NUMBER 133

DATE: 08/05/94

Sector Processing Center (SPC):

During June, NOA, ESA, JMA, and CSU processed greater than 96% of A data into B1/B2 data.

Normal operations for June were reported by NOA, ESA, JMA and CSU. B2 data for April 94 were received from AES, but no B2 data or report was received for June 94.

Special Area Processing Center (SAPC):

No report received.

Satellite Calibration Center (SCC):

No report received.

Global Processing Center (GPC):

The GPC continued to receive B2 data, BC data and correlative data in a routine manner.

Production of NOAA-11 (preliminary) B3 data is current. New BT (calibration) data have been produced for the "afternoon" polar orbiters: NOAA-7 (calibration has been completed back to September 1981), NOAA-9 and NOAA-11.

Production of NOAA-12 (preliminary) B3 data is current. New BT data have been produced for the "morning" polar orbiters: NOAA-8 and NOAA-10. BT data for NOAA-12 will be produced as the final B3 data are produced.

Production of GOES-7 (preliminary) B3 data is current. The processing of the UWS replacement B2 data continues.

Production of METEOSAT-3 (preliminary) B3 data is current. The processing of B3 data for May 92 through October 92 continues.

Production of METEOSAT-5 (preliminary) B3 data has commenced and is now current. New BT (calibration) data have been produced for the period before July 91 including METEOSAT-2, METEOSAT-3 (at 0 longitude), and part of the record of METEOSAT-4. Final B3 data for METEOSAT-4 for July - December 91 have also been completed.

Production of GMS-4 (preliminary) B3 data is current. New B3 data for GMS-4 for July - December 91 have been completed.

Production of all INSAT-1 (preliminary) B3 data has been completed.

Production of the ice/snow and TOVS correlative datasets is current. The availability of the sea ice data for 93 is still in doubt.

Investigation and testing of the new ice microphysics radiative model have been completed. This was an interesting problem. The difficulty with the subroutine received from P. Minnis occurs at scattering geometries where the observed cloud reflectivity is larger than the asymptotic limit for an ice cloud but not for a liquid water cloud. Thus, for some of the largest optical thickness values obtained using the liquid water microphysics, there is no corresponding ice cloud solution. The code supplied by P. Minnis simply extrapolated to find a match, producing very large (> 1000) optical thickness values. The "surprise" is that the differences between the liquid water and ice cloud optical thicknesses required to produce the same reflectivity are quite large, even at the largest optical thicknesses. We even produced our own calculations using a slightly different ice crystal phase function from that used by Minnis, but the differences between these two results are insignificant. In fact, as the asymptotic limit for both microphysics is approached, the ratio of ice to liquid optical thickness is approximately 0.55. So, even though other "intuitive" facts are correct -- that the difference in cloud albedos at very high optical thicknesses is very small and the angular dependence of the radiation is very small, the optical thickness difference remains large.

These undefined situations will be handled by employing the ratio of ice to liquid water optical thicknesses at the asymptotic limit, which is justified by three observations: (1) the situation occurs mostly for clouds with large optical thicknesses (> 25) and cold tops (< 260K), (2) the angular dependence of the reflected visible radiation decreases for larger optical thicknesses, so that use of a ratio determined from albedo rather than reflectance is not too much in error, and (3) the error in the cloud albedo at optical thicknesses > 100, corresponding to an optical thickness ratio of 0.55, is actually small (< 0.03). In any case, the uncertainty in the optical thickness is probably already very large for what are probably deep convective clouds, composed of both liquid and ice phases with broad, probably multi-modal particle size distributions and exhibiting large deviations from plane-parallel structure. Overall, use of the ice microphysics model for colder topped clouds will reduce the optical thicknesses retrieved. One important consequence of this change is that the amounts of deep convective and cirrus clouds identified by the current criteria will decrease and increase by 30-40%, respectively. However, there was some evidence that the original amounts were a little high and low, respectively.

Re-processing will commence next week: first, we will process specific months of data that have been requested by various other projects (Surface Radiation Budget, FIRE, ICE, TOGA-COARE). Then we will begin full re-processing of the data before July 91, as well as production of new data beyond July 91.

ISCCP Central Archive (ICA):

The ICA continued to receive B1 data from NOA, AES, ESA, CSU and JMA and B3 and C data from the GPC in a routine manner.

The health of GMS-4, METEOSAT-3 (at 75 W), METEOSAT-5, GOES-7, NOAA-11 and NOAA-12 remained good. Check out of the new GOES-8 continues with very few problems being encountered. The launch of NOAA-J (to replace NOAA-11) may be delayed further by unexplained noise in some HIRS channels.

OTHER

The third meeting of the Working Group on Data Management for Radiation Projects of the WCRP was held in Budapest, hosted by G. Major and R. Roth who hosted the first ISCCP planning meeting in 1980. Although all centers reported essentially routine data collection and processing activities, it is interesting to note that, for a project that has lasted so long, many institutional changes and evolutions are underway. One technical consequence was much discussion of ways to modernize the exchanges of data, either by changing to newer media or, for the smaller datasets, beginning to use Internet for exchanges. As the satellite systems evolve, new spectral channels are added which make the B1 and B2 datasets significantly larger. All SPCs agreed to add these new channels as they become available and to explore using denser or more convenient media with the centers that they interact with. In particular, an investigation will be made to see whether most of the exchange of AC and BC datasets can be conducted over Internet.

The main topic of discussion concerned possible changes being studied for the second phase of ISCCP as part of GEWEX -- it was decided to call this the second phase of ISCCP rather than referring to it as ISCCP-2. Of most concern were proposals to increase the spatial resolution of the products to 10 km (B1 resolution) and the time resolution to one hour. The former change would not impact the SPCs (they would just change the mailing address for B1 data), but would require a 9-fold increase in the processing at the GPC that does not seem feasible (the GPC processing is already I/O bound). The latter change would significantly impact the SPCs, as well as increasing the GPC processing by another factor of 3. Combined with the advent of new spectral channels that already will increase the size of B1 and B2 datasets by about a factor of 2, a change to B1 data at one hour frequency amounts to a 50-fold increase in the dataset sizes. Other proposed changes to ISCCP involve datasets that would be obtained from a single source, either NASA or NOAA, and, for the most part, involve only a modest increase of processing at the GPC. Some ideas are more ambitious, so that no decisions were made until the scientific studies are completed.

Also reviewed was the status of other projects, mainly the Surface Radiation Budget (SRB) project and the Baseline Radiation Network. Both of these appear to be developing well and can be expected to begin delivery of complete datasets either later this year or early next year. SRB has already issued a four-year downwelling shortwave flux dataset.

Stage B3: July 83 - June 91

Stage CD: July 83 - December 91

Stage C1: July 83 - June 91

Stage C2: July 83 - June 91

Delivering sea ice correlative data for 93 to GPC.

Delivery backlog for B3 data = 30 months (with respect to planned schedule). B3 data for 96 months have been archived.

Delivery backlog for C1/C2 data = 30 months (with respect to planned schedule). C1 data for 96 months have been archived.