Vol. 105, No. 4 * Pages 189–303 * October 2001 - March 2002 (Vol. 106, No.1)

The University of Maryland, Global Energy and Water Cycle Experiment, Shortwave Radiation Budget algorithm was used with satellite data from the D1 product of the International Satellite Cloud Climatology project to estimate monthly values of the shortwave top of atmosphere (TOA) and surface fluxes for the period of July 1983–June 1994. Based on this data set, regional, zonal and global averages of absorption and cloud radiative effects are examined in the total shortwave spectrum. Absorption of radiation varies significantly both regionally and with latitude. Out of the 342 W m^–2 of radiation arriving annually from the Sun at the top of the atmosphere 240 W m^–2 is absorbed by the Earth. Annual global mean atmospheric and surface absorption are larger by 11 and 34 W m^–2, respectively, for land than for ocean. The top of atmosphere radiation budget is in good agreement with those from the ERBE and ScaRaB missions.The global annual average SW flux absorbed at the surface from this study is 165 W m^–2, and agrees well with other satellite studies. However, this value is about 20 W m^–2 larger than the absorption indicated by studies that employ radiation data measured at ground sites. SW cloud radiative forcing at the TOA and the surface from this study is –46 W m^–2 and –49 W m^–2, respectively.

Based on the short- and long-wave radiative fluxes in the high-mountain positions Skalnaté Pleso and Stará Lesná the influence of the atmospheric boundary layer on the radiative fields in the High Tatras is studied. Besides the time and space variability of the radiative fluxes the attention to the emissivity, radiative cooling and heating in the investigated atmospheric layer as well as the influence of the low clouds on the long-wave radiation balance is paid.

In the past 10 years an enormous progress was made in measuring long-wave radiation. In this study the measurements of atmospheric long-wave radiation, which have been carried out with different types of pyrgeometers at the German BSRN station at Lindenberg since October 1994, are compared with model calculations by LOWTRAN 7 at clear sky. It was found that since November 1996, the measurements have been matched the calculated values. Comparison of measured and modeled values is a good method to check the data quality.

The First Global Positioning System (GPS) satellite was launched in 1978, but the history of GPS meteorology (radio occultation) began in the 1960’s with the beginning of the interplanetary flight. The studies of GPS meteorological applications appeared shortly after GPS system started to operate. In the literature many papers can be found about both the space-based and ground-based GPS applications; several papers discuss the opportunity of using GPS data for NWP and climatological applications. In this paper another GPS meteorological application is described. Radiances from current polar orbiting infrared (IR) and microwave (MW) sounders are used to derive temperature and moisture profiles in the troposphere in a physical retrieval algorithm. Specifications of the tropopause and the surface are necessary information in the profile retrieval. While the altitude of tropopause can be difficult to define in radiometric approaches, the GPS provides an opportunity to derive very accurate upper atmospheric temperature profiles by using radio occultation (RO) techniques. In this paper we show that the combination of radiometric (IR and MW) and geometric (RO) information yields improved tropospheric temperature and moisture profiles when compared to those inferred from either system alone.

The increasing number of IR spectral channels of recent satellite imagers implies the more accurate retrieval of surface skin temperature. In this paper the theoretical accuracy limits as a function of channel numbers, viewing angles and noise equivalent radiances have been studied. Based on LBL computations of the channel radiances regression type relationships have been established between the brightness temperatures and the skintemperatures. In this study model filter functions of the 7 IR channels of the GLI imager of the ADEOS-I have been used with a large set of temperature profiles. The global and seasonal distribution of temperature profiles were considered by groups of climatologically representative temperature profiles. Standard singular value decomposition program package was used to solve the multivariable linear regression problem. Results show that the accuracy limits of the skin temperature retrievals are depending mainly on the noise equivalent radiances and up to around 60 degree viewing angles remaining in the range of 0.1–0.2 K.

The long data record from operational environmental satellites allows the long-term, large-scale mapping of various characteristics of the Earth. This paper focuses on the diurnal temperature range (DTR) over land. The availability of necessary daytime and nighttime infrared radiation measurements within the various long-term satellite-based data sets is discussed. We conclude that currently the most suitable data sets for global DTR mapping are those that were produced by projects originally designed for the observation of atmospheric parameters: the International Satellite Cloud Climatology Project (ISCCP; mostly geostationary at low and mid-latitudes) and the Pathfinder Atmosphere from AVHRR (PATMOS; polar only). While ISCCP has more frequent sampling of the diurnal temperature cycle, there exists a large data gap over Asia due the lack of geostationary satellites. It is suggested that PATMOS data be used to fill this gap. However, daytime and nighttime PATMOS data need to be corrected for the effects of the local times of observation, which are not coincident with the times of occurrence of the daily maxima and minima. We demonstrate that, over hot surfaces, significant differences exist between coincident PATMOS surface temperatures and those from ISCCP corresponding to the same local time, indicating differences in calibration and/or in deriving surface skin temperature from top-of-the-atmosphere radiances. Residual differences also exist between coincident ISCCP and PATMOS daytime-nighttime temperature differences, suggesting that appropriate ISCCP to PATMOS normalization be made prior to adjusting PATMOS data for diurnal cycle effects to derive DTR maps.

This paper examines three-dimensional (3D) radiative effects, which arise from horizontal radiative interactions between areas that have different cloud properties. Earlier studies have argued that these effects can cause significant uncertainties in current satellite retrievals of cloud properties, because the retrievals rely on one-dimensional (1D) theory and do not consider the effects of horizontal changes in cloud properties. This study addresses two questions: which retrieved cloud properties are influenced by 3D radiative effects, and where 3D effects tend to occur. The influence of 3D effects is detected from the way side illumination and shadowing make clouds appear asymmetric: Areas appear brighter if the cloud top surface is tilted toward, rather than away from, the sun. The analysis of 30 images by the Moderate Resolution Imaging Spectroradiometer (MODIS) reveals that retrievals of cloud optical thickness and cloud water content are most influenced by 3D effects, whereas retrievals of cloud particle size are much less affected. The results also indicate that while 3D effects are strongest at cloud edges, cloud top variability in cloud interiors, even in overcast regions, also produces considerable 3D effects. Finally, significant 3D effects are found in a wide variety of situations, ranging from thin clouds to thick ones and from low clouds to high ones.

Satellite based cloud information is an essential input for the nowcasting system recently developed at the Hungarian Meteorological Service (HMS). This paper describes the cloud classification method used operationally for half-hourly Meteosat images. It was worked out in frame of a co-operation between HMS and Météo-France. A pixel by pixel algorithm is applied using auxiliary data as well such as the actual surface temperature and surface reflectivity. For the classification we use the results of the test runs of clustering type cloud classification methods earlier developed by Météo-France and by HMS separately. The monthly averaged kernels of those clusters are actualized as functions of the surface parameters pixel by pixel and images are classified according to these actual kernels. Verification was made by the forecasters of HMS. We also examined the possibility of applying the method for filtering noisy radar images. The results of a special study comparing the radar and satellite based cloud top height values according to cloud types gave new considerations for the development of the cloud classification method as well.

Two random variables are defined, one by an initial ensemble perturbation at a given geographical location, and another by the analysis uncertainty at the same location. The initial ensemble perturbations are said to be statistically consistent with the analysis uncertainty if the probability density functions of the two aforementioned random variables are equal. In this paper, it is shown how observed data can be used to test this statistical consistency. The usefulness of the proposed approach is demonstrated by an application to the global ensemble forecasting system of the National Centers for Environmental Prediction. Targeted dropsonde observations, collected during the 2000 Winter Storm Reconnaissance program, are used to show that increasing the rescaling frequency in the breeding cycle from once a day to four times a day improves the consistency of the initial ensemble perturbations.