How to get the same lifetime of every particle in output file

[jodua009]

Hi all.

I'm running OpenDrift using model fields from Nordfjords160. I want to seed a certain number of particles every 3rd hour and then give every particle a lifetime of two days (48 hours). The results should be written to a netcdf file (i.e. "output.nc")

My problem is that in my output the particles have different lifetimes. That means:

• Particles seeded first have a lifetime of 48 hours (this is what I want)
• Particles seeded after 3 hours have a lifetime of 45 hours (I want 48)
• Particles seeded after 6 hours have a lifetime of 42 hours (I still want 48)
• and so on...

Any suggestions on how I can fix this?

Best regards Johannes

[AndresSepulveda]

Hi,

How long is your run?

I use this option for that purpose and it works fine

o.set_config('drift:max_age_seconds',2592000) # 30 days in my case.

[jodua009]

Thanks a lot for your reply, and sorry for my late follow-up.

I want to run sets of say 100 particles that are seeded every three hours and that are given a lifetime of 48 hours. I want to do 4 seedings and have a time step of 10 minutes.

When I do this my output NetCDF-file gives me lon,lat variables with dimensions 289x400 (289 = 48x6+1). For the first 100 particles everything is fine. However, my problem is:

• For the next 100 I have (using particle 101 as an example) lon(1:18,101) = nan
• Then for the next 100 (using particle 201 as an example) lon(1:36,201) = nan
• Then for the next 100 (using particle 301 as an example) lon(1:54,301) = nan

So my particles are seeded when they are supposed to, but I will miss the last 3/6/9 hours for set 2/3/4. A workaround could be to run the simulation for 57 hours giving me output variables with dimensions 343x6. But I would prefer not to do it this way. At some stage I might want to do seedings on the first of every month, and I then believe this way would give me output variables with (too) huge dimensions.

Is there a way to force the output of each variable to start at index 1 independent of when the particle was seeded? If there are other better ways to do this, any suggestions would also be highly appreciated.

Johannes

[AndresSepulveda]

I think this is "normal" OpenDrift output. I am no sure your file size will be smaller. The idea is that with this approach there is a single time vector.

What I do to select the trajectories is to read the status variable and select, for each trajectory the time steps when status indicates that the particle is active.

Hope this helps,

Andres

[jodua009]

I see. I will try this out.
Thank you.