The Single Scanner Footprint TOA/Surface Fluxes and Clouds (SSF) product
contains one hour of instantaneous Clouds and the Earth's Radiant Energy System
(CERES) data for a single scanner instrument. The SSF combines instantaneous
CERES data with scene information from a higher-resolution imager such as
Visible/Infrared Scanner (VIRS) on TRMM or Moderate-Resolution Imaging
Spectroradiometer (MODIS) on Terra and Aqua. Scene identification and cloud
properties are defined at the higher imager resolution and these data are
averaged over the larger CERES footprint.
Top-of-Atmosphere (TOA) and surface fluxes (shortwave, longwave, window)
Full footprint area parameters (imager coverage, wind, temperature,
humidity, extrapolation)
Clear area parameters (clear amount, aerosol)
Cloud area parameters for up to 2 layers (cloud amount, height,
temperature, pressure, optical depth, emissivity, and water path and particle
size for both water phases)
MODIS land aerosols (starting with Edition1A Terra, not available
for TRMM)
MODIS ocean aerosols (starting with Edition1A Terra, not available
for TRMM)
CERES footprints which fall outside of the imager swath do not appear on the
SSF data product. The maximum VIRS viewing zenith angle is ~48°, and the
maximum MODIS viewing zenith angle is ~65°. The imager swath restriction
does not apply to the CERES ES-8 product which contains a full set of CERES
footprints.
Additional information about the format and content of the SSF can be
found in the CERES
Data Products Catalog. A detailed description of the SSF can be found
in the SSF
Collection Guide. Data set specific information can be found in the
associated Quality Summary.
The CERES Data Management Team and the Atmospheric Science Data Center
(ASDC) at Langley use a Sampling Strategy, a Production Strategy, and a
Configuration Code (CCode) to track versions of CERES primary data products. In
general, minor reprocessing changes are tracked by increasing the Configuration
Code while major reprocessing changes result in a new Production Strategy. The
Sampling Strategy identifies the satellite and instruments which acquired the
data in the product.
A summary of changes made to the CERES SSF product is shown in the following
tables. Please note that Field of View (FOV) and footprint are used
interchangeably.
SW Surface Flux Model B module was modified to switch back from MATCH
broadband to Match GFDL aerosols. The following SSF SDSs will vary: SSF-46 and
SSF-48.
Software recompile to support upgrade to Toolkit version 5.2.12.
SW Surface Flux Model A module was modified to switch from Match GFDL to
MATCH aerosols at 550 nm. The following SSF SDSs will vary: SSF-38, SSF-41 and
SSF-44.
SW Surface Flux Model B module was modified to switch from Match GFDL to
MATCH broadband aerosols. The following SSF SDSs will vary: SSF-46 and
SSF-48.
LW Surface Flux Model B module was modified to correct cloud pressure
calculations. The following SSF SDSs will vary: SSF-39, SSF-47 and SSF-49.
Cloud Phase Algorithm was modified to have a more accurate cloud phase
selection. The following SSF SDSs will vary SSF-82 through SSF-114.
Water content map is re-generated to change defaults to zeros over
Antarctica region. The following SSF SDSs will vary: SSF-25, SSF-66, SSF-81,
SSF-116, SSF-117, SSF-82 through SSF-114.
The problem over polar/non-polar transition ocean areas (mostly occured in
Sea of Okhotsk and Hudson Bay in Terra-MODIS_Edition2-QC) is resolved by having
a fix ocean albedo over non-polar region. SSF SDSs: SSF-66, SSF-81, SSF-116,
SSF-117, SSF-82 through SSF-114 will vary over the transition ocean areas.
12 months of clear sky start-up maps are re-created for 0.6 um and 2.1 um
respectively. Clear sky maps are the inputs for all algorithms, therefore the
new maps will influence many SSF SDSs (mostly SSF-66, SSF-81, SSF-116,
SSF-117,SSF-82 through SSF-114) to some degrees.
The following summaries the changes in CERES mask algorithm, which will
affect SSF SDSs: SSF-66, SSF-81, SSF-116, and SSF-117.
Daytime non-polar:
Consolidated the Terra and Aqua daytime cloud mask into one set for
easier maintenance and higher efficiency.
Implemented the new daytime thin Cirrus detection tests over land and
desert using MODIS 1.38, 8.55 - 11, 11 - 12 tests and the MOA precipitable
water. It significantly improved the thin Ci detection over land. Daytime
CERES cloud mask structure has been changed.
In C3_land tests, improved the low clouds tests and the dark snow
detection at high latitudes, smoothed the transition regions.
Daytime polar:
Improved both Aqua and Terra mini-mask for a better classification of
the TBD pixels.
Reorganized and cleaned up the Aqua and Terra masks.
Twilight polar:
Improved the twilight Aqua and Terra cloud and snow detection for a
smoother transition between polar and non-polar.
When VISST algorithm returns no cloud retrieval, LBTM algorithm is applied.
This modification will improve SSF SDSs: SSF-82 through SSF-114.
CERES Overlapping Algorithm was implemented. None of SSF SDS is affected.
Clouds QC structure is modified to include histograms for several cloud
properties. None of SSF SDS is affected.
Footprints influenced by a solar eclipse will be removed from the file.
The Surface type percent coverage (SSF-25) land coverage is reduced based
on the water content map. The reduced area is added to the ocean areal
coverage.
The Snow/Ice percent coverage from vis albedo (SSF-30) now includes
permanent snow.
The GRing module was modified to use correct coordinates for the no
footprint case. A GRing that covers the entire globe was added.
Aqua dynamic Spectral Correction Coefficients were used.
All footprints with a viewing zenith greater than or equal to 63 degrees
are included in the SSF. When the viewing zenith angle is less than 63 degrees
and the packet number is even, then only footprints with an even value Scan
sample number are placed in the SSF. When the viewing zenith angle is less than
63 degrees and the packet number is odd, then only footprints with an odd value
Scan sample number are placed in the SSF.
All PSF weighted imager
parameters on the SSF (SSF-24 through SSF-26, SSF-30, SSF-54, SSF-64 through
SSF-78, SSF-81 through SSF-113, and SSF-131) will be obtained using only imager
pixels from every 2nd scanline and every 4th pixel (2x4) and a new point spread
function.
The Surface type percent coverage (SSF-25) sea ice areal coverage is
reduced based on the ice fraction reported on the microwave ice map. The
reduced area is added to the ocean areal coverage.
ADM geo (SSF-30) now contains the percent snow or ice for polar
region based on the overhead-sun albedo from CRH 0.6 µm map updated from
the day before. The threshold over ocean is equal to the average IGBP type
overhead sun albedo plus 3-sigma. The threshold over land IGBP types is its
average overhead sun albedo plus 2-sigma.
The Radiance and Mode flags (SSF-34) will identify when the
footprint was under a solar eclipse.
CERES ADM types and TOA fluxes-upward (SSF-27 through SSF-29 and
SSF-38 through SSF-41) will be derived using the new Terra Edition2A ADMs.
Notes on General Procedures (SSF-64) only footprints that have Stowe
aerosol calculated on them are flagged as using the second generation algorithm.
Total aerosol A optical depths (SSF-73 and SSF-74) will use satellite
specific solar constants as provided by Dr. Ignatov to determine reflectance.
The radiance channel for clear air optical depth using the Stowe algorithm
was made selectable based on satellite. For Aqua, it will be 0.6 and 2.13.
Clear, Clear/layer/overlap, All subpixel clear, All subpixel overcast
cloud area percent coverage (SSF-66, SSF-81, SSF-116, and SSF-117) will
vary due to
Added skin T restriction for Tlim test: (T > 270K)
Daytime Polar:
Added restriction to Cold Cloud Test, avoid calling Super Cold Plateau
clouds.
Added thin cirrus clouds detection.
Tuned various thresholds to improve the results
Added mini-mask for TBD pixels, no more TBD in CERES daytime mask.
Nighttime Polar:
All the improvements in the Terra Edition 2 (SCCR467).
Filter out the bad T3.7 data (T3.7 > 300K)
Daytime non-polar:
Added sun-glint probability dependent T3-T4 thresholds in SunGlintTest.
Improved smoke detection over ocean.
Added a snow test region flag to avoid "tropic snow"
Special cloud detection tests over super cold plateau for night and
twilight.
Welch mask will no longer be used when CERES mask in undetermined over
polar region during daytime.
Added 30% adjustment on IR thresthold for cloud test.
The polar mask is no longer applied in nonpolar regions where skin
temperature is less than 270 K
The imager pixel will not have an undetermined cloud fraction anymore.
Cloudy Footprint Area SDS (SSF-82 through SSF-114) will be improved
because over snow-ice non-elevated land, where an inversion cloud height was
calculated from GEOS lapse rate, the MOA skin temperature used in calculations
was replaced by the daily averaged MOA air surface temperature. Cloud thickness
and clouds top emissivity calculations now uses cloud phase instead of cloud
temperature.
Footprint Imager Radiance Statistics (SSF-118 through SSF-131) for
daytime footprints contain radiance for 2.13 micrometer.
The Percentage of CERES FOV with MODIS land and ocean aerosol
(SSF-132 and SSF-146) was corrected to represent the true coverage instead of
anything over 1 percent being reported as 100 percent.
PSF-wtd MOD04 corrected optical depths land, effective optical depth
average ovean, and optical depth small average ocean (SSF-136 through
SSF-138 and SSF-150 through SSF-159) now have an upper limit of 5.0
Aqua dynamic Spectral Correction Coefficients were used.
The Aqua processing variations began with a Terra Edition1A baseline.
Clear, Clear/layer/overlap, All subpixel clear, All subpixel overcast
cloud area percent coverage (SSF-66, SSF-81, SSF-116, and SSF-117)
will vary due to
New start-up maps for 2.1 µm albedo and 2.1 µm albedo standard
deviation were implemented.
The Kriebel Bidirectional model for 2.1 µm was implemented.
Both sub-Arctic summer and sub-Arctic winter snow reflectance models for
2.1 µm were implemented.
Welch mask will no longer be used when CERES mask in undetermined over
polar region during daytime.
2.1 µm daytime polar cloud mask used.
Improved polar nighttime and non-polar datime CERES mask.
Cloudy Footprint Area SDS (SSF-82 through SSF-114) will be improved
because
The 2.1 µm cloud model used in polar regions was implemented so
that the NoVis cloud retrieval algorithm automatically uses the 2.13 µm
cloud model.
The observed 1.6 µm reflectance and clear sky predicted 1.6
µm reflectance were replaced by 2.1 µm observed reflectance and
clear sky predicted 2.1 µm reflectance respectively.
PSF-wtd MOD04 dust weighting factor land (SSF-135) is filled with
default values.
Mean and Stddev of imager radiances (SSF-118 through SSF-129)
contain the same radiance channels as used for Terra.
Static Aqua spectral correction coefficients were used.
The Aqua processing variations began with a Terra Edition1A baseline.
All footprints with a viewing zenith greater than or equal to 63 degrees
are included in the SSF. When the viewing zenith angle is less than 63 degrees
and the packet number is even, then only footprints with an even value Scan
sample number are placed in the SSF. When the viewing zenith angle is less than
63 degrees and the packet number is odd, then only footprints with an odd value
Scan sample number are placed in the SSF.
All PSF weighted imager
parameters on the SSF (SSF-24 through SSF-26, SSF-30, SSF-54, SSF-64 through
SSF-78, SSF-81 through SSF-113, and SSF-131) will be obtained using only imager
pixels from every 2nd scanline and every 4th pixel (2x4) and a new point spread
function.
Clear, Clear/layer/overlap, All subpixel clear, All subpixel overcast
cloud area percent coverage (SSF-66, SSF-81, SSF-116, and SSF-117)
will vary due to
New start-up maps for 2.1 µm albedo and 2.1 µm albedo standard
deviation were implemented.
The Kriebel Bidirectional model for 2.1 µm was implemented.
Both sub-Arctic summer and sub-Arctic winter snow reflectance models for
2.1 µm were implemented.
Welch mask will no longer be used when CERES mask in undetermined over
polar region during daytime.
2.1 µm daytime polar cloud mask used.
Improved polar nighttime and non-polar datime CERES mask.
Cloudy Footprint Area SDS (SSF-82 through SSF-114) will be improved
because
The 2.1 µm cloud model used in polar regions was implemented so
that the NoVis cloud retrieval algorithm automatically uses the 2.13 µm
cloud model.
The observed 1.6 µm reflectance and clear sky predicted 1.6
µm reflectance were replaced by 2.1 µm observed reflectance and
clear sky predicted 2.1 µm reflectance respectively.
PSF-wtd MOD04 dust weighting factor land (SSF-135) is filled with
default values.
Mean and Stddev of imager radiances (SSF-118 through SSF-129)
contain the same radiance channels as used for Terra.
The 1.6 micrometer imager radiance values are saved for all footprints with
CERES solar zenith <= 90 degrees.
Cloud Phase Algorithm was modified to correctly classify cloud phase for
thin cirrus cases.
The CERESmask was updated to improve cloud detection over strong sunglint
ocean (sunglint probability > 40%) to reduce false clouds. Over sunglint ocean,
the clear sky standard deviation of the brightness temperature difference
between 3.7 µm and 11 µm was changed from a constant number (2.5K)
to 0.0255K*sunglint_probability + 2.45K.
The extinction ratios and single scattering albedos were updated.
SW Model A surface flux algorithm was modified to retrieve the GFDL aerosol
parameters from the correct hemisphere.
SW TOA flux algorithm was updated to apply earth-sun distance correction
to unfiltered SW radiance in the SW flux calculation for snow and the default
SW ADM type was set to 592.
LW model A Surface algorithm module: Effective pressure (used to compute
air temp) now computed based on ocean/land and surface pressure (previously
based on tropics/extra-tropics). New sets of coefficients for Land (no longer
split based on tropics/extra-tropics).
Aerosol correction removed from SW model B surface flux algorithm.
Cloud retrieval model for 1.6µm was changed to higher angle resolutions.
Cloud retrieval models for 3.7µm were changed to higher angle
resolutions.
CERES Mask algorithm was updated. The updates include:
Clear sky night time.
Polar daytime mask.
Polar nighttime mask.
CERES non_polar night time.
The twilight definition was changed for solar zenith angle range from
82.0-87.5 degrees to 82.0-88.5 degrees.
Improved version of MODIS classifier.
New IR cloud phase algorithm.
Updated overlapping algorithm.
Spectral correction algorithm modified to use dynamic Spectral Correction
Coefficients based on instrument gains.
Combined NSDC and NESDIS snow/ice maps used as input.
Aerosol A algorithm has been changed to process each channel independently
(single-channel) using the input tables from the third-generation algorithm.
Clear sky snow and ice reflectance models and albedo models for 0.6µm,
1.6µm and 3.7µm changed to higher angle resolutions.
Clear sky, snow and ice reflectance models and albedo models updated for
0.6µm at TOA, 1.6µm and 3.7µm at surface
The Cloud Phase Algorithm modified for a better cloud detection, especially
for thin cirrus clouds.
A MODIS 3.7µm calibration implemented. This calibration includes
Response-Function-Weighted wavelength and VIRS radiance normalization.
The CERES mask algorithm (including polar mask algorithm) re-tuned based
on the 3.7µm calibration.
The Relative Azimuth was changed from CERES to the imager for the Stowe
aerosol optical thickness calculation.
Over polar regions and over snow-ice surfaces, the NoVis algorithm was
turned on.
Welch MODIS classifier implemented.
Monthly clear sky start-up maps for Terra used.
Additional Vgroups and SDSs containing the MODIS Aerosol parameters are
on the SSF HDF files. There are 14 land and 15 ocean parameters.
TRMM Edition 2B SW, LW, and WN ADMs are used.
SW Model A and SW Model B use GFDL climatology to compute aerosol optical
depth.
LW Model A Air Temperature at 950mb was replaced with air temperature at
effective pressure. The effective pressure was computed based on FOV location
(tropics/extra tropics) and surface pressure. Parameter name remained temp950mb
even though contents changed.
Expanded imager viewing zenith bins in the QC report to support MODIS
angles beyond 50 degrees. This corrects the bogus cloud detection problem found
in Beta1.
SW and LW Model A surface fluxes only computed for FOVs for which clear
coverage exceeds 99.9%
SW TOA flux algorithm updated to apply earth-sun distance correction to
unfiltered SW radiance in the SW flux calculation for snow.
LW model A Surface algorithm module: Effective pressure (used to compute
air temp) now computed based on ocean/land and surface pressure (previously
based on tropics/extra-tropics). New sets of coefficients for Land (no longer
split based on tropics/extra-tropics).
The SW Model B surface flux algorithm no longer uses GFDL climatology to
compute optical depth.
The Model A algorithm for CERES downward/net SW surface fluxes corrected
to include the GFDL aerosol correction and process only for clear-sky.
Aerosol A algorithm has been changed to process each channel independently
(single-channel) using the input tables from the third-generation algorithm.
Aerosol A algorithm processes each channel independently (single-channel)
with updated tables for the spectral responses for 0.66 and 1.6 micrometer
MODIS channels.
Vgroups added to hdf file; Vgroups map to parameter groupings
SSF_Header Vdata updated (Added Beta angle)
Added SSF FOV parameters Column averaged relative humidity, Cloud
property extrapolation over cloudy region, Cloud-mask clear-strong
percent coverage, Cloud-mask clear-weak percent coverage,
Cloud-mask percent coverage supplement, Aerosol A supplements
1-4, Mean water particle radius for cloud layer (1.6), Mean ice
particle effective diameter for cloud layer (1.6), and Mean cloud
particle phase for cloud layer (1.6)
Deleted SSF FOV parameters: Colatitude of CERES FOV at TOA,
Longitude of CERES FOV at TOA, CERES spectral albedo, Shadowed imager
pixels percent coverage, Imager sun glint percent coverage, Imager-based fire
percent coverage, Area percent coverage for cloud layer, and Stddev of
cloud particle phase for cloud layer.
Other SSF FOV parameter updates include: parameter names adjusted, clear
and cloudy percent coverage data types switched to 32-bit real.
MOA input generated from ECMWF data. Using MOA ozone information.
VIRS Release 5 imager data used as input
CRH startup maps replaced (corrects problem over tundra ecosystems)
Surface emissivity maps replaced (now based on ECMWF/AVHRR 1986 data)
CERES calibration adjustment to 1.6 micron VIRS radiances. Most adjustments
to imager radiances included in SSF imager radiance statistics.
Nighttime cloud algorithm added
Polar snow and ice cloud mask added
Potential cloud layer overlap algorithm added
Correction to twilight drop in cloud fraction
New Welch algorithm incorporated
CERES Cloud Mask (improvements in sunglint areas, added shadowed algorithm)
updated. Mask decision tree uses Welch results as needed.
Added code to handle saturation of 3.75 channel
Added reflectance calculation for 3.75 channel
VISST algorithm modified to accommodate new science algorithms. New models
for 3.7 channel ice and water phases. Handles cloud inversions.
Particle size calculation adjusted
Aerosol optical depth calculations use 2x2 pixel uniformity test for channel
1 reflectance and land/water map on 10 minute grid
Keep cloudy FOVs with limited or no cloud property coverage; extrapolate
layer coverages and properties where possible
Single broadband surface albedo and surface emissivity calculation for
FOV (previously calculated for every pixel)
New Slope Intercept Spectral Correction algorithm and coefficients to
unfilter CERES radiances
ADMs updated to VIRS12B
TOA fluxes referenced at surface
Only FOVs which are 99.90r more clear have valid CERES flux values. All
others CERES derived flux values set to CERES default.
HDF SSF format updated - each parameter is a separate SDS
VIRS Release 4 imager data used as input
Note for cloud layer contains strong cloud, weak cloud, and sunglint
from cloud assessments
Notes on cloud algorithms contains strong clear and weak clear
assessments
Visible optical depth no longer logarithmically averaged. Corresponding
parameters now Mean logarithm of visible optical depth for cloud layer
and Stddev of logarithm of visible optical depth for cloud layer
VIRS based direction models used
Correlated K profile derivation improved
VINT phase determination algorithm improved
Cloud mask improvements
CRH startup maps improved for 0.6 albedo, 1.6 albedo, ratio of 1.6 to 0.6
reference, and 10.8 brightness temperature
Low cloud detection improved
VIRS 1.6 micron channel thermal leak correction made before computing
Total aerosol optical depth at 1.6 microns in clear area
Imager-based aerosol percent coverage and associated flag modified
Algorithms for computing Percentiles of visible optical depth for cloud
layer and Percentiles of IR emissivity for cloud layer expanded
VIRS12A Angular Direction Model used to invert radiances to fluxes
Missing microwave precipitable water corrected - microwave values properly
flagged on MOA and copied onto SSF
The sample images for clear area percent coverage depict one
hour, or one granule, of SSF data. The sample images for SW unfiltered radiance
and SW TOA flux depict an area over northeast Africa (Egypt) and illustrate the
CERES crosstrack scan pattern. Two sets of images are provided to illustrate
the similarities and differences between TRMM Edition1 SSF and TRMM Edition2A
SSF. In TRMM Edition1 SSF, only clear footprints have fluxes associated with
them. In the other SSF data sets, TOA fluxes are not limited to clear footprints.
Mail Stop 420
Atmospheric Sciences Research
Building 1250
21 Langley Boulevard
NASA Langley Research Center
Hampton, Virginia 23681-2199
USA
FAX: (757) 864-7996
Author/Data Center:
User and Data Services Office
Atmospheric Science Data Center
NASA Langley Research Center
Mail Stop 157D
2 South Wright Street
Hampton, Virginia 23681-2199
USA
Telephone: (757) 864-8656
FAX: (757) 864-8807
E-mail: larc@eos.nasa.gov URL: http://eosweb.larc.nasa.gov
The requested form of acknowledgment for any publication in which these data
are used is: "These data were obtained from the NASA Langley Research Center
Atmospheric Science Data Center."
The Langley Data Center requests a reprint of any published papers or
reports or a brief description of other uses (e.g., posters, oral presentations,
etc.) of data that we have distributed. This will help the Data Center
determine the use of data distributed, which is helpful in optimizing product
development. It also helps us to keep our product related references current.
To assist the Langley Data Center in providing the best service to the
scientific community, we request a notification if you transmit these data
to other researchers.
Reference:
The CERES Team has gone to considerable trouble to remove
major errors and to verify the quality and accuracy of these data. Please
provide a reference to the following paper when you publish scientific
results with the CERES data:
Wielicki, B. A., B. R. Barkstrom, E. F. Harrison, R. B. Lee III, G.
L. Smith, and J. E. Cooper, "Clouds and the Earth's Radiant Energy System
(CERES): An Earth Observing System Experiment," Bull. Amer. Meteor.
Soc., 77, 853-868, 1996.
Web Page Information:
Document Creation Date: Dec 01, 2000
Last Date Modified: Dec 14, 2000; May 18, 2001; June 4, 2001; July
6, 2001; Sep 10, 2001; Oct 17, 2001; Jan 25, 2002; Aug 2002; Nov 2002; May 1,
2003; Mar 24, 2004; Apr 26, 2004; Mar 3, 2005; Jun 2, 2005; Oct 29, 2007; Oct 21, 2008
Review Date: October 2008
Document ID: Author: User and Data Services, Langley ASDC
ASDC Help Desk: Phone (757) 864-8656;
E-mail larc@eos.nasa.gov URL: http://eosweb.larc.nasa.gov/GUIDE/dataset_documents/cer_ssf.html
