Effective immediately, a new source of cloud data is
available--the satellite-derived cloud cover product (AFOS
category SCP). NOTE: CURRENTLY, THIS PRODUCT IS ONLY
AVAILABLE
INTERNALLY TO NWS USERS, VIA THE AFOS NETWORK.
DISTRIBUTION TO
EXTERNAL USERS IS SCHEDULED TO BEGIN BY THE END OF THE
YEAR.
The National Environmental Satellite, Data, and Information
Service (NESDIS) and the Cooperative Institute for Meteorological
Satellite Studies at the University of Wisconsin have recently
adapted a technique called "carbon dioxide slicing" which uses
data from GOES-7 to estimate cloud top heights, opacity, and
total sky cover. Infrared radiation measurements at several
frequencies are used. At three of these channels, carbon dioxide
is an effective absorber and re-emitter of electromagnetic
radiation. Since each of these channels has a peak absorption
response at a different level (850 mb, 500 mb, and 200 mb), this
cloud estimation technique is effective with clouds at various
levels. The technique is most effective with middle- and
high-level clouds. To identify whether each field-of-view (FOV)
is clear or cloudy, infrared data from a transparent "window"
channel are also used. Radiative transfer principles are then
used to estimate the cloud parameters from the measured radiance
data in combination with NMC numerical model output.
These satellite-derived cloud estimates, which are available in
the satellite cloud product, are used to complement
ASOS observations with cloud data above 631 mb (approximately
12,000 feet MSL). Since ASOS measures cloud BASES below 12,000
feet AGL, and the carbon dioxide slicing technique is less
effective at estimating low clouds, only cloud top estimates
above 631 mb are transmitted. Locations for which the data are
provided are selected from a list of ASOS sites. Data are
usually available hourly for most of the United States; however,
competing requirements for the instrument cause data in certain
locations to be late or unavailable. Due to the oblique viewing
angle that currently exists from the present position of the
GOES-7 (over the equator at 112 deg West longitude) relative to
the northeast United States, satellite-derived cloud data will
not be available over the northeast United States until the
GOES-I-J system becomes operational in mid-1995. Also, there are
no data using this technique for Alaska nor Hawaii.
The AFOS PILs and WMO headers of the satellite-derive cloud cover
products (in NWS regional collectives) are listed below:
AFOS PIL WMO Header NWS Region
cccSCPER1 TBUS20 KWBC Eastern
cccSCPCR1 TBUS21 KWBC Central
cccSCPSR1 TBUS22 KWBC Southern
cccSCPWR1 TBUS23 KWBC Western
The strengths of the SCP are:
1. It complements ASOS with cloud data above 631 mb.
2. Data are generally available hourly.
3. After the GOES-I-J system becomes operational, data will
be available across the contiguous United States.
4. The average Effective Cloud Amount (ECA) values in the
report can assist in estimating cloud opacity.
The weaknesses of the SCP are:
1. Currently, with only one GOES, cloud estimates are not
available over the northeast United States.
2. The technique is generally capable of detecting only the
highest cloud layer. Individual tops of multi-layered
clouds aren't resolved. In the case two layers of clouds,
with a
top layer of thin clouds, the reported cloud top
height will likely be a level between the two layers.
3. Sometimes,
very thin clouds (usually cirrus) go undetected
or are misidentified as low clouds. Hence, clouds with low
tops (at or below 631 mb) are not reported. Whenever there
is sufficient doubt about whether the radiances above a
station represent low clouds or higher thin clouds, the sky
cover for that station is designated MCLR.
4. Whenever thin clouds are present (especially when MCLR
[mostly clear] is reported - see Part 2 for an explanation
of MCLR), the cloud top height estimates should be used
judiciously.
5. A strong surface-based inversion (i.e., during an Arctic
air mass across the northern tier) may cause the satellite
to report middle clouds when skies are actually clear.
6. Horizontal resolution averages 10 km in the mid-latitudes.
Hence, the technique tends not to resolve small cloud
masses and small cloud breaks.
Note: One technical problem has not yet been corrected. It
pertains to the reports at ACY and Y62. At these locations the
CLD TOP values are often mistakenly printed as "zero." This
software "bug" is in the process of being corrected.
A sample SCP message and a line-by-line explanation follows in
part 2 of this message, which follows immediately.
A Technical Procedures Bulletin on this new product is in
preparation. If you have any questions, call Chuck Kluepfel at
the Office of Meteorology at (301) 713-1867.
END OF PART 1 - PART 2 FOLLOWS
TO: ALL NWS FIELD OFFICES
FROM: MARY M. GLACKIN
CHIEF, SERVICES DEVELOPMENT BRANCH
SUBJECT: THE NEW SATELLITE-DERIVED CLOUD COVER
PRODUCT -
SAMPLE MESSAGE AND KEY TO READING PRODUCT
(THIS IS PART 2 OF A 2 PART MESSAGE...PART 1 WAS JUST SENT)
A sample of the satellite-derived cloud cover product (SCP) is
given below:
.................................................................
NMCSCPER1 (stored in AFOS database under your local node id)
TBUS20 KWBC 241950
SATELLITE-DERIVED CLOUD INFORMATION FOR MID (CLD TOPS 631-
400 MB)
AND HIGH LEVEL (CLD TOPS ABOVE 400 MB)
(JUN)
STA DA/TIMEZ MID HIGH CLD TOP ECA
ACY 24/1923 CLR 0
IAD 24/1924 MCLR 190-230 5
ORF 24/1925 BKN 250-300 44
PIT 24/1923 SCT 120-140 13
. . .
. . .
ROA 24/1925 OVC 330-400 90
.................................................................
A line by line description of the SCP follows:
The first line of the product gives the AFOS header. The second
line gives the WMO header and the date/time (UTC) group of
product issuance. The third and fourth lines give the title of
the product. The fifth line gives the month. Line 6 contains
the column headers, while lines 7 through the end depict the
following data for each station in the collective from left to
right under the appropriate column headers:
1. Designated ground observing station for which the data are
calculated - usually a 3 letter identifier.
2. Date and time of observation (UTC).
3. Cloud coverage (CLR, SCT, BKN, OVC) at the MID levels,
defined as 631 to 400 mb (approximately 12,000 to 23,600 feet
MSL) in a standard atmosphere. Mostly clear (MCLR) is used
when very thin, transparent clouds may be present above
631 mb, but there is insufficient data to accurately resolve
either total cloud coverage or cloud top heights. It should
be noted that MCLR is a new category, which is currently
being added to the software. It is a permanent change, but
has not yet taken effect. MCLR is expected to begin
appearing in the reports late in 1993.
4. Cloud coverage at the HIGH levels (defined as above 400 mb).
Note: A cloud coverage of BKN or OVC will be given for
EITHER the MID or HIGH levels. SCT layers may be reported at
BOTH MID and HIGH levels. A CLR (MCLR) report indicates
clear (mostly clear) skies above 631 mb and implies nothing
about cloud coverage at or below 631 mb.
5. Range of cloud top heights, given in hundreds of feet MSL,
from all fields-of-view (FOVs) within approximately 25 to
35 km (40 to 56 miles) of the station. The lower limit is
631 mb, and the upper limit is set to the first guess of the
tropopause height from the NGM. These values are unreliable
whenever the sky cover report is MCLR.
6. Average Effective Cloud Amount (ECA) - given in percent.
Average ECA, a measure of cloud cover and cloud opacity, is
an average of up to 25 ECAs from different FOVs around the
station. The average ECA for clear skies is zero, while the
average ECA for an opaque overcast is 100. An average ECA of
less than 33 (66) in the case of broken (overcast) cloud
conditions indicates that the cloud layer is likely to be
thin, while a value of 33 (66) or greater indicates the
likelihood of an opaque cloud layer. For example, an average
ECA of 55 with a broken cloud cover indicates an opaque
broken deck. That same average ECA value (55) with an
overcast cloud cover indicates thin overcast conditions.
If you have any questions, call Chuck Kluepfel at the Office of
Meteorology at (301) 713-1867.
END OF PART 2 OF 2