NUMBER 52

DATE: 12/04/87

Sector Processing Center (SPC):

During October, ESA and JMA collectively processed greater than 93% of A data into B1 and/or B2 data. AES processed 70% of the A data into B1/B2. No processing matrix was received from NOAA for October; the processing matrix for September was received.

Normal operations were reported by ESA, JMA and NOA.

AES reported complete loss of data on 28 October caused by equipment malfunction. Other data losses were mostly caused by incomplete transmission from the satellite or from the ground-reception system.

No report was received from CSU.

Special Area Processing Center (SAPC):

No reports were received.

Satellite Calibration Center (SCC):

BC results for GOES-6 were received from the SCC for July 1984 and for January 1987. Results for GOES-6 (March 1987, last of the old format data), GMS-3 (April 1987) and METEOSAT-2 (July 1987) are being processed.

Good cloud targets have been very difficult to find in January 1985 and 1986 for GOES-6 (indeed a recent look by the GPC at AVHRR data for Decembers from 1983, 1984 and 1985 shows that a major "clearing" of cloudiness in this area occurred over this period); however, some new AC sectors have been found by the SCC using their photographic archive. These new AC sectors have been requested from UWS and NOA.

The change of GOES transmission format caused the collection rate of AC data to be very low from April through June 1987, so the SCC is working on March and July BC values. AC data covering the period from April onward have to be reformatted into the old CSU-AC format; these data have not yet been delivered.

The new AC sector selection procedure for GOES-6, agreed to at WGDM 5 in Paris, has not yet been implemented by CSU. In this procedure, CSU was to provide coordinates for AC sectors, based on image inspections, to the SCC in time for the selection of NOAA orbits and provide these selected sectors for GOES-6, rather than collecting fixed sectors. The SCC has, instead, attempted to select better sectors by inspection of the GAC mosaics received on the GOES-WEFAX, but this does not appear to be a very reliable procedure.

AC data from AES have been received regularly, except for September 1987. Adaptation of the processing software to accommodate their format will begin early in 1988.

The corrected GMS AC data were received from JMA; however, the images appear to have a spurious "stripping" that prevents their use for normalization. Discussions with JMA are underway to determine the problem.

A recent paper by Frouin and Gautier (1987, Int. J. Remote Sens., 22, 73-101), proposes calibrations for GOES-5, GOES-6 and NOAA-7. These results give a NOAA-7 Channel 1 calibration about 10% higher than the nominal adopted as the ISCCP reference and GOES calibrations about 25-30% lower than the nominal adopted by ISCCP. The relative relations among the GOES and NOAA calibrations appear consistent with the normalizations obtained by the SCC and confirm the estimated absolute uncertainty of the ISCCP standard. (More recent results by Whitlock, comparing Gautier's results with the NOAA aircraft method, show good correspondence btween the two results when obtained at the same time, but suggest that the variability obtained by Gautier is caused by changes of the White Sands surface albedo. Intercomparisons of all these results will be conducted early next year.)

Global Processing Center (GPC):

The GPC continued to receive B1 and/or B2 data from JMA, ESA, CSU, AES and NOA and correlative data from NOAA/NESDIS in a routine manner.

NOAA-9 B3 production has been completed through February 1986; chronic tape drive difficulties have slowed the pace of B3 data production, however. The calibration monitor results must be inspected before delivery. There may be a small change in the Channel 1 calibration over the first year; however, part of this effect may also be caused by a systematic difference in the snow cover in the two winters (January 1985 and 1986). An early look at the monthly mean radiance map for the whole Earth suggests that the snow cover may be different and also showed a significant change in the "amount" of subtropical cloud cover over the oceans. We are examining the NOAA snow data to confirm the apparent change in snow cover. There does appear to be a very small change in the average relfectance of the Sahara, but we need to examine these results more carefully to understand whether a calibration change has actually occurred.

With the receipt of GOES -6 BC data for January, April and July 1984, the B3 data will be re-copied next week to add the BC information. The GOES-6 B3 for November 1983 through October 1984 will then be delivered to the ICA. Only the BC data from January 1985 and 1986 remain unavailable.

Production of the GMS B3 data will begin next week and should be completed through January 1985 before the end of December.

Remapping of the nothern hemisphere snow data for 1983 and 1984 has been completed; the snow data for the southern hemisphere have not been obtained as yet, but are expected to arrive this month.

Comparisons between the TOVS climatology and that of Oort show some differences that are probably due to specific variations in 1983; however, these differences are large enough (2-6 K) that we will use the monthly-average TOVS data to fill as many holes in daily data as possible and the Oort climatology to fill only those holes caused by the complete lack of TOVS data for a whole month. Before averaging, we will extend the replication of the original observations to two grid cells distance (about 500 km), in the east-west direction only. Error tests indicate that the errors are still only about 3 K. These changes in strategy will delay the beginning of TOVS processing; however, first delivery should still occur in January, once a tape documentation is written.

Test runs of the revised cloud algorithm are underway to re-tune the performance to be consistent with the earlier version. Refinements in the use of topography and snow/ice information are also being tested. The pace of this testing is slowed by a continuing problem of tape drive reliability; despite repeated "repairs", it still requires multiple submissions of a tape job to get an error-free read or write completed. We estimate that two, or possibly three, weeks of testing remain, covering, specifically, refined sun-glint model, improved clear IR "spike" detection using topography, minor changes in clear-sky logic, and additional use of snow/ice information.

Final tests of the radiation code have begun and should be completed within one to two weeks. These include tests of the ocean surface VIS relfectance model, revised procedures for calculations involving small values of cloud optical thickness, revised treatment of the TOVS temperature profiles, and further tests of the adjustment of thin cloud top pressures to account for transmission of IR radiation from the surface.

Changes in the diagnostic step that produces the C1 data will begin next week.

Development of the cloud data production system has continued, allowing us to start production with two complete systems in place, instead of one. This will increase the intital pace of C data production. Although the first production will be cautious, we should be able to begin producing two months of C data every month almost from the beginning.

ISCCP Central Archive (ICA):

The ICA continued to receive B1 data from JMA, CSU, ESA, AES and NOA in a routine manner.