ISCCP D2 SELECTED VARIABLE DESCRIPTIONS
This variable represents the frequency of occurrence of cloudy conditions in individual satellite image pixels, each of which covers an area of about 4 to 49 square kilometers. Comparisons to other measurements confirm that this quantity also represents fractional areal coverage at one time for the larger 280 km grid cell areas. Clouds are detected by tests of infrared radiances in nighttime and by separate infrared and visible radiance tests in daytime.
In the gridded ISCCP products, several cloud types are defined to give more detailed information on the variations of cloud properties. These cloud types are defined by the VIS/IR cloud top pressure and optical thickness or by the IR cloud top pressure alone. In the D-series datasets, there are three IR cloud types: low, middle and high: low clouds have top pressures greater than or equal to 680 mb, high clouds have top pressures less than 440 mb, and middle clouds are in between. The VIS/IR cloud types divide clouds in the three pressure levels by optical thickness range as shown in Figure 1. In the D-series datasets, low and middle cloud types can be either liquid or ice depending on temperature. The names used to refer to particular combinations of cloud top pressure and optical thickness are the classical morphological cloud types, but the association of particular ranges of cloud properties with these morphological types is only qualitative.
Cloud Top Temperature/Pressure
These variables are determined from the infrared radiances measured by the satellites in cloudy conditions and the correlative data that indicate the temperatures and humidities at various pressure levels in the atmosphere. The temperature represents the amount of infrared radiation emitted by the cloud and the pressure represents the vertical location above mean sea level of the cloud top that corresponds to its temperature. Since some clouds are partially transparent to infrared radiation, their actual top temperatures may be somewhat colder and their estimated cloud top pressures may be somewhat lower than given by these values. In daytime conditions when cloud optical thickness values are obtained from the visible radiances, the cloud top temperature/pressure values can be corrected for transmission of radiation through partially transparent clouds. The resulting values of temperature/pressure are lower than the IR cloud top temperature/pressure values. The magnitude of the difference between these values is larger for smaller optical thicknesses.
Cloud Optical Thickness
This variable is determined from the visible radiances measured by the satellite in cloudy conditions and represents the amount of solar radiation at visible wavelengths reflected by the clouds (the amount of infrared radiation absorbed by the clouds is also related to this variable). Since some clouds are partially transparent to solar radiation, this variable is obtained by accounting for the radiation reflected from the surface using the surface visible reflectances obtained from clear scenes. No results are reported for nighttime conditions or in the un-illuminated (winter) regions near the poles. The cloud optical thickness, together with the surface reflectance and cloud amount, determine how much solar radiation is reflected back to space by each location on Earth.
Cloud optical thickness is retrieved from observed visible radiances using two microphysical models in the D-series datasets. One model is a liquid water cloud composed of spherical droplets with a cross-section-weighted average radius of 10 microns and a variance of the size distribution of 0.15. This model is used for clouds with top temperatures greater than or equal to 260K in the D-series datasets. The second model is an ice cloud composed of fractal polycrystals with a cross-section-weighted radius of 30 microns and a variance of the size distribution of 0.1. This model is used in the D-series datasets for clouds with top temperatures less than 260K.
Cloud Effective Particle Radius
This parameter represents an average over the distribution of particle sizes in a cloud. It can be estimated from satellite measurements by contrasting the amount of radiation scattered and absorbed by the cloud: the absorption depends on the particle volume and the scattering depends on the particle cross-section, so their ratio gives an estimate of a linear dimension of the particle. For liquid water droplets, which are spherical in shape, this relationship can be precisely defined. For ice crystals in a variety of shapes, this relationship can only be approximated. Measuring reflected sunlight at two wavelengths provides a particle size near cloud top; combining measurements of scattered sunlight with microwave absorption provides a particle size averaged over the whole cloud layer.
This variable is retrieved from the IR radiances measured under clear conditions and correlative data describing the temperature and humidity profile in the atmosphere. The surface temperature represents the amount of IR radiation emitted by the liqiud or solid surface and is equal to the actual temperature of the surface when the surface emissivity is equal to one. Most surfaces have emissivities less than but nearly equal to one, so the actual surface temperatures are slightly larger than the values given here: water surface temperature will be about 1.0 K larger and land surface temperatures will be 1.0 - 3.0 K larger depending on the amount of vegetation.
This variable is retrieved from the VIS radiances measured under clear conditions and correlative data describing the ozone column abundance. The surface reflectance, particularly of water surfaces, varies with viewing and illumination geometry. Brighter areas over oceans at lower latitudes are caused by the occurrence of sunglint. Large reflectances in polar areas are caused by snow and ice cover.
Total Column Water Vapor
This parameter represents the total precipitable centimeters of water vapor in the atmosphere and is determined from analysis of satellite infrared sounder data (NOAA operational analysis). Since the result comes from measurements of absorption at infrared wavelengths, the results are strictly valid only for relatively cloud-free locations (cloud cover fraction over a 300 km region less than about 60%). The original data report water vapor amounts for three layers covering the lowest part of the troposphere (approximately from the surface to the 300 mb level -- there is only a very small amount of water vapor above this level) and are sampled at intervals of about 300 km and 1 day.
Ozone Column Abundance
This parameter represents the total amount of ozone in the atmosphere (in Dobson units) and is determined from satellite infrared sounder data (NOAA operational analysis). The ozone amount is determined from the absorption at about 9.6 microns wavelength. The original data are sampled at intervals of about 300 km and 1 day.
Atmospheric Temperature at 500 mb
This parameter represents the average atmospheric temperature over the pressure layer from 560mb to 440mb and comes from the analysis of satellite infrared and microwave sounder data (NOAA operational analysis). Comparison of the microwave data, which is relatively unaffected by the presence of clouds, to the infrared data is used to remove cloud effects. Infrared emissions at several wavelengths in the 13-15 micron absorption band of CO2 are used to determine temperatures for 9 layers in the troposphere. The original data are sampled at intervals of about 300 km and 1 day.
This parameter represents the minimum layer-mean temperature found in the upper atmosphere from the analysis of satellite infrared and microwave sounder data (NOAA operational analysis). The temperature is inferred from infrared emissions. The original data are sampled at intervals of about 300 km and 1 day.
This parameter represents the pressure at the level of minimum layer-mean temperature in the upper atmosphere from the analysis of satellite infrared and microwave sounder data (NOAA operational analysis). Pressures are assigned to temperature values by the amount of absorption by CO2 expected as a function of wavelength; this value is determined by an analysis of the temperature profile shape. The original data are sampled at intervals of about 300 km and 1 day.
Ice/Snow Cover Fraction
This parameter combines the snow cover fraction determined from analysis of visible wavelength satellite imagery (NOAA operational analysis) and the sea ice fraction determined from analysis of satellite microwave imagery. The presence of snow is determined from a brightening of the cloud-free scene and snow cover fraction from the number of image pixels containing snow for regions about 100 km across. The sea ice fraction is determined from the microwave brightness temperatures at two wavelengths. The ISCCP dataset comes from the Navy analysis for 1983-1992 and the NASA analysis (performed at NSIDC) after 1992. In regions with both land and water, the snow and sea ice fraction are combined to give a single fractional coverage value.
Surface Skin Temperature
This parameter represents the solid surface physical temperature. As part of the ISCCP cloud analysis, the clear-sky infrared (wavelength of about 11 microns) brightness temperature is estimated at intervals of 30 km and 3 hr. The surface skin temperature is retrieved from these values by correcting for atmospheric emission and absorption of radiation, using data for the atmospheric temperature and humidity profiles, and for the fact that the surface infrared emissivity is less than one. The values of surface emissivity used are shown in the Narrowband Infrared Emissivity dataset. Because these values are determined under clear conditions, they will over-estimate the average daytime - summertime maximum temperatures and underestimate the average nighttime - wintertime minimum temperatures.
Surface Narrowband Infrared Emissivity (11 micron)
This parameter is the infrared emissivity at about 11 microns wavelength used to correct surface infrared brightness temperatures to physical temperatures. This compilation is constructed from a classification of surface types (for land areas this uses surveys of vegetation and land-use) and documented emissivity values for different kinds of surfaces. These values are also used in the NASA GISS climate model.
Surface Broadband Infrared Emissivity (5-200 micron)
This parameter represents the spectrally averaged emissivity of the surface over the infrared wavelength range from 5 to 200 microns. To provide the correct spectral average, this value is determined by calculating the ratio of emitted infrared fluxes from Earth's surface without an atmosphere, using 3-hr and 30 km temperature data from each month in 1992, assuming a spectrally dependent emissivity compiled from surface classification and documented values for different surface types and an emissivity value of unity. Averaged in this fashion, the values depend on temperature and are, therefore, shown as a function of time and location.
Surface Microwave Emissivity
These parameters represent the surface emissivity determined at four microwave wavelengths (three of which are determined with both vertical and horizontal polarizations) from a combined analysis of satellite microwave and infrared emissions. The analysis uses the satellite infrared (and visible) data to identify cloud-free scenes; the infrared emissions are then used to determine the surface skin temperature for each scene. Co-located microwave measurements, corrected for atmospheric emission/absorption, can then be used to determine microwave emissivities since the temperature is known from the infrared analysis. The original data are sampled at intervals of about 30 km.
Surface Total Albedo
This parameter represents the spectrally averaged reflection of sunlight from the surface over the solar wavelength range from 0.2 to 5 microns. To provide the correct spectral average, this value is determined by calculating the reflected solar flux from Earth's surface without an atmosphere assuming a spectrally dependent albedo compiled from surface classification and documented values of albedo for different surface types. This compilation is also used in the NASA GISS climate model, but have been modified here so that the visible wavelength albedo agrees with clear-sky values determined from the ISCCP cloud analysis. Albedo averaged in this fashion is strictly a property of the surface; however, the actual reflected flux depends on the spectrum of sunlight actually reaching the surface, which is altered by the effects of water vapor, ozone and clouds. In addition these values (called the plane albedo) depend on the direction of solar illumination and hence, vary with season and location.