This data set is only available for a restricted user group, please contact us if you want to access these data.
RESTRICTED only accessable in CEN/MPI net or via CliSAP login What does that mean?
This data set is based on observations of two vertically profiling satellite sensors: a radar (CloudSat) and a lidar (CALIPSO-CALIOP). The CloudSat radar operates at a frequency of 94 GHz (wavelength: about 2 mm). The CALIPSO lidar operates with three channels: two at 532 nm which perpendicular to each other polarization and one at 1064 nm.
CloudSat and CALIPSO (both launched April 28, 2006) are part of the so-called "A-Train" constellation and fly along the same orbit at a distance of only 10-15 seconds (CloudSat leads) so that footprints of both sensors overlap, allowing to sense the same volume with the combinations of different frequencies involved.
CloudSat provides vertical profiles of the radar cross section and thus the refraction index, which is a measure for the concentration of cloud and precipitation particles in the sensed part of the atmospheric column, at a range resolution of about 240 m. CALIPSO-CALIOP provides three vertical profiles of the laser energy scattered back by aerosol and cloud particles at a range resolution of about 15 m; again the amount of the energy scattered back is a measure of the concentration of scattering particles in the sensed part of the atmospheric column.
Combination of the data obtained with both sensors allows to obtain an almost complete characterization of the vertical cloud structure because the two sensors couple with their different wavelengths used (CloudSat: about 2 mm, CALIPSO: 532 nm and 1064 nm) to different cloud and precipitation particles of both phases, liquid and solid. While CloudSat is rather insensitive to thin, high cirrus clouds CALIPSO can reliably detect these even if these are very thin. In contrast, some clouds can be opaque enough to prevent CALIPSO from seeing what is below, while CloudSat can penetrate these clouds, resolving what is underneath. Another advantage is, that in contrast to most other vertical profiling sensors, both sensors used here provide temperature independent results.
A big disadvantage is, however, the footprint size (about 1 km for CloudSat, about 70 m for CALIPSO) - similar to satellite altimeters used to remotely sense the ocean surface (see SSH and sea ice thickness). Because of this, data from many satellite overpasses have to be averaged / combined to a product with comparably coarse spatio-temporal resolution (in this case: 2° grid resolution and monthly temporal resolution) in order to obtain a reasonable statistics.
The data set contains the cloud fraction per grid cell. Different data sets exist for low, medium high and high clouds as well as the total cloud cover; we also offer the mean vertical profile for each grid cell. Each file contains the number of cloud pixels (per grid cell), the total number of pixels (per grid cell), and the ratio of both = cloud fraction of the respective grid cell.
|Number of cloud pixels per grid cell||--|
|Number of data pixel per grid cell||--|
|Cloud fraction||-- (0 ... 1)|
Period and temporal resolution:
- 07/2006 to 02/2011
Coverage and spatial resolution:
- Spatial resolution: 2° x 2°, cartesian grid
- Geographic longitude: -180°E to 180°E
- Geographic latitude: -90°E to 90°E
- Dimension: 180 columns x 90 rows for total ("total") cloud cover, and coverage of high ("high"), middle ("mid") and low level ("low") clouds; 180 columns x 90 rows x 40 levels for product "profile"
- Altitude: variable for "total", "high", "mid", and "low"; every 480 m for "profile"
Upon request we offer seasonal averages as well as monthly averages over the entire period.
The offered data sets do not contain uncertainty estimates. However, according to the report of the GEWEX Radiation Panel cloud assessment CALIPSO-CloudSat cloud cover data compare well with other cloud cover products - in particular when it comes to the cloud cover of the single cloud layers (high, middle and low).
A clear disadvantage is, however, the coarse spatial and temporal resolution of the data set. Even the monthly averages offered here seem noisy and suggest to go for seasonal averages. This is caused by the nadir looking viewing geometry of both sensors which result in a narrow sub-satellite track to be monitored as the satellite passes over.
More details can be found in Kay and Gettelman, 2009.
UCAR, Colorado, U.S.A.
email: jenkay (at) ucar.edu
ICDC / CEN / University of Hamburg
email: stefan.kern (at) uni-hamburg.de