Home >> Forums >> Mercator data

Mercator data

Hello,
I don't succeed to run ichtyop with dataset come from Corpernicus. I recived this message :  Index 0 out of bounds for length 0 | java.lang.IndexOutOfBoundsException: Index 0 out of bounds for length 0. But first, i have the last version of java second, i download netbeans, and then i put netcdf data in input folder with this name : mercator2d-Caraibe.nc and xml files in folder cfg  Caraibe_config_mercactor2d.xml.
How can i resolve this problem ?
 

Ichthyop version: 
Ichthyop 3.3alpha
Hydrodynamic dataset: 
Mercator
Java version: 
Java SE 8
Operating system: 
Windows 10
Attachments: 

Hi

Can you please provide your configuration file? Thanks

Nicolas

Hi,
Thanks for your answer.
Jérôme

Hi Jerome

First, there is a mistake in the configuration file. You need to set the equal to "gridU file pattern" and "gridV file pattern" equal to "mercator2d-Caraibe.nc" (where both velocity fields are stored).

Second, I have encountered issues when initialising the run (I had to modify the code slightly). But still the results are "weird" (a lot of particles die at the domain edges).

Can you try to run the code with the updated configuration (filename + more precise release stain)? If it still does not work, I will provide you with the updated code.

Nicolas

Hi Nicolas,
Thanks a lot but when i do all modification, i recieved a new message : Ichthyop [servere] [Dataset]  Index -1 out of bounds for length 25 | java.lang.ArrayIndexOutOfBounds.
What do i do ?
Jérôme 

Hi Jerome

Can you try the enclosed NetCDF and tell me if it works?

https://filesender.renater.fr/?s=download&token=508b5d2b-c6f7-777b-6f57-...

Thanks

Nicolas

Hi Nicolas,
Thank you for your answers, it works now ! But i have a last question about wind drifter, How can i use NCEP wind (10 m above sea level) nc files to perform a simulation that include windage in ichthyop ?
Thanks
Jérôme

Hi Jerome

Sorry for the very late response (I skipped your request).

To use 2D wind data files to infer wind drift, you need to manually add a new block of options to your configuration file (cf. enclosed file).

When this is done, you should be able to see these new parameters in the Ichthyop GUI. Note however that this module has, to my knowledge, never been fully tested.

Nicolas

Attachments: 

Hi Nicolas,
Thank's for this answer, but i don't success to manually add this new block with enclosed xml file. Can you tell me more carefully, how to do it please ? and when i have to go to do it ?

Hi Jerome.
 

Send me your configuration file, I will add it for you. You can do it by opening the file with Notepad or any text editor.

Nicolas

Hi nicolas sorry for late.
Thank's a lot nicolas, I succeed

Hi,
I cannot success modelling with the new nodule windage. I change time origine but I don't know why but i already have a problem with time. How i can resolve this trouble please ?
Time origine from netcdf = year1900 month12 day31 00:00
In this nc files inclose, i compute time = time_value from nc *24*3600

Hi Jerome.

If you look at the NetCDF MT variable, the diff is approximately 3600. Therefore, I assume that the time unit in your file is seconds, and you have a time step of 3600 = 60 min * 60 sec = 1h.

So in your configuration file, you should set time_uinit as seconds, not hours.

Nicolas

Hi Nicolas,
I'm going to detail my problem, I have current nc file where the time origin is equale to 2000y 01month 01day 00:00:00. I make a conversion to have the time in seconds and ichthyop work very well. Infortunatly, want a add the wind nc file where the time origin is equal to 1900y 12month 31day, nothing work. First of all, i change time origin in hidden parameter to 1900y 12month 31day, i recieved this message :
java.lang.IndexOutOfBoundsException: {Wind dataset} Time value 2015/01/01 12:00:00 (3.5976528E9 seconds) not contained among NetCDF files.
Then, i try to translet time of my nc row wind in second, but i don't succeed to get a start of 3.5976528E9 second.
 after i edite my configuration xml file and i set time_unit as seconds. But i have the same error message :
java.lang.IndexOutOfBoundsException: {Wind dataset} Time value 2015/01/01 12:00:00 (3.5976528E9 seconds) not contained among NetCDF files.
Do you think all nc files must have the same time origin ? if yes do know, how to get all nc files with same tim origin ?
a other hand, do you know what kind of transformation i have to do in nc wind files to have good value ?