Movement parameters in Osmose Version 3 Update2

Dear Sir/Ma'am,

I tried to initiate a trial run  of the Osmose model (Version 3 update 2) while specifying the parameter "movement.distribution.method.sp#=random" for all the species integrated into the model. I also set the parameter "movement.distribution.ncell.sp#=50". However, when I tried to run the model (with all the other parameters set at their respective default values), I encountered an error as follows:

java.lang.ArrayIndexOutOfBoundsException: 58
at fr.ird.osmose.grid.AbstractGrid.getCell(
at fr.ird.osmose.step.DefaultStep.init(
at fr.ird.osmose.Simulation.init(
at fr.ird.osmose.Osmose$

With respect to the error that I have encountered, I have a query

Query 1

If the user decides to go with the random type of spatial distribution, are there any parameters other than "movement.distribution.method.sp#" & "movement.distribution.ncell.sp#" that needs to be specified explicitly in the osm_param-movement.csv file  ? If not, are there any additional steps that needs to be followed for proceeding the model run with random spatial distribution?

Kind regards,

Tarun Joseph.

Hello Tarun,

What you did for setting up random spatial distribution for all species looks good to me, there are no parameters other than  "movement.distribution.method.sp#" & "movement.distribution.ncell.sp#".

For a reason that I do not understand yet, it looks like there is a problem with your grid. Maybe we should dig in this direction. How did you define your grid ? How many cells are land / ocean ? etc.

Cheers, Philippe

Hello Philippe,

For the demo run, I defined the spatial grid exaclty as it was configured for the Southern Benguela region. The grid consist of 62 rows with 56 columns, where 1865 cells represent land and 1607 cells represent ocean. The only parameters that were changed for the run include "movement.distribution.method.sp#" & "movement.distribution.ncell.sp#". I suspect that I may have overlooked any factors that I ought to have considered while specifying the "movement.distribution.ncell.sp#" parameter? Could you share your insight on this issue?

Kind regards,

Tarun Joseph

OK, now I understand perfectly what you did. You started from the default OSM configuration and just changed the movement from maps to random.

I've just checked out the source code and found a bug in the class. I updated the source code with a fix for the problem. I send you a link by email to download the patched version.

Cheers, Philippe