to develop the benchmark databases are briefly described. Some practical results are presented that involve the utilization of line-by-line benchmarks in the ScaRaB measurement validation. The narrow-band model designed for the ISTOK-1 measurement processing and developed on the basis of the LBL precalculations is presented.

The study of the Earth radiation budget is important for understanding the long-term impact of industrialization on the environment. Clouds exert the single most important influence on the Earth radiation budget. The ideal situation for predicting the effect of clouds is to have one of two extremes: either a cloud-free sky or a completely overcast sky. Each of these cases can easily be treated with a simple one-dimensional model. However, the more usual partly-cloudy sky condition introduces the possibility of an unlimited variety of cloud geometries, size and spatial distributions, and properties. In most of these cases the interaction among neighboring clouds presents a set of problems which cannot be accurately simulated with plane-parallel or even two-dimensional models. Only in recent years have researchers begun to understand the relationships between cloud three-dimensional effects and the anisotropy of Earth-reflected radiance. In the real world, the influence of clouds on radiation depends not only on liquid water content, or optical depth, but also on cloud microphysical properties such as particle shape, size distribution, and on cloud morphology and spatial distribution (Parol, et al., 1994). As shown by Stevens and Greenwald (1991), cloud morphology is likely to have a larger impact on the Earth radiation budget than cloud microphysics. Thus, it is very important to quantify and parameterize the influence of cloud inhomogeneities on the radiation field. The anisotropy of any radiation field can be studied by normalizing the emitted or reflected radiance by the equivalent Lambertian directional distribution of energy leaving the field. The work in this paper deals entirely with the shortwave portion of the spectrum; longwave emitted energy is not treated.

The influence of the top altitude of terrestrial liquid water clouds on the polarization of sunlight reflected by a cloudy atmosphere is theoretically investigated. It is compared to the influences of other cloud properties and of stratospheric aerosols. Using a typical atmosphere model, we show how accurately the cloud top pressure may be derived from nadir measurements performed by satellites of the polarization at 350 nm. The accuracy of the derived pressure is about 5 mb under favorable conditions, when the accuracy of polarization measurements is 0.1%, and it depends mainly on knowledge of the density of the cloud and of the stratospheric aerosol optical thickness. The stratospheric aerosol optical thickness may be estimated with an accuracy of 0.02 using observations of the polarization at 670 nm having an accuracy of 0.1%.

POLDER (polarization and directionality of the Earth's reflectances) is a recent French instrument devoted to the global observation of the polarization and directionality of solar radiation reflected by the Earth surface-atmosphere system. The instrument will be launched on the Japanese ADEOS platform in 1996. One of the scientific objectives of the satellite mission concerns cloud characteristics and the Earth radiation budget. An original feature of POLDER is its ability to provide quasi-simultaneous multidirectional radiance measurements of any scene, that means to observe a part of its bidirectional reflectance distribution function (BRDF). This paper focuses on the analysis of BRDFs acquired by the airborne version of POLDER during the CLEOPATRA, ASTEX and EUCREX campaigns over various cloud types (stratocumulus, altostratus and cirrus clouds).

The global distribution of total column ozone ((Omega) ) is attracting great international attention as concerns over reduced global abundances escalate. Detection of a trend is an arduous task, made difficult by numerous natural inter- and intra-annual fluctuations, many of which are not well understood. Accordingly, this study analyzes these natural variations (across all spatial and temporal scales) through the application of rotated principal component analysis (PCA) to the (Omega) data derived from version 6.0 TOMS (total ozone mapping spectrometer) for the period 1984-1989. Utilization of Kaiser's varimax orthogonal rotation allowed delineation of eleven homogeneous subregions that together accounted for 74.08% of the total variance. Each subregion displayed statistically unique (Omega) characteristics that were further examined through time series and spectral analysis, allowing identification of the probable phenomena (i.e. annual and semiannual cycles, QBO, ENSO, baroclinic waves) responsible for the variability of (Omega) .

The authors present a method that can be used to retrieve the stratospheric ozone concentration, the aerosol extinction, and the microphysical aerosol characteristics. The method is based on the direct solar radiation extinction measurements from the satellite multichannel spectrometer 'Spektr-256'. The Angstrom's approximation of the aerosol extinction of the wavelength dependence in 0.45 - 0.83 mkm is being realized to separate the aerosol and ozone optical thickness. The Tichonov's method for reconstruction is being used. The aerosol type and the vertical aerosol refractive index suppositions give possibility for every altitude step -- 65 m to calculate the aerosol size distributions, the average aerosol radius and width size distributions. The above-mentioned remote sensing method is being applied by the authors to the serial experimental dates 1993-1994. In remote sensing time the aerosol size distributions have been taking of the single-modal shape distributions, generally, are being shown the calculating maps of the size distributions. The presence of the considerable fluctuations, the average aerosol radius, and width size distributions are being permitted on assumption that the stratospheric aerosol was a band-like structure with thickness of the bands about 1 km.

The relation between the reflectivity of the atmosphere-surface system and the optical thickness of a homogeneous cloud layer in the atmosphere is investigated at 0.63 micrometer, which is the central wavelength of NOAA-AVHRR (advanced very high resolution radiometer) channel 1. A detailed radiative transfer model is employed, in which the multiple scattering and absorption resulting from cloud particles, molecules, aerosols, ozone and surface are fully taken into account. To estimate the sensitivity of the relation between atmospheric reflectivity and cloud optical thickness, the influence of variation of solar zenith angle, surface albedo, cloud particle effective radius and viewing geometry are examined. The relation between reflectivity and cloud optical thickness is mainly sensitive to solar zenith angle, surface albedo and viewing geometry. Therefore, these parameters have to be known for a retrieval of cloud optical thickness from reflectivity measurements. For actual retrievals, a database is prepared with atmospheric reflectivities for many solar zenith angles, viewing zenith angles, azimuth angles, surface albedos and cloud optical thicknesses. The retrieval method is applied to an AVHRR image of Sept 11, 1994. First results show a promising correlation with cloud optical thickness estimations from ground based measurements of direct solar irradiance.

During the ASHOE/MAESA campaign a series of observations of the UV-visible radiation field were obtained from the ER-2 aircraft with the composition and photodissociative flux measurement (CPFM) instrument. Observations were made at the limb and in the zenith and nadir directions over the spectral range 300 - 775 nm at 1 nm spectral resolution. Analysis of these data yield surface and cloud effects on the radiation field as well as the effects of polar stratospheric clouds and changes in column ozone. We have conducted model-data comparisons of the direct and scattered radiation field and compared these results to TOMS satellite overflights. The radiation field models used in the data analysis include the integral equation solution described by Anderson et al., DISORT and MODTRAN3. In addition, the resultant radiation field observations are utilized in photochemical models for comparison to the in situ trace constituent measurements. Finally, we have found that the CPFM oxygen atmospheric band observations can be used to detect effective cloud heights.

The polar ozone and measurement (POAM II) instrument has been collecting data on the vertical distribution of ozone and other species in the polar stratosphere since its launch on the SPOT 3 satellite early in autumn 1993. POAM II makes solar occulation measurements at nine wavelengths from 35 to 1059 nm, allowing it to retrieve ozone, aerosols, water vapor, and nitrogen dioxide with about a 1 km vertical resolution. We will present data on the development and dissipation of the Antarctic ozone hole. We find that the bulk of the depletion occurred during the month of September, with ozone destruction rates exceeding 2% per day. The POAM II measurements clearly demonstrate the strong containment properties of the polar vortex.

Vertical concentration profiles of three trace species, HCl, CO, and NO, in the middle atmosphere have been retrieved from solar occultation infrared spectra recorded by the Grille spectrometer during the ATLAS-1 mission with a spectral resolution of about 0.1 cm^-1. HCl and NO profiles are compared with ATMOS interferometer measurements obtained during the ATLAS-1 mission, and with HALOE (Halogen Occultation Experiment on the Upper Atmosphere Research Satellite) measurements performed over the same period (24 March - 1 April 1992) and latitude range. A large increase in HCl abundance is estimated by comparison of HCl profiles produced by the Grille spectrometer with profiles obtained from the ATMOS interferometer during the 1985 Spacelab-3 mission. The estimated increase rate is in accordance with the rate deduced from ATMOS measurements performed in 1992 during the same mission ATLAS-1, and with model predictions. NO mixing ratios are in general agrement with HALOE and ATMOS profiles and reported measurements from other instruments. CO profiles extend up to 100-110 km and show a large vertical gradient in CO volume mixing ratio in the mesosphere and the lower part of the thermosphere, as expected by chemical models and observed by various instruments.

The POLDER (polarization and directionality of the earth reflectances) instrument, to be launched in 1996 on the Japanese ADEOS platform, includes a channel which covers the 910 nm water-vapor absorption band (near IR), as well as a channel at 865 nm. It is expected that the ratio of the two reflectance measurements can yield an estimate of the total atmospheric water vapor content. The major uncertainties on this estimate result from (1) the surface reflectance spectral signature, (2) scattering by atmospheric aerosol, and (3) the water-vapor vertical profile. A radiative transfer model has been developed in order to quantify these uncertainties. From radiative transfer simulations, an uncertainty on the order of 10% is expected on the total water vapor amount. Moreover, an airborne version of the POLDER instrument has been developed and flown over various targets. These targets include semi arid surfaces (Sahel), bog, coniferous and deciduous forest (Boreal forest) and the ocean. Water vapor measurements from radiosondes, concomitant with the POLDER measurements, are available. These data are used for the method validation.

The monochromatic and color twilight atmosphere images based on the numerical solution of the radiation transfer equations are presented. The twilight atmosphere brightness field calculations are carried out for the wavelength range of 400 - 750 nm. The calculation results are compared with monochromatic brightness values and color characteristics of the twilight aureole of the Earth's atmosphere obtained from the manned spacecraft 'Soyuz-5'. Effects of the multiple radiation on the twilight atmosphere brightness field are investigated. The calculated images for the atmosphere with aerosol component and without it are compared. The comparison analysis results are aimed at possible diagnostics of the aerosol atmosphere.

To analyze the optical tomography reconstruction (OTR) methods applications for reconstruction of meso- and large-scale irregularities of minor gas and aerosol components in the atmosphere and near Earth space (NES) the principal model problems of OTR are considered. The results of computer simulation to reconstruct the model irregular structure in NES are presented and discussed.

Since the conception of the Global Positioning System (GPS) its applications have grown to provide not only a navigation, but also a geodetic, and more recently a meteorological tool. This latter application has become termed GPS Meteorology. Atmospheric research in GPS initially centered around the premise that the atmosphere was a nuisance parameter (NP). To obtain the highest precisions (particularly in height) this NP has to be modeled, or estimated, in addition to the geodetic parameters. One of the most successful techniques for dealing with this NP has been to stochastically estimate the NP using a Kalman filter. As a by-product to the precise geodetic parameters obtained, tropospheric wet delays (the NP) can also be output. These tropospheric wet delays (TWD) can be related to integrated precipitable water vapor (IPWV) through the simplified equation: TWD equals k(DOT)IPWV. The dimensionless value of k is dependent on the weighted mean temperature of the atmosphere and can be predicted from surface temperature alone (but typically ranges from 6.0 to 6.5). The expansion of the International GPS Geodynamics Service (IGS) network provides the possibility of near-continuous measurements of IPWV anywhere on the Earth by using GPS as a passive remote sensing tool. Such a possibility provides applications for numerical weather prediction (NWP) models and climate and global change research. This paper outlines the current status of atmospheric research in GPS. A project aimed at validating the use of GPS for meteorology at Nottingham is discussed using observations from ground-based water vapor radiometers (WVRs) for validation.

The water vapor is an atmospheric component which is not evenly distributed. If we consider the total water vapor content in a vertical column, i.e. the precipitable water, W it may change from 0.5 g/cm² for high latitudes to about 6 g/cm² for equatorial zones. All that makes it impossible to know its distribution with only radiosondes data, because their validity is restricted to a few km². In order to correctly characterize the water vapor content over the sea around the Canary Islands, we have used brightness temperatures of TOVS sensors onboard NOAA satellites. We have compared the total water vapor content obtained with radiosondes data, launched by Instituto Nacional de Meteorologia at Santa Cruz de Tenerife, with the nearest free-clouds data from satellite for a set of fifty days of 1994. The statistics generated with the comparison are shown and the validity of the humidity fields determined with only TOVS data are discussed.

Mean daily values of the total atmospheric optical depth have been obtained from measurements of spectral solar irradiance at ground level in Valencia, Spain. These measurements have been taken during ten days in the years 1993 and 1994. The total columnar ozone amount and aerosol optical depths have been calculated using both King and Byrne and by Flittner at al. methods. The results obtained show that these algorithms lead to big errors if they are employed to determine instantaneous values of total ozone content. If they are used to calculate mean daily values, both methods give similar results either for the total ozone content or the aerosol optical depth, with quite acceptable errors. Considering the errors introduced by any one of the two methods, King's algorithm leads to higher imprecision in the aerosol optical depth determinations. This imprecision is particularly significant when the curve of the aerosol optical depth as a function of wavelength differs from the exponential law proposed by Angstrom.