Load bio profiles
Jaime Pitarch Portero
I am trying to load measured profiles of phytoplankton into GOTM. As far as I
investigated, only the last developers version is prepared to do it, but the
relevant lines in observations.F90 are between those "#ifdef BIO". (What is
BIO?) I removed all them, but then I get the linking problem:
ifort -o gotm_prod_IFORT ./gotm/main.o -DIFORT -DNETCDF_FMT -DNETCDF3
-DSEAGRASS -DNO_0D_BIO -DPRODUCTION -DFORTRAN95 -DREAL_4B=real\(kind=4\) -O3
-module /home/jaime/gotm-git/modules/IFORT -I/home/jaime/netcdf40/include
-I/usr/local/include -I/home/jaime/gotm-git/include
-I/home/jaime/gotm-git/modules/IFORT -fPIC -L/home/jaime/gotm-git/lib/IFORT
-lgotm_prod -lseagrass_prod -lairsea_prod -lmeanflow_prod -lturbulence_prod
-lobservations_prod -loutput_prod -lutil_prod
/home/jaime/netcdf40/lib/libnetcdf.a
/home/jaime/gotm-git/lib/IFORT/libobservations_prod.a(observations.o): In
function `observations_mp_get_all_obs_':
observations.F90:(.text+0x3f7): undefined reference to `get_bio_profiles_'
/home/jaime/gotm-git/lib/IFORT/libobservations_prod.a(observations.o): In
function `observations_mp_init_observations_':
observations.F90:(.text+0xb1fa): undefined reference to `get_bio_profiles_'
Has anyone a version of the code with the whole "loading-simulating-storing"
process working? If not, is it possible to fix this with a reasonable effort?
Thanks in advance,
Jaime.
Karsten Bolding
> Hello all users and developers,
> I am trying to load measured profiles of phytoplankton into GOTM. As far as I
> investigated, only the last developers version is prepared to do it, but the
> relevant lines in observations.F90 are between those "#ifdef BIO". (What is
> BIO?) I removed all them, but then I get the linking problem:
they were thee for a reason - check conditional compilation.
>
<snip>
>
> Has anyone a version of the code with the whole "loading-simulating-storing"
> process working? If not, is it possible to fix this with a reasonable effort?
resonable effort - is that yours or ours?
>
> Thanks in advance,
> Jaime.
>
Karsten
Jorn Bruggeman
> I am trying to load measured profiles of phytoplankton into GOTM. As
> far as I investigated, only the last developers version is prepared to
> do it, but the relevant lines in observations.F90 are between those
> "#ifdef BIO". (What is
> BIO?) I removed all them, but then I get the linking problem:
By default, GOTM compiles without the biogeochemical models. To active
those, you need to set environment variable "BIO" to "true" (e.g., "export
BIO=true" in bash) before running make. That will in turn set the
preprocessor macro BIO during compilation, which activates any lines
protected by #ifdef BIO ... #endif, and includes all bio files.
As long as you are moving to the developers version anyway, it is worth
noting that the old "bio" framework will be deprecated in favour of the new
Framework for Aquatic Biogeochemical Models
(http://sourceforge.net/projects/fabm/; manual at
http://www.meece.eu/documents/deliverables/WP2/D2.14.pdf). This framework
supports everything the old bio framework used to, and much more. In
particular, it allows fine-grained prescription of per-variable initial
conditions for biogeochemistry (p 18-20 in the manual).
Cheers,
Jorn
Markus Kreus
I once coded/modified the routine for reading in observed biogeochemical
properties from separate files.
Unfortunately, it did never enter the repository and the patch will
likely be outdated for just a 'plug-in' to the recent devel version.
I'll have a look whether I can fix it, ...
Cheers Mark
Jaime Pitarch Portero
I found the line BIO=false in Rules.make, and changed it to BIO=true. Now, it
seems to compile with all the bio modules.
Unfortunately, when I execute the namelists, it gives the following error:
forrtl: severe (24): end-of-file during read, unit 10, file /media/Data
/greifensee_nml/obs.nml, line 509, position 0
Image PC Routine Line Source
gotm_prod_IFORT 081F0979 Unknown Unknown Unknown
(...)
This error comes when trying to execute the line:
read(namlst,nml=bioprofiles,err=93)
In observations.F90, the list of variables loaded is:
namelist /bioprofiles/ bio_prof_method,bio_prof_file
And in obs.F90, the corresponding lines are:
&bioprofiles
bio_prof_method = 2,
bio_prof_file = 'bioprofs.dat',
Everything seems to be ok, and that's why I do not understand this error. Any
idea?
Thanks,
Jaime.
Mensaje citado por Markus Kreus <markus...@zmaw.de>:
--
Markus Kreus
I merged the routines mentioned beyond with a fresh git download. So
far, I got it work again, but only without support for output to the
netcdf-file.
If you are still interested, you can try out.
There are two new files (observation/get_obs_profiles.F90 and
output/save_obs_profiles.F90). Modifications to existing files are shown
in read_obs_...diff. The code runs without 'BIO' defined, as
observations may be more general than that.
I did a quick test with the gotland testcase. Necessary information on
the files to open are specified at the end of obs.nml
Cheers Markus
Jaime Pitarch Portero
I could compile and run the code from observed NPZD profiles. It does the time
loop but at the end, it gives the terrific message:
Output has been written in NetCDF
*** glibc detected *** /home/jaime/gotm-git/bin/gotm_prod_IFORT: free(): invalid
next size (fast): 0x0877fe78 ***
======= Backtrace: =========
/lib/i386-linux-gnu/libc.so.6(+0x6ebc2)[0xb7632bc2]
/lib/i386-linux-gnu/libc.so.6(+0x6f862)[0xb7633862]
/lib/i386-linux-gnu/libc.so.6(cfree+0x6d)[0xb763694d]
/home/jaime/gotm-git/bin/gotm_prod_IFORT[0x816b42c]
/home/jaime/gotm-git/bin/gotm_prod_IFORT[0x8161fb7]
/home/jaime/gotm-git/bin/gotm_prod_IFORT[0x8162172]
/home/jaime/gotm-git/bin/gotm_prod_IFORT[0x8162201]
/home/jaime/gotm-git/bin/gotm_prod_IFORT[0x815e29b]
/home/jaime/gotm-git/bin/gotm_prod_IFORT[0x815ff3b]
/home/jaime/gotm-git/bin/gotm_prod_IFORT[0x80ef5eb]
/home/jaime/gotm-git/bin/gotm_prod_IFORT[0x80cc83a]
[0x0]
======= Memory map: ========
08048000-082bc000 r-xp 00000000 08:05 919797
/home/jaime/gotm-git/bin/gotm_prod_IFORT
082bc000-082c3000 rw-p 00274000 08:05 919797
/home/jaime/gotm-git/bin/gotm_prod_IFORT
082c3000-0849c000 rw-p 00000000 00:00 0
08744000-087b8000 rw-p 00000000 00:00 0 [heap]
b7400000-b7421000 rw-p 00000000 00:00 0
b7421000-b7500000 ---p 00000000 00:00 0
b759f000-b75a0000 rw-p 00000000 00:00 0
b75a0000-b75a3000 r-xp 00000000 08:05 12583856
/lib/i386-linux-gnu/libdl-2.13.so
b75a3000-b75a4000 r--p 00002000 08:05 12583856
/lib/i386-linux-gnu/libdl-2.13.so
b75a4000-b75a5000 rw-p 00003000 08:05 12583856
/lib/i386-linux-gnu/libdl-2.13.so
b75a5000-b75c1000 r-xp 00000000 08:05 12583866
/lib/i386-linux-gnu/libgcc_s.so.1
b75c1000-b75c2000 r--p 0001b000 08:05 12583866
/lib/i386-linux-gnu/libgcc_s.so.1
b75c2000-b75c3000 rw-p 0001c000 08:05 12583866
/lib/i386-linux-gnu/libgcc_s.so.1
b75c3000-b75c4000 rw-p 00000000 00:00 0
b75c4000-b773a000 r-xp 00000000 08:05 12583845
/lib/i386-linux-gnu/libc-2.13.so
b773a000-b773c000 r--p 00176000 08:05 12583845
/lib/i386-linux-gnu/libc-2.13.so
b773c000-b773d000 rw-p 00178000 08:05 12583845
/lib/i386-linux-gnu/libc-2.13.so
b773d000-b7740000 rw-p 00000000 00:00 0
b7740000-b7768000 r-xp 00000000 08:05 12583875
/lib/i386-linux-gnu/libm-2.13.so
b7768000-b7769000 r--p 00028000 08:05 12583875
/lib/i386-linux-gnu/libm-2.13.so
b7769000-b776a000 rw-p 00029000 08:05 12583875
/lib/i386-linux-gnu/libm-2.13.so
b777b000-b777d000 rw-p 00000000 00:00 0
b777d000-b777e000 r-xp 00000000 00:00 0 [vdso]
b777e000-b779c000 r-xp 00000000 08:05 12583832 /lib/i386-linux-gnu/ld-2.13.so
b779c000-b779d000 r--p 0001d000 08:05 12583832 /lib/i386-linux-gnu/ld-2.13.so
b779d000-b779e000 rw-p 0001e000 08:05 12583832 /lib/i386-linux-gnu/ld-2.13.so
bfad6000-bfaf8000 rw-p 00000000 00:00 0 [stack]
forrtl: error (76): Abort trap signal
Image PC Routine Line Source
. B777D424 Unknown Unknown Unknown
libc.so.6 B7632BC2 Unknown Unknown Unknown
In fact, it remains hung and doesn't release the control back. Ayy, I am close
to solve the problem!!
Thanks all,
Jaime.
Mensaje citado por Markus Kreus <markus...@zmaw.de>:
> Hello Jamie,
--