Update below, but first a note to the Endeavor team:

Please select any or all of the following files. We will
email them as soon as we get the ok from you. 

-ICS_X.s3r files from EL9905 bongo grid 11-14 May 
-ICS_X.s3r files from EN323 VPR grid 6-8 May "grid2"
-ICS_X.s3r files from EN323 VPR grid 8-? May "grid3"
-EL9905a4.m3d from EL9905 ADCP update 
      (let us know what your last el9905 adcp obs time was) 

Note the above ICS files (approx 100Kbytes each) are now embedded 
in OC341 broadscale but we will incorporate AL99XX broadscale as soon as 
they finish sampling the sflank next week.

-JiM.

-----------------------------------------------------

Hello all,

  We just completed the case examining the sensitivity
to the value of ekmin.  We changed the value from 
2e-3 to 2e-5 and kept all other parameters as in
Case 3 (May18_FCAST2) described in this morning's message.

The mean separation between 4 model and observed drifters
is now 1.38km!! -- our best result in this series of expts.  

Here's a summary table:

Case  name                                Mean Separation
------------------                                -------------------------
May17_FCAST                              4.65 kms
May17_FCAST2                            3.80 kms
May18_FCAST                              3.70 kms
May18_FCAST2                            2.56 kms
May18_FCAST3                            1.38 kms

The model fields displayed in the defplots page do not
display features that cause concern but we will continue
to dissect the solutions to make sure we haven't overlooked
anything.  The solution still does not hold the stratification
in the top 5-10 m of the water column and we will continue
to explore this.

Comments from the shore-based and Endeavor team on the 
reduction of ekmin would be greatly appreciated.  We are 
not quite ready to embrace the 2e-5 value as the new value for 
the remaining runs based on just one run.

Cheers,
Cisco and Jim.