Andrew Winkelman
 NOAA/GLERL
 2205 Commonwealth Blvd.
 Ann Arbor, MI  48105-2945
 Phone:    (734) 741-2336
 Fax:      (734) 741-2055
 e-mail:   andy.winkelman@noaa.gov

     On each transect, stations were made at water depths of 15, 30, and 45m. At 
each station, a CTD profile was taken and water samples collected at 3,4, and 5 depths 
respectively. Each station had a sample from approximately 1m below the surface and 1m 
above the bottom. The latter was collected using a Niskin equipped with a tripping 
mechanism that closed the bottle when the weight touched bottom. Water samples were 
collected in 8 liter Niskin bottles that were cleaned with soap and water prior to the 
beginning of the cruise. Subsequently, the bottles were flushed with lake water while on 
the wire by moving them up and down prior to closing. 

     Particle size analysis was measured on the suspended matter using a Spectrex 
Laser Particle Counter (Model ILI 1000, Spectrex Corp, Redwood, CA). This instrument 
had several shortcomings, but was available and was used to acquire very rare data on the 
pilot cruise. Water samples were collected directly from the Niskin bottles into Spectrex 
glass sample bottles. The instrument uses a He-Ne laser (632.8 nm) to estimates particle 
concentration and calculates size distribution over the range of 1 to 100um from near-
forward scattered light. The instrument is calibrated by regularly analyzing a set of 3 
sealed bottles with a known size distribution of styrene spheres supplied by Spectrex. 
Background and offsets are adjusted to assure that the readings are within +/-15% of the 
values provided by the manufacturer. During the cruise, the 5um standard read high by an 
average of 35%, outside the range recommended by the manufacturer. Thus the small 
particle size counts should be considered high. Background counts were subtracted and % 
mass was calculated for each bin.

     A serious problem with this instrument is its restriction to less than 1000 
particles per ml; greater concentrations result in coincidence counting. Based on the 
ambient concentrations encountered, all samples had to be diluted with distilled, filtered 
water from between 20:1 to 150:1 (DW:Lake Water).  This results in a bias against rare, 
large particles. 

     There is very little particle size data available for the suspended matter of the 
Great Lakes. However, because of the limitations of the instrument, users should 
consider the poor precision and bias when evaluating the results. Samples were analyzed 
from each station and depth sampled on the cruise.  Samples were shaken prior to 
analysis.  However, experiments prior to the cruise clearly showed that the size spectrum 
of lake samples begins to drift into larger aggregates approximately 4 hours after 
collection. The frequency of water sample collection restricted replication to 3 separate 
scans for each sample, in order to complete analyses in less than four hours. A total of 48 
samples were collected and analyzed on the cruise and the average coefficient of variation 
for all size categories was approximately equal at 104%.