This is the readme file for the NOAA Climet OPC processed data for 
Kosan during the ACE-Asia experiment, March-April 2001.  

Measurements were made at Gosan (33.292 N, 126.162 E, Time = UTC + 9 
hrs, Elevation = 50 m) during the period of April 5 to April 27, 2001 
during the ACE-Asia intensive period.  The only day with no data at 
all is April 12, although there are a number of days with very limited 
amounts of data (4/17 to 4/20)

Contact person:
Patrick Chuang
pchuang@es.ucsc.edu
831.459.1501 (w)

The NOAA Climet OPC is a white light counter, and therefore we expect it 
to be less sensitive to changes in refractive index versus single 
wavelength OPCs.  The size range that we feel it can reliably measure is 
0.5 to 8.7 microns in 9 nearly equally spaced bins. The specific values 
are shown in the below table.

Column 1:  The channel number (1 through 9).
Column 2:  The geometric mean bin diameters = square root of (upper bin 
           boundary diameter*lower bin boundary diameter), in micrometers.
Column 3:  The geometric bin width = (Upper bin boundary diameter/lower 
           bin boundary diameter).

1	0.50	1.60
2	0.77	1.50
3	1.12	1.40
4	1.57	1.40
5	2.20	1.40
6	3.07	1.40
7	4.30	1.40
8	6.02	1.40
9	8.72	1.50

In reality, the OPC actually measures 64 channels worth of data (although 
the first 8 exhibit too much instrumental noise to be worth anything), 
which we then bin into the 9 channels shown above.  Nine channels were 
chosen because it reflects a reasonable bin width compared to the transfer 
function of the instrument (i.e., particles of size 1.12 microns are very 
unlikely to be measured as belonging in either bin 2 or bin 4).  

IMPORTANT NOTE:  The reported concentrations are actually dN, in units of 
cm-3, rather than dN/dlogDp.  To obtain the latter from the reported 
values, divide the reported dN values by the logarithm of the geometric 
bin width (column 3).  

---

The uncertainty in the concentration of each bin is reported.  A number 
of different error sources are considered:

1.  The width of the OPC transfer function.  Based on calibrations of the 
    OPC using PSL and glass bead aerosols, the maximum broadening of a 
    monodisperse aerosol was measured.  This is then used in calculating 
    the uncertainty in counting for any particular bin.

2.  Counting statistics.  Counting errors based on a low sample size can 
    be expressed (assuming Poisson statistics) as square root(counts).  
    This is also factored into the reported uncertainty.

3.  Flowrate uncertainty.  This is assumed to be 10% of the total flowrate. 

By correctly propagating these uncertainties, a concentration uncertainty 
can be reported.  These can be considered the one (1) sigma uncertainty in 
concentration.

---

The format of the data files, named opc_spec_day???.txt is as follows:

Column 1:  Year

Column 2:  Day of Year (1 to 365)

Column 3:  Hour (1 to 24)

Column 4:  Minute (0 to 59)

Column 5:  Second (0 to 59)

Column 6:  Decimal day (1 to 265.999).  

Column 7:  Input file name (not really useful information for anyone but 
           myself)

Column 8:  Integrated OPC measured concentration in units cm-3.  That is, 
           measured concentration of all particles in the OPC size range 
           (no uncertainty for this value is presently reported). 

Column 9 through 17:  Concentration for channels 1 through 9 in units cm-3, 
                      respectively, i.e. dN(1), dN(2),..., dN(9) where the 
                      values in () represents the channel number.  NOTE:  
                      values are reported as dN, and can be readily 
                      converted to dN/dlogDp if desired as described above.

Column 18 through 26:  One sigma uncertainty in concentration in units cm-3 
                       for channels 1 through 9, respectively, i.e. 
                       sigma(dN(1)),..., sigma(dN(9)).

Times are reported as UTC, which is Korean Standard Time minus 9 hours.  
The data are averaged to roughly 1 minute (though not exactly).  
There are lots of data gaps because of OPC computer crashes, and overheating.  

---

Patrick Chuang
April 8, 2003