The IIR Level 3 (L3) Global Energy and Water cycle Experiment (GEWEX) Cloud Product
reports IIR cloud effective radius and water path averages and histograms on a uniform
2-dimensional 1° latitude by 1° longitude spatial grid. This product is designed
to follow the general guidance of the GEWEX Cloud Assessment. Cloud amount, radiative
temperature, effective emissivity, and optical depth characterize the cloud samples for
which IIR microphysical retrievals are reported. Cloud properties are provided for
atmospheric columns containing only ice clouds, only liquid water clouds, and only high
ice clouds of layer centroid pressure lower than 440 hPa. All level 3 parameters are derived
from the IIR version 4.20 level 2 (L2) track products
(Garnier at al., 2021),
with a temporal averaging of one month. The IIR algorithms take full advantage of
the co-located characterization of the atmosphere provided in the CALIOP version 4.20 level 5-km layer products.
Though the averages and histograms are the same as in the IIR CALIPSO-ST cloud
product provided to the GEWEX project, the data structure is slightly different.
Instead of reporting each cloud property as yearly files, this product includes all
cloud properties as monthly files. The product is represented as Hierarchical Data
Format (HDF) 4 format instead of Network Common Data Form (netCDF) 4 format. The
changes are made to keep consistency with other current L3 CALIPSO cloud and aerosol products.
The major categories of the IIR Level 3 GEWEX Cloud Product are:
Cloud amount averages
Cloud radiative temperature averages and histograms
Cloud effective emissivity averages and histograms
Cloud effective radius averages and histograms
Cloud water path averages and histograms
Cloud optical depth averages and histograms
Quality Filters Information
Valid IIR pixels: The valid IIR pixels are those with a valid radiance
in each of the three channels centered at 8.65 μm, 10.6 μm, and 12.05 μm,
and with a valid scene classification parameter.
Candidate clouds: the candidate clouds or “candidate cloudy scenes”
are selected as a pre-requisite for meaningful microphysical retrievals. The
selection criteria, which are listed and described below, are implemented using
the IIR scene classification and various quality flags reported in the
IIR L2 product.
The retrieved cloudy scene includes at least one cloud layer.
At least one cloud layer has a low, medium, or high confidence in cloud feature type.
No mixed aerosols/clouds scenes with aerosols above clouds (scene types 63 and 66 excluded).
One cloud layer, or several layers separated by less than 1 km.
No single shot cleared cloud in the column for scenes composed of semi-transparent layers.
Full column retrievals
The retrievals apply to all the clouds included in the column: the background
reference used for emissivity retrievals is the surface.
Centroid altitude ≤ 20 km
150 K ≤ Radiative temperature ≤ 320 K
Following the rationale developed in
Garnier et al. (2021),
we select the so-called confident retrievals, for which the two microphysical indices
derived from the pairs of IIR channels 12.05/10.6 and 12.05/8.65 are within the
expected range of values according to the look-up tables. These look-up tables are
computed using ice habit models
(Bi and Yang, 2017;
Yang et al., 2013)
and Lorenz-Mie theory for water clouds. In addition, we require Ice or Liquid Water
Path to be reported, meaning that the scenes include only ice clouds or only water
clouds with visible optical depth smaller than about 20.
Confident microphysical retrievals
Candidate cloudy scenes with good or medium confidence in particle shape index, and ice or liquid water path is reported.
Ice and water clouds
The cloud layer thermodynamic phase assignment is from the CALIOP version 4.20
level 5-km layer products. The CALIOP V4.20 algorithm
(Avery et al., 2020)
provides four phase assignments: Randomly Oriented Ice crystals (ROIs), Liquid water,
Horizontally Oriented Ice crystals (HOIs), and unknown phase. The presence of HOIs is
identified mostly from the beginning of the mission until November 2007, when the
platform viewing angle was close to nadir (0.3°). In contrast, HOIs are rarely
detected since December 2007 when the platform viewing angle was permanently changed
to 3°. For this first version of the product, we chose to define “Ice”
clouds as clouds classified as ROIs. We also require a high confidence in the phase
assignment. Cloud layers are classified as ROIs more frequently at tilt angle of 3°
than at 0.3°, because HOIs, if any, are typically not detected at 3°, and
therefore possibly included in the population of clouds classified as ROIs.
Differences between statistics derived from IIR observations conditioned by CALIOP
measurements at near nadir and 3° viewing angles are thus to be expected.
Candidate cloudy scenes with all the layers composed of Randomly Oriented Ice (ROI) crystals with high confidence in the phase assignment.
Candidate cloudy scenes with all the layers classified as liquid water clouds with high confidence in the phase assignment.
Quick description of the statistics
The statistics are provided for candidate cloudy scenes with confident microphysical retrievals.
The cloud amounts are derived with respect to the number of valid IIR pixels. The mean cloud
amounts in each grid point are the mean values of the cloud amounts derived from each individual
orbit track. Cloud amounts are reported for all confident microphysical retrievals, regardless
of the phase assignment, as well as for ice and water clouds as defined above. High ice
clouds are those ice clouds with centroid pressure smaller than 440 hPa.
Cloud radiative temperature
For a meaningful characterization of the layer temperature, we chose to report
the estimated cloud radiative temperature rather than the layer top temperature reported in the
CALIOP L3 GEWEX products.
This radiative temperature is estimated as the temperature at the centroid altitude
of the 532-nm attenuated backscatter, further corrected in case of ice clouds to
account for the difference in attenuation between visible and IR wavelengths, as
described in section 3.4 of
Garnier et al. (2021).
Mean values and histograms are for the ice, water, and high ice pixels with confident microphysical retrievals.
Cloud effective emissivity, cloud water path, cloud optical depth
The IIR parameters are from the IIR L2 track product, and they were derived as detailed in
Garnier et al. (2021).
Note that the effective radii reported in this product are simply derived as half
of the L2 effective diameters. Mean values and histograms are for the ice, water,
and high ice pixels with confident microphysical retrievals.
CALIOP optical depth
(Young et al., 2018)
means and histograms are provided for ice and high ice clouds. They are from the
Version 4.20 CALIOP 5-km layer products used as input to the IIR algorithm.
Retrievals in liquid water clouds
Microphysical retrievals in the thermal infrared atmospheric window perform best
for clouds that are sufficiently cold to ensure a sufficient radiative contrast with
respect to the surface. Therefore, the population of liquid water clouds with microphysical
retrievals is skewed toward supercooled water clouds.
Last Updated: November 22, 2021
Curator: Charles R. Trepte
NASA Official: Charles R. Trepte