Opened 2 months ago

Last modified 2 months ago

#248 accepted Defect

Inconsistency between Pisso et al paper and model behavior

Reported by: ahilboll Owned by: saeck
Priority: major Milestone:
Component: FP other Version: FLEXPART 10.3
Keywords: Cc:

Description

In the Pisso et al. paper, it is said that

In simulations where a particle represents several species, all species are transported with the settling velocity of the first species. If this is undesired, simulations for the different species must be run separately.

However, when running the model with two species per release point, I get the following:

 WARNING: more than 1 species per release point, settling disabled

This seems to be contradicting. What is actually the case?

Given that the paper describes v10.3, I believe that either the model code (specifically, the exact words of the warning) or the paper should be updated.

Change History (5)

comment:1 follow-up: Changed 2 months ago by saeck

  • Owner set to saeck
  • Status changed from new to accepted

thank you for the notice, it seems the paper needs to be corrected. I checked the code and it is set that settling is disabled if there is at least one release point, where there are more than 1 species with mass>0 and different settling velocities released. If it is needed to have different settling velocities, the particles (with different species definition) have to be released from different releasepoints.

comment:2 in reply to: ↑ 1 Changed 2 months ago by ahilboll

Replying to saeck:

If it is needed to have different settling velocities, the particles (with different species definition) have to be released from different releasepoints.

But how would I do that without running FLEXPART once per species?

From what I understand (of course I'm probably wrong) I set the species in the RELEASES file header (i.e. in the &RELEASES_CTRL section) and the paper says that in case of multiple species I should have multiple entries for MASS in the &RELEASE section. Now if I set NSPEC=2, SPECNUM_REL=15,16 and then have multiple &RELEASE sections, how would FLEXPART know what to do?

Maybe you could post an example how I need to set this up?

comment:3 Changed 2 months ago by saeck

In the RELEASES file you would define e.g. 2 different species like

NSPEC= 2,
SPECNUM_REL= 1, 2,

and then you have one releasepoint where you release spec 1 with mass and spec 2 with 0 mass and one where you release species 2 with mass and spec 1 with 0 mass:

MASS= 0.00000000 , 1.00000000 ,
MASS= 1.00000000 , 0.00000000 ,

comment:4 Changed 2 months ago by ahilboll

Thanks a lot, Sabine, that's very useful!

comment:5 Changed 2 months ago by ahilboll

I think I found another inconsistency between code and paper:

When it comes to the SPECIES definitions, Table 8 in the paper lists [m] as units for psigma. However, when I look at the SPECIES_NNN files, the psigma has values between 0.3 and 1.2. That can hardly be in meters ... I suspect that the code expects psigma to be relative standard deviation (i.e., divided by the mean), so the units in the table should be 1 and the description should read [...] Species diameter relative standard deviation.

Note: See TracTickets for help on using tickets.
hosted by ZAMG