Opened 5 years ago
Closed 5 years ago
#215 closed Defect (fixed)
Bug in flex_extract v7.0.3 ERA5 data extraction
| Reported by: | anphi | Owned by: | anphi |
|---|---|---|---|
| Priority: | blocker | Milestone: | |
| Component: | flex_extract | Version: | flex_extract_v7.0.3 |
| Keywords: | ERA5 | Cc: |
Description
Recently, I found out that flex_extract version 7.0.3 has an error in the extraction of ERA5 data!
PLEASE STOP USING ERA5 DATA EXTRACTION!!!!
Times 17 and 18 do not contain any flux fields (e.g. precipitation).
And the rest of the times it seems that the flux fields are shifted in time.
I'll provide a bug fix as soon as possible!
Explanations will follow ...
Change History (9)
comment:1 Changed 5 years ago by pesei
- Status changed from new to assigned
comment:2 Changed 5 years ago by ahilboll
comment:3 Changed 5 years ago by anphi
Yes, I will!
comment:4 Changed 5 years ago by ahilboll
I'm just guessing here (don't have the time to really get into this at the moment), but could it be that in ll. 1468 of FlexpartTools.py,
if not (step==c.maxstep and c.maxstep>12 or sdates==elimit):
vdp.pop(0)
svdp.pop(0)
there's something wrong? It seems strange to pop from vdp as well here. But then again, I don't really know what I'm talking about ...
comment:5 Changed 5 years ago by anphi
This should be the correct behavior. We are going over the time series keeping just the last 4 time steps in these vpd and svdp lists.
comment:6 Changed 5 years ago by pesei
- Milestone set to flex_extract_v7.1
comment:7 Changed 5 years ago by anphi
- Milestone flex_extract_v7.1 deleted
comment:8 Changed 5 years ago by anphi
The problem is solved and the bugfix version is announced as 7.0.4 through the mailing list.
I am closing the ticket.
comment:9 Changed 5 years ago by anphi
- Resolution set to fixed
- Status changed from assigned to closed
Hi, thanks for the warning! Can you please announce on the mailing list once it's fixed? I just found out about this problem accidentally while I was browsing the timeline ...