Opened 17 years ago

Closed 17 years ago

Last modified 16 years ago

#126 closed enhancement (fixed)

Thinner version ID

Reported by: Dave Offiler Owned by: Dave Offiler
Priority: normal Milestone: 1.1
Component: ROPP (all) Version: 1.0
Keywords: Thinner, version ID Cc:

Description

Kent suggests adding a separate version ID for the ROPP thinner code to the ropp_io version ID so as to be able to know which thinner is being used by DMI vs EUM (or whoever) for compariosn/validation purposes.

The simplest solution would appear to be to add a version ID to the new (v1.1) thin_method string in the ROPP structure which is also saved as a global attribute in the netCDF file.

Change history (3)

comment:1 by Dave Offiler, 17 years ago

Resolution: fixed
Status: newclosed

a) Extended thin_method string in ROPP main structure in ropp_io_types.f90 from 60 to 80 characters.

b) Added software ID string parameter definition to ropp_io_thin() [ "v1.1" ] and added this to rodata%thin_method string after the method ID & title.

c) Re-compiled (ifort) and tested by running (in ropp_io/data):

$ ../tools/ropp2ropp ropp_test.nc -d -p 100 -o t1.nc
Reading ropp_test.nc ...

Thinning method: SAMPLE (Sampling to a maximum of 100 levels) [v1.1]
Level 1b data (BA) not thinned
Refractivity data not thinned
Level 2b data (T,q,P) not thinned
Ensuring all profiles are in ascending height order...

Profile   1
 FmtVer:  ROPP I/O V1.1
 OccID :  OC_20071120122120_CHMP_G004_GFZ_
 LEOID :  CHMP
 GNSID :  G004
 STNID :  UNKN
 ProCen:  GFZ GeoForschungsZentrum Potsdam
 DTocc :  12:21:20.000 20-Nov-2007
 DTpro :  23:59:59.999 20-Nov-2007
 GEOref:  158.000   -62.2   156.9 6385605.5     0.0     0.0     0.0   174.3   -33.6
 PCD   :  0000000000000000
 Qual  :    100.0
 PODmth:  UNKNOWN
 LVLmth:  UNKNOWN  UNKNOWN  UNKNOWN
 THNmth:  SAMPLE (Sampling to a maximum of 100 levels) [v1.1]          <<<<<-----
 SW-ID :  UNKNOWN
 B/G   :  NONE
 ModTyp:  NONE
 Npts  :      0  319  319  319    0    0

Actually, something is wrong here: there are 319 BA, Ref & TqP samples in the input file, and the command should sample to <=100 - which it hasn't done, and even says its hasn't. This needs separate investigation'''

And further:

> ncdump -h t1.nc
...
               :thin_method = "NONE (Thinning disabled) [v1.1]" ;
...

d) Re-generated test files in ropp_io/data using gentestfiles.sh

e) Committed as [1307]

comment:2 by (none), 16 years ago

Milestone: 1.1

Milestone 1.1 deleted

comment:3 by Huw Lewis, 16 years ago

Milestone: 1.1
Note: See TracTickets for help on using tickets.