Memorandum to: 	All ACE participants

From:	       	Darrel Baumgardner, NCAR/RAF

Subject:	Composite particle probe spectra


A new set of data files are now available that contain a composite
size distribution synthesized from measurements with the C-130
PCASP, FSSP-300, FSSP-100 and 260-X. This composite size
distribution represents a reprocessing of the FSSP-300, FSSP-100
and 260-X measurements. The changes from the netCDF data files that
are presently on the NCAR MASS store are as follows:

1) The composite distribution uses only portions of each
   probe's measurements and takes advantage of the overlapping
   size ranges of each of the probes. The nominal size ranges
   of these four probes are:

   PCASP     -  0.1 to 3.0 um.
   FSSP-300  -  0.3 to 20.0 um.
   FSSP-100  -  2.0 to 47 um
   260-X     - 10.0 to 620 um.

   Normally, the first channel of each of these probes is suspect
   because of elecrtonic noise or sunlight contamination, so the
   overlapping provides a means of covering these channels with
   another instrument, with the exception of the PCASP, where the
   first channel is not included in the processing at all. The composite
   distribution is made up as follows:

   PCASP     -  0.161 to .303 um
   FSSP-300  -  0.415 to 4.1 um
   FSSP-100  -  5.8 to 49  um
   260-X     - 45.0 to 605 um.

2) The FSSP-300 calibrated to match a particle refractive index of 1.40,
   instead of the factory calibration at 1.58. Some channels were
   combined to minimize the problems of Mie sizing ambiguities.

3) A better calibration was applied to the FSSP-100 and the last three
   channels were combined to take care of Mie scattering ambiguities
   in that size range.

4) The OAP was recalibrated and found to have a size resolution of 17 um
   rather than 10 um. There was a change made in the field to increase
   signal to noise ratio and the resolution was changed at that time. This
   affects not only the sizing, but also the sample volume.
  
   
   The data files are presently stored in the anonymous ftp directory
pub/darrel/rfnn.spec, where nn is the flight number. These files
have been compressed using the gnu toolkit utility gzip. To
obtain these files, ftp to chinook.atd.ucar.edu.

   These files will remain in this ftp directory until they are
transferred to the NCAR MassStore at which time I will notify
ACE participants of the transfer. I am also sending these data
files to OFPS and I assume they will make the data available
through CODIAC. While we are deciding their final resting place,
however, the best place to retrieve the data is in the ftp
directory mentioned above.

   The size distributions are ten second averages throughout the flights.
I can make one second data available upon request. The 43 channels
of data are in units of cm-3. In the header of each file I have
included the midpoint and width of each channel. The concentrations
should be divided by the width when plotting size distributions.
The average channel diameter can be used when calculating volumes
and surface areas from the size distributions.

   There are several caveats. I have not yet done an extensive
quality assurance check on these files. I'll be doing so over time
but at this point caution is warranted. There could be occasional
spikes and noise that will contaminate the spectra. Also, the PCASP
had a leak in its plumbing until flight 11, so I have zeroed its
values in flights 4-10. I will appreciate any feedback regarding
problems that may be detected in the data files.