The two attached graphs update LBNL's view of V-K2 and HOT.
The revised plot VK2_HOT_MULVFS_GRN_OD.tif reflects changes to the data 
only for HOT MULVFS cast 4. The new results change very little my description of
differences of light absorbing particles at HOT but now use exactly the same imaging
methodology as used at sea
I've attached a first view of transmissometer derived POC. The profile
systematics from intensive observations during cycles 1 and 2 show the
CTD casts 17:24 and 62:68 are chosen as represeting cycles 1 and 2.
Cycle 1: upper 50 m: lows of 2.5 to highs of 4.5 uM POC
          100-250 m: ~0.6 to 0.7 uM POC
          >500:      ~0.35 to 0.4 uM POC
Cycle 2: upper 50 m: lows of 1.6 to highs ~ 2.1 uM POC
          100-250 m: ~0.6 uM POC
          >500:      ~0.35 to 0.4 uM POC
Deep water POC values are probably accurate to ~0.1 uM POC. See note below for computation
of POC levels.
In surface waters significant variability is due to a diurnal cycle in
POC. In absolute terms there was a 40% to 55% decrease of POC in the upper 50 m between 
observation cycles. There was a 10% decrease of POC in waters between 100 and about 250 m.
These changes follow the trends seen in MULVFS Optical density profiles.
A preliminary analysis of CTD temperature variability (attached) shows
that there was significant watermass interleaving in the 80 to 500 m depth interval. 
I believe that changes in surface and intermediate water masses may contribute significantly
to the differences in particle abundances observed.
Comparisons of up and down trace temperature data reveal internal
wave amplitudes at depths of ~30m as follows:
     NOTE CAST 20 T AND S BAD due to pump failure
   CTD cast   up/down displacement of isotherms (m)
      6        8m
     10        5m
     11       12m
     13        5m
     14       22m **** this looks very real to me!
     17        7m
     25        4m
     27        6m
     31        6m
     33        5m
     37       note mixed layer T clearly different up vs down
     43        8m
     48        6m
     52,53,54  4m
     59        8m
     62        5m
     67,68     6m
     72       10m
     73        6m
     86       10m
     All other CTD casts showed little up/down cast difference
Implications of this are that internal wave induced mixing must be at
times a significant factor in maintaining euphotic zone productivity and possibly export. 
I think Erik and Dave might have already thought about this.
Note on POC estimates: 
Conversion of transmissometer voltages to beam attenuation coefficient was performed using 
standard methods using factory calibration
(Vair,Vref,Vzero) for the sensor, adjustment of Vref using Vair and Vzero readings of the 
instrument as installed on the CTD (prior to use aboard ship). Drift in sensitivity due to
light source aging was determined by Air and Zero calibrations on the CTD after cast 86. The
drift behaviour was verified using data from 3000 m casts (27 and 72). linear downwards drift
in sensitivity during CTD operation occurred at a rate of 0.0033 V/day over the 4 days of
cumulative CTD ops. 
Conversion of beam attenuation coefficient to POC followed Bishop et al. (1999) with 
compensation for differences in transmissometer acceptance angle as described in Bishop et al. 2004.
      POC = cp / 0.062 * 1.3 uM
The 0.062 comes from Bishop et al. 1999; the 1.3 is the adjustment due to the factor of three 
difference in transmissometer receiver acceptance angle (1.5 - SeaStar vs 0.5 degrees : 1-m seatech).
Basically, the wider acceptance angle allows more forward scattered light to be detected,
thus compensating for photons lost from the beam due to absorption and scattering. This effectively
decreases the sensitivity of the CStar instrument used during VERTIGO by ~30%. See Bishop et al. 2004.
Jim Bishop
Earth Sciences Division,
Lawrence Berkeley National Laboratory,
One Cyclotron Road, M/S 90-1116, 
Berkeley, CA  94720, USA
tel:  (510) 495 2457
fax:  (510) 486 5686