Title: Cape Grim Filter Samples

Contact:

   Barry J. Huebert
   Department of Oceanography
   University of Hawaii
   1000 Pope Road
   Honolulu, HI, USA, 96822
   Telephone: (808) 956-6896
   Facsimile: (808) 956-9225
   E-Mail Address: huebert@okika.soest.hawaii.edu


Data Set Description:

All concentrations are in ppt molar ratio (10E-12 moles analyte/mole air).

As a quality-control check, two filters were being exposed at all times.
A set of criteria, outlined below, was used to choose which of those
numbers would best represent any given time interval.

A. If both MSA and NSS 3 hr avg>+-20% from 6 hr conc, entire 3 hr samples
   were deleted for all species
B. If only one of MSA or NSS >20% from 6 hr conc, only 3 hr data for
   deviating species was deleted
   For any time-period, only 3 or 6 hour data has been saved for each species
C. In cases where the NSS is impacted by heavy sea-salt loading,
   the shorter sample was retained if 3&6 were more than 20% but less than
   40% different.
D. After this culling by MSA and NSS:
   Na, Cl, and SO4 kept for the shortest retained sample
     (cals best for smaller loading)
   NH4+ retained with the NSS sample (although NH4 sensitivity was
     often marginal)
   NO3- retained for longest sample (to maximize sensitivity)

The heavy loading of seaspray made blank subtraction a nightmare;
more so since simultaneous blanks suggest that different filter positions
received different amounts of seasalt. High sodium made ammonium IC analyses
difficult, as well. Although a filter-by-filter error analysis has not yet
been completed, it is reasonable to assume +-20% except near the detection
limits. Two-sigma (for all project blanks) in 6 hr samples gives:

MSA  0.2  ppt  DL
NSS  20   ppt  DL
NH4  5    ppt  DL        Times 2, for 3 hour samples
Na   1000 ppt  DL        Divided by 4, for 24 hour samples
Cl   1300 ppt  DL
NO3  20   ppt  DL
SO4  80   ppt  DL

Note, however, that these are upper limits to uncertainty, because each
sample was in fact corrected using a simultaneous blank. For Na, Cl, and SO4,
the blanks varied widely, as a function of wind-speed (and thus sea-spray),
so the zero-order uncertainty here is large.

So values below the limits above may still be significant, and have been
left in the table.


Format: TAB-delimited ASCII:

SampName
MidTime (decimal day) (UTC)
Jday (DDD) (UTC)
Date (YYYY/MM/DD) (UTC)
Start (HH:MM:SS) (UTC)
Stop (HH:MM:SS) (UTC)
Hrs
MSA
NSS
NH4
Na
Cl
NO3
SO4


File Name Conventions:
   cgfilter.dat

References: None.