These are e-mails regarding LTRANS v.2b that report bugs and solutions. 
Many thanks to the contributors!

**************** FROM Romain Watteaux, 3/2014***************************
*******  Problem with Passive Behavior (Behavior = 0) *****************

Dear Elizabeth,

I actually found what was the problem: it is necessary to initialize
Xbehave,Ybehave and Zbehave to 0.0d0 in Ltrans.f90 instead of in the
subroutine behave, because otherwise for a passive scalar behave=0 and
you don't go into the routine behave, and X,Y,Zbehave take the RAM
value.
At least this fixed the problem on my computer.

Best regards,

Romain Watteaux

romain.watteaux@szn.it a écrit :

> Dear Elizabeth,
>
> I am writing to you concerning your code LTRANS. I am Romain Watteaux,
> a freshly arrived research associate at the Stazione Zoologica Anton
> Dorhn in Naples, and work with Daniele Iudicone on various projects. We
> just started a project on eggs and larvae tracking in the Tyrrhenian
> sea, and decided to use your code to do so.
>
> I seek some help too since I encounter a problem. I found a bug that
> may come from the code (if not from my computer). I donwloaded your
> code from http://northweb.hpl.umces.edu/LTRANS.htm. It runs perfectly
> when I launch it as it is, but when I change in LTrans.data the
> behavior to 0, the code gives NaNs. It seems to be a memory problem
> related to ix or perhaps any parameter next to it in the RAM, because
> when I added "write(*,*) 'ix',ix" right after "call update_particles"
> in Ltrans.f90 (I was doing few checks), the code works and does not
> give NaNs. Any chance you have an idea what the problem is ?

>
> Best regards,
>
> Romain Watteaux 

************************************************************************
************************************************************************



*************** From Elizabeth North, 6/13/12 ******************************

This is not a bug, but a precaution to users.

In the LTRANS.data file, if you have SphericalProjection set to '.TRUE.' (I would recommend this), then 
it is important to adjust the following variables to match the hydrodynamic model domain:

  latmin = 36               ! Minimum longitude value, only used if SphericalProjection is .TRUE.
  lonmin = -77              ! Minimum  latitude value, only used if SphericalProjection is .TRUE.

To do this, find the minimum longitude and minimum latitude in the entire model domain, the set latmin and 
lonmin to the nearest whole number less that the minimum latitude and longitude (i.e., if the minimum longitude 
in the hydrodynamic model domain is -75.666, then set lonmin to -76).

If your numbers for latmin/lonmin in your LTRANS.data file don't correspond to your model domain, 
your particles could go the wrong way.

******************************************************************************
************************************************************************



********************* From Annalisa Bracco, 6/23/15*****************************

Problem: We were "able to successfully run the test case, but when we turn the behavior parameter
to zero, then the execution of the test hung at stepf 5 despite all attempts to change 
other parameter settings (no apparent error. The code simply does not move forward)."

Solution:
"The problem is that the update due to behavior is done for all Behavior values. The 
solution (and yes, happy to share) is to change line 1153

!Update due to Behavior
            newZpos = NewZpos + ZBehav
  into

!Update due to Behavior
      if(Behavior.NE.0)then
          newZpos = NewZpos + ZBehav
      endif


Annalisa Bracco, Professor
School of Earth and Atmospheric Science
Georgia Institute of Technology"

************************************************************************
************************************************************************