Search found 77 matches

by mathieu
Sun Jun 24, 2018 6:00 pm
Forum: ROMS FAQ
Topic: Segmentation Faults
Replies: 12
Views: 30846

Re: Segmentation Faults

Hernan, a remark on your point on "wrap-around integer". Actually in Fortran (and C/C++) the integer overflow is undefined behavior. See for example https://stackoverflow.com/questions/405 ... r-overflow
So, gfortran is right to stop at that.
by mathieu
Fri Jun 22, 2018 3:56 pm
Forum: ROMS FAQ
Topic: Segmentation Faults
Replies: 12
Views: 30846

Re: Segmentation Faults

The temporary arrays are when you pass a A(1,:) array to a subroutine. Since the values are not aligned there is a need for a new array which of course slows things down. But it is no problem if done only in the input parameter reading. It is of course a problem if wclock_on is called recursively. S...
by mathieu
Mon Jun 18, 2018 2:34 pm
Forum: ROMS FAQ
Topic: Segmentation Faults
Replies: 12
Views: 30846

Re: Segmentation Faults

Hi Kate, in my experience it has never happened that the compiler was wrong. Bug detected for one compiler but not for the other means bug. In order to detect the bug in gfortran you can use compilation option -fcheck=all -fsanitize=address -fsanitize=undefined. For Intel Fortran Compiler, options a...
by mathieu
Wed Jul 05, 2017 9:33 am
Forum: ROMS Usage
Topic: COAWST problem when using 2way WRF-ROMS
Replies: 6
Views: 12134

Re: COAWST problem when using 2way WRF-ROMS

One possibility to consider is whether the WRF grid completely covers the ROMS grid.
A variant of that possibility is that the interpolation matrix is not well computed for some reason.
by mathieu
Sat Sep 17, 2016 5:23 pm
Forum: ROMS Discussion
Topic: blowing up with ubar and vbar with min and max as NaNs
Replies: 9
Views: 9249

Re: blowing up with ubar and vbar with min and max as NaNs

If you compile with ifort, you can detect the precise line where the NaN first occur with the compile option -fpe0. With gfortran the option is -ffpe-trap=invalid.
by mathieu
Wed Jul 29, 2015 1:50 pm
Forum: ROMS Problems
Topic: forrtl: severe (174): SIGSEGV, segmentation fault occurred
Replies: 24
Views: 17860

Re: forrtl: severe (174): SIGSEGV, segmentation fault occurr

As Kate said, you have to use USE_DEBUG and maybe edit the corresponding file in Compiler directory for changing the debug compiler flags. The "Unknown" will then be replaced by a precise statement of the nature of the error. In truth, it is more likely that there is an error in your confi...
by mathieu
Tue Oct 23, 2012 12:19 pm
Forum: ROMS Discussion
Topic: MPI parallelization
Replies: 1
Views: 2409

MPI parallelization

I put this post because I wonder if the MPI parallelization of ROMS could be improved. The exchange code is in mp_exchange.F and contains typical pairs mpi_irecv/mpi_send . What about using the slightly more advanced MPI_SENDRECV function? A priori it moves some of the buffering to the MPI library a...
by mathieu
Thu Apr 19, 2012 7:57 pm
Forum: ROMS Discussion
Topic: model blows up when using bulk forcing..
Replies: 4
Views: 4637

Re: model blows up when using bulk forcing..

use DIAGNOSTICS_UV and then select the right Dout.
But it is possible that the bathymetry is not the problem.
You need to look at what happened at the time steps before
the blow up. Another idea is to simplify the problem by
removing some feature of your setting, this can help in
some cases.
by mathieu
Wed Apr 18, 2012 6:39 am
Forum: ROMS Discussion
Topic: model blows up when using bulk forcing..
Replies: 4
Views: 4637

Re: model blows up when using bulk forcing..

What makes you think bulk forcing is responsible? There are many other things that could create the blow-up: horizontal pressure gradient (here you have relatively large rx1=10), NaN or inconsistent forcing, too high speed (you have automatic blow at 20m/s), etc. You give too few information for us ...
by mathieu
Tue Apr 03, 2012 12:00 pm
Forum: ROMS Problems
Topic: SOURCE POINT FOR RIVER RUNOFF (topic locked)
Replies: 115
Views: 2387605

Re: SOURCE POINT FOR RIVER RUNOFF

You have to be aware that by setting river_Vshape(k,i)=1/N; you are not setting uniform flux over the vertical profile. Let us write the vertical positions as z_w(i) with i between 0 and N. In your case, what this means is that 1/N of the flux goes into the layer between level z_w(i) and z_w(i+1) fo...
by mathieu
Tue Dec 20, 2011 12:04 pm
Forum: ROMS Discussion
Topic: How can make sure proper river point position in ROMS
Replies: 15
Views: 19677

Re: How can make sure proper river point position in ROMS

Some general remarks without looking at the details. First, the points that are put in the river file are u - and v -points. Second the coordinates are put with respect to the internal coordinate system of ROMS and there is an index shift documented by Hernan recently in the forum. Lastly, the way r...
by mathieu
Tue Dec 20, 2011 9:18 am
Forum: ROMS Discussion
Topic: How can make sure proper river point position in ROMS
Replies: 15
Views: 19677

Re: How can make sure proper river point position in ROMS

Nothing so complicated. it is simply the mean with the NaN values removed.
by mathieu
Sat Nov 05, 2011 7:50 am
Forum: ROMS Usage
Topic: creating periodic grid ...
Replies: 4
Views: 5111

Re: creating periodic grid ...

First of all you have to realize that ROMS does not use the longitude/latitude internally. It uses pm , pn , Coriolis factor f and a few others. That is why you can make periodic grid at all because if longitude were used then you would have a problem with the ambiguity 0=360. Hence you can actually...
by mathieu
Tue Oct 25, 2011 11:34 am
Forum: ROMS Discussion
Topic: The Segmentation fault of ROMS 3.6
Replies: 15
Views: 11548

Re: The Segmentation fault of ROMS 3.6

The key warning is Clock skew detected. Your build may be incomplete. This happens when you are compiling on a distant computer and that there is a clock shift between it and your computer. The solution to that is to set your clocks correctly by using the NTP protocol and/or to do make from a clean ...
by mathieu
Wed Jul 13, 2011 10:22 am
Forum: ROMS Discussion
Topic: Hydrostatic consistency and the rx0 rx1 factors
Replies: 4
Views: 6706

Re: Hydrostatic consistency and the rx0 rx1 factors

Yes the number given for rx0, rx1 correspond to the other one given in this forum. I have not checked the formula but as far as I remember the hydrostatic consistency means that two adjacent vertical cells share one level and it is equivalent to rx1 <= 1 . Yes the erroneous pressure gradient depend ...
by mathieu
Wed Jun 29, 2011 6:02 am
Forum: ROMS Problems
Topic: Changing AKT_BAK values has no effect on model results
Replies: 8
Views: 7728

Re: Changing AKT_BAK values has no effect on model results

AKT_BAK is the background vertical tracer advection mixing coefficient for tracers. It is thanks to it that temperature/salinity got slowly uniformized vertically. The mixing coefficient is described by the turbulence parameterization and this is of course a difficult problem with no ideal solution...
by mathieu
Wed Jun 01, 2011 8:58 am
Forum: ROMS Problems
Topic: mpirun problem
Replies: 7
Views: 6246

Re: mpirun problem

I think the problem is now probably in SWAN. Look after the Errfile* for the problem.
by mathieu
Tue May 31, 2011 7:56 am
Forum: ROMS Problems
Topic: mpirun problem
Replies: 7
Views: 6246

Re: mpirun problem

One possible explanation, that you need to check, is that the mpif90 and mpirun/mpiexec do not come from the same origin.
by mathieu
Tue Apr 05, 2011 10:09 am
Forum: ROMS Source
Topic: MCT and ARPACK
Replies: 3
Views: 6465

Re: MCT and ARPACK

Ok, so ARPACK has been modified. But are those modifications documented?
Has MCT been modified as well? Will those modifications be included in current distribution of those softwares?
by mathieu
Mon Apr 04, 2011 3:27 pm
Forum: ROMS Source
Topic: MCT and ARPACK
Replies: 3
Views: 6465

MCT and ARPACK

ROMS comes with included MCT and ARPACK libraries. What is the reason for that?
Is it for convenience so that we can compile it easily or were there some modifications to the code?
If so, what were those modifications?
by mathieu
Mon Mar 21, 2011 8:50 am
Forum: ROMS Discussion
Topic: How can make sure proper river point position in ROMS
Replies: 15
Views: 19677

Re: How can make sure proper river point position in ROMS

I have not derived the formula myself. I followed for that the map_rivers.m file which I put in attachment. This is not the original file but a hacked version of it. I do not know the original author. I think it is mandatory to use this program before any ROMS run using rivers. The formulas are that...
by mathieu
Fri Mar 18, 2011 9:44 am
Forum: ROMS Discussion
Topic: How can make sure proper river point position in ROMS
Replies: 15
Views: 19677

Re: How can make sure proper river point position in ROMS

The river inflow in ROMS is quite complicated. In the netcdf file of rivers you have the following arrays: double river_Xposition(river) ; river_Xposition:long_name = "river ETA-position at RHO-points" ; river_Xposition:units = "nondimensional" ; river_Xposition:field = "riv...
by mathieu
Mon Jan 31, 2011 8:10 am
Forum: ROMS Problems
Topic: Intel’s new i7 980x CPU gives disappointing speedup
Replies: 48
Views: 78506

Re: Intel’s new i7 980x CPU gives disappointing speedup

In the 90s the talk was that well optimized fortran libraries were about 30% faster than the same C code. The common explanation was that in fortran there is no pointer and all arrays are directly declared which gives the compiler much room to optimize the code. This is as opposed to C code where po...
by mathieu
Wed Nov 24, 2010 8:16 am
Forum: ROMS Discussion
Topic: Is Evaporation acting as a sink of mass?
Replies: 4
Views: 4196

Re: Is Evaporation acting as a sink of mass?

The key code to look at is bulk_flux and there you see that the evaporation is modeled as a salt flux : stflx(i,j,isalt)=cff*(evap(i,j)-rain(i,j)) Thus the model does not lose water but increases its total quantity of salt. SELFE do the same thing as ROMS but SHYFEM does it correctly as a water flux.
by mathieu
Fri Nov 05, 2010 12:27 pm
Forum: ROMS Source
Topic: Problem with nearshore_mellor05.h
Replies: 1
Views: 4871

Problem with nearshore_mellor05.h

I found a number of problems with the implementation of the nearshore formulation of the radiation stress. The problem can be detected very easily since there is both a formulation of the stress for barotropic and baroclinic in nearshore_mellor05.h . Hence if one removes the #ifdef SOLVE3D ... #else...
by mathieu
Fri Sep 10, 2010 7:03 am
Forum: ROMS Problems
Topic: help regarding the time step
Replies: 5
Views: 6256

Re: help regarding the time step

I think the value of rx1 at 14 is too high and that you need to smooth more your bathymetry. Your choice of vertical stratification Vtransform=1 and Vstretching=1 is no longer the common one. I would use Vtransform=2 and check if the vertical stratification is really the one that I need. Otherwise, ...
by mathieu
Thu Sep 02, 2010 5:58 am
Forum: ROMS Discussion
Topic: Wet-points only ROMS
Replies: 2
Views: 2926

Re: Wet-points only ROMS

I think computing only on wet points would lead to a big performance loss. You see the processors are pipelined and thus an if test wrongly predicted can lead to a several cycles lost. That is why you have zeta(i,j)=zeta_new(i,j)*rmask(i,j) and not IF (rmask(i,j).eq.1) THEN zeta(i,j)=zeta_new(i,j) E...
by mathieu
Tue Jun 15, 2010 10:17 am
Forum: ROMS Discussion
Topic: Problem on coupling
Replies: 4
Views: 4222

Re: Problem on coupling

One possible explanation for this problem is that SWAN reduces his grid when it has full land mask on it. This allows it to gain a little in speed but creates a mismatch in grid size.
The solution to this problem is to set up the bathymetry to zero in the swan grid when there is land.
by mathieu
Sun Apr 18, 2010 11:07 am
Forum: ROMS Results
Topic: some suggestion with history file and average file
Replies: 7
Views: 10716

Re: some suggestion with history file and average file

I asked to do something like that some years ago and it was refused for understandable reasons: ROMS is already sufficiently complex. For your needs a workable solution is the following: have a a perl script that run in an infinite loop and check if a file his_0044.nc and his_0045.nc exist. If so th...
by mathieu
Thu Apr 15, 2010 9:08 am
Forum: ROMS Discussion
Topic: Salt conservation
Replies: 1
Views: 2302

Salt conservation

When sea water evaporates from the surface of the ocean some denser waters are formed on the surface, which then goes deeper. This physical phenomena is modelled in ROMS by a salt flux , that is an increase in salt concentration at the surface of the model with the same free surface. The relevant co...
by mathieu
Fri Apr 09, 2010 2:05 pm
Forum: ROMS Problems
Topic: GET_2DFLD - unable to find requested variable: wind_time
Replies: 5
Views: 5804

Re: GET_2DFLD - unable to find requested variable: wind_time

Well try replacing bulk_time by wind_time or srf_time, etc. and tell us how it goes. Splitting file in several forcing files also helps in practice.
by mathieu
Fri Apr 09, 2010 7:16 am
Forum: ROMS Problems
Topic: GET_2DFLD - unable to find requested variable: wind_time
Replies: 5
Views: 5804

Re: GET_2DFLD - unable to find requested variable: wind_time

Please show us the header of your wind forcing file by doing ncdump -h wind_frc.nc One header that work well (other combinations are possible) is netcdf aladin_wet_wind { dimensions: eta_rho = 291 ; xi_rho = 266 ; wind_time = 489 ; variables: double wind_time(wind_time) ; wind_time:units = "day...
by mathieu
Sun Mar 21, 2010 11:10 am
Forum: ROMS Problems
Topic: a question about NUDGING_SST || defined NUDGING_T
Replies: 3
Views: 3757

Re: a question about NUDGING_SST || defined NUDGING_T

The assimilation file has to be put in APARNAM if I am not mistaken.
Of course you have to interpolate fields to the grid of ROMS.
by mathieu
Fri Mar 19, 2010 7:27 am
Forum: ROMS Discussion
Topic: mpirun
Replies: 1
Views: 3211

Re: mpirun

I think this is ROMS agrif that you are trying to compile.
See https://gforge.inria.fr/forum/?group_id=735 and http://roms.mpl.ird.fr/ for questions.
by mathieu
Thu Mar 18, 2010 7:56 am
Forum: ROMS Problems
Topic: a question about NUDGING_SST || defined NUDGING_T
Replies: 3
Views: 3757

Re: a question about NUDGING_SST || defined NUDGING_T

The basic idea of NUDGING_T is to use measurement of temperature and nudge the model towards those values. The problem is of course that temperature measurements are sparse: they are available typically at some point by CTD measurements or at the surface by satellite measurement. Thus enters the fun...
by mathieu
Wed Mar 17, 2010 7:51 am
Forum: ROMS Discussion
Topic: forcing files
Replies: 4
Views: 4722

Re: forcing files

What happens is the following. There is a mechanism in ROMS named regridding where you put as input to ROMS the non-interpolated output of a meteorological model and ROMS does the interpolation by himself. The code is fairly hackish as opposed to the rest of ROMS but it is useful to save disk space....
by mathieu
Sat Jan 30, 2010 12:30 pm
Forum: ROMS Usage
Topic: Unphysical velocities using custom salinity
Replies: 6
Views: 7584

Re: Unphysical velocities using custom salinity

I believe you are a little too aggressive with the vertical parametrization. If you choose theta_s=4 , theta_b=0.35 , Vtransform=2 , Vstrecthing=1 then the need for rx0 smoothing is much reduced. In principle the higher theta_s the better the surface is resolved (at the expense of realism) and the l...
by mathieu
Tue Jan 19, 2010 8:51 am
Forum: ROMS Usage
Topic: Unphysical velocities using custom salinity
Replies: 6
Views: 7584

Re: Unphysical velocities using custom salinity

Dear eivinds, yes smoothing your bathymetry is probably a good idea. But you must look at the rx1 factor, which must be around 5 or 6 at most. With the vertical stretching that impose something like rx0 around 0.18. The key point is the error in the computation of the pressure gradient. You have an ...
by mathieu
Thu Oct 22, 2009 11:22 am
Forum: ROMS Source
Topic: Wave output
Replies: 0
Views: 3836

Wave output

The ROMS model can output the variable "Lwave", "Dwave", "Hwave" obtained from the SWAN model running coupled but apparently it is not possible to output the variables "Pwave_top", "Pwave_bot", "Ub_swan" and "Wave_dissip". If this...
by mathieu
Thu Aug 27, 2009 11:30 am
Forum: ROMS Problems
Topic: help~really confused about the so called add tidal forcing~
Replies: 7
Views: 5801

Re: help~really confused about the so called add tidal forcing~

Some remarks: The code in set_tides.F has the code # if defined WEST_FSOBC || defined WEST_M2OBC in the section for adding the tidal variability to zeta_west , but zeta_west exist only if WEST_FSOBC is triggered. Thus this simplifies to the clearer #ifdef WEST_FSOBC In this case, SOUTH_M2NUDGING has...
by mathieu
Wed Aug 26, 2009 8:52 am
Forum: ROMS Tools and Techniques
Topic: Offline nesting
Replies: 3
Views: 3828

Re: Offline nesting

This is a classic operation.
Everybody has different scripts. Mine are available from http://www.liga.ens.fr/~dutour/Roms_Interpolation/.
The idea is to interpolate from the history file (or from stations
if prefered) to the boundary points and then create a boundary forcing
file.
by mathieu
Mon Jul 06, 2009 8:43 am
Forum: ROMS Problems
Topic: unable to find requested variable: zeta_time
Replies: 10
Views: 6589

Re: unable to find requested variable: zeta_time

Then, my next suggestion would be to discard completely the variable bry_time and use instead zeta_time , v2d_time , v3d_time , temp_time , salt_time . There is indeed a cdl file bry_unlimit.cdl in the ROMS distribution, but the variable bry_time that it uses does not appear anywhere in the shipped ...
by mathieu
Thu Jul 02, 2009 5:20 am
Forum: ROMS Problems
Topic: unable to find requested variable: zeta_time
Replies: 10
Views: 6589

Re: unable to find requested variable: zeta_time

Try replacing
nc{'zeta_time'} = ncdouble('bry_time');
by

Code: Select all

nc{'zeta_time'} = ncdouble('zeta_time');
by mathieu
Mon Jun 22, 2009 1:20 pm
Forum: ROMS Problems
Topic: My seashore test
Replies: 36
Views: 27599

Re: My seashore test

Ok, so it is not compiler. What I would do in your place is: --Disable NEARSHORE_MELLOR , this radiation stress can be large and has some problems. --Do not use waves. I am not sure what you mean with Lwave=0.0 but if it means constant values of Lwave and Hwave in your ocean_frc.nc file then it is m...
by mathieu
Sun Jun 21, 2009 10:11 am
Forum: ROMS Problems
Topic: My seashore test
Replies: 36
Views: 27599

Re: My seashore test

Dear fancer, thank you for bug report. I corrected my program. Otherwise, for your problem, look at the following: --Your maximum barotropic number is about 0.958 which is almost at the limit of 1. Try running with a reduced dt and a value of 0.2 for maximum barotropic. --Try to get a running system...
by mathieu
Fri Jun 19, 2009 8:10 am
Forum: ROMS Problems
Topic: My seashore test
Replies: 36
Views: 27599

Re: My seashore test

The thing that is wrong with the application is the smoothing of the very high roughness. You had Maximum grid stiffness ratios: rx0 = 9.926508E-01 (Beckmann and Haidvogel) rx1 = 2.196218E+04 (Haney) Those are incredibly large value. Maximum rx0 should be 0.2 and maximum rx1 should be 6. Above those...
by mathieu
Thu Jun 18, 2009 2:07 pm
Forum: ROMS Problems
Topic: Lfloat option
Replies: 3
Views: 3189

Re: Lfloat option

Ok, so Hernan corrected the problem in revision 354. Thank you. Another problem I encountered is with the entry at ocean_time=1 of the netcdf FLTNAME file. At present the field lon and lat are set to zero and the remaining fields are set to fillvalue. Maybe it is a feature and not a bug but this is ...
by mathieu
Wed Jun 10, 2009 12:33 pm
Forum: ROMS Problems
Topic: Lfloat option
Replies: 3
Views: 3189

Lfloat option

One of the nice feature of ROMS is the ability when running with STATIONS or FLOATS to actually disable those options. For STATIONS the way to do is to have the following station files. Lstations == F For FLOATS the corresponding file is Lfloats == F But unfortunately this does not work, there are s...
by mathieu
Tue Jun 02, 2009 8:37 am
Forum: ROMS Problems
Topic: How to do nesting using ROMS?
Replies: 7
Views: 6462

Re: How to do nesting using ROMS?

One possible set of matlab script for doing this is available from http://drobilica.irb.hr/~mathieu/Roms_Interpolation/ Two methods are available: interpolating using stations or using history files. The grids do not have to be parallel. I was the only user so far of those script so please ask me if...
by mathieu
Fri May 29, 2009 10:42 am
Forum: ROMS Problems
Topic: Perfect restart and 4dvar
Replies: 0
Views: 2440

Perfect restart and 4dvar

Is it possible to combine IS4DVAR with PERFECT_RESTART ? Apparently not. ROMS compiles correctly with those options but when running the error is NETCDF_PUT_FVAR_1D - error while inquiring ID for variable: ocean_time The reason I wanted to use perfect restart is for the AKt fields: when using GLS, t...
by mathieu
Fri Apr 24, 2009 12:45 pm
Forum: ROMS Problems
Topic: Perfect restart problem
Replies: 1
Views: 3181

Re: Perfect restart problem

I got it. In order to read the perfect restart file, it is necessary to use the option ANA_INITIAL . Maybe this should be said in the cppddefs.h , i.e. replace ** ANA_INITIAL use if analytical initial conditions ** by ** ANA_INITIAL use if analytical initial conditions or from perfect restart file **
by mathieu
Wed Apr 22, 2009 11:02 am
Forum: ROMS Problems
Topic: Perfect restart problem
Replies: 1
Views: 3181

Perfect restart problem

I understand that in order to have perfect restart we should select PERFECT_RESTART in the cppdefs.h file. I do obtain the restart file with the additional variables rzeta , AKt , AKs and others. But how can we read those variables? The crux of the matter appears to be in the get_state subroutine. T...
by mathieu
Wed Apr 22, 2009 7:56 am
Forum: ROMS Tools and Techniques
Topic: RNT libcxa.so.5
Replies: 5
Views: 14992

Re: RNT libcxa.so.5

Another method to avoid the libcxa problem is to compile with the option -static-libcxa.
by mathieu
Thu Apr 16, 2009 8:23 am
Forum: ROMS Usage
Topic: Help My first application
Replies: 16
Views: 11484

Re: Help My first application

Inside of roms the attributes south , east , north , west refer to the Grid . More precisely if the grid has size (eta_rho,xi_rho) , then: The points of the southern boundaries are all the points (1,iXi) for iXi between 1 and xi_rho . The points of the eastern boundary are all the points (iEta,xi_rh...
by mathieu
Wed Feb 25, 2009 11:49 am
Forum: ROMS Adjoint
Topic: which option can I use for vertical mixing in adjoint model?
Replies: 6
Views: 6524

Re: which option can I use for vertical mixing in adjoint model?

Would it be possible to have a similar option FORWARD_BULKFLUXES for the variables shflux, ssflux, sustr and svstr?
by mathieu
Mon Feb 23, 2009 9:13 am
Forum: ROMS Adjoint
Topic: which option can I use for vertical mixing in adjoint model?
Replies: 6
Views: 6524

Re: which option can I use for vertical mixing in adjoint model?

One possibility that you have is to select the option ANA_VMIX. You then have to edit ana_vmix.h to put the values of Akt, Akv that you want to use.
by mathieu
Thu Jan 22, 2009 9:23 am
Forum: ROMS Bugs
Topic: CORRELATION
Replies: 0
Views: 2161

CORRELATION

I tried to compute the covariance normalization factor file for the DOUBLE_GYRE application using the CORRELATION option as explained in s4dvar.in During the compilation, one gets fortcom: Error: mod_fourdvar.f90, line 335: This name does not have a type, and must have an explicit type. [INIVAL] DTs...
by mathieu
Fri Dec 19, 2008 10:39 am
Forum: ROMS Problems
Topic: ROMS and SWAN
Replies: 1
Views: 2194

ROMS and SWAN

I post the message here since there is no ROMS and SWAN forum. The file coupling_inlet_test.in give the impression that one can choose which variables are sent from roms to swan. That illusion is dispelled by the code section in waves_coupler.F : COMPDA(INDX,JWLV2)=TEMPMCT(IP,2) COMPDA(INDX,JVX2)=TE...
by mathieu
Mon Dec 01, 2008 11:02 am
Forum: ROMS Bugs
Topic: Closed boundary problem
Replies: 2
Views: 3576

Closed boundary problem

Suppose you run your model from initial state Initial.nc . Should the first record in the history file His_0001.nc coincide with Initial.nc ? I would say so but this is definitely not the case. For the sake of the argument, suppose that all boundaries of the rectangle (1:eta_rho,1:xi_rho) are closed...
by mathieu
Sun Nov 23, 2008 7:24 pm
Forum: ROMS Discussion
Topic: Question about "READ_WATER" option
Replies: 1
Views: 2391

Re: Question about "READ_WATER" option

The problem of masked grid is that much of the computational domain is taken by land points. In practice this means that if 30% of rho, u, v points are land then 30% of disk space is taken by numbers equal to zero or another filling variable. The idea of READ_WATER, WRITE_WATER is to write only the ...
by mathieu
Tue Nov 04, 2008 10:50 am
Forum: ROMS Discussion
Topic: "missing_value" for masked regions in NetCDF output files?
Replies: 16
Views: 12085

Re: "missing_value" for masked regions in NetCDF output files?

Having coherent filling values is good, but for us the real problem of the mask is the excessive use of disk space. The READ_WATER, WRITE_WATER options of ROMS allow to write only the sea points and are certainly nice. However, Hernan indicated me that the resulting files are not CF-compliant. What ...
by mathieu
Mon Nov 03, 2008 11:48 am
Forum: ROMS Bugs
Topic: excess smoothing with smth_bath tool
Replies: 2
Views: 4080

Re: excess smoothing with smth_bath tool

In general, I do not recommend to use the Shapiro filter. There are two problems, even for this modified version: ---There is a termination problem, i.e. for some bathymetries the filter does not converge to the right bathymetry. ---The perturbation to the bathymetry is large. There are some matlab ...
by mathieu
Fri Sep 19, 2008 1:02 pm
Forum: ROMS Bugs
Topic: NUDGING_SST bugs
Replies: 1
Views: 2751

NUDGING_SST bugs

In the file ROMS/Nonlinear/set_data.F line 1022, please replace cff1=Zr*cff+pcoef_SST(order,ng) by cff1=Zr*cff1+pcoef_SST(order,ng) In the file ROMS/Modules/mod_scalar.F please eliminate the variables assi, pert, Prms they are not used anywhere in the code and they are wrongly allocated on line 1635...
by mathieu
Mon Sep 01, 2008 9:17 am
Forum: ROMS Discussion
Topic: ROMS grid stiffnesses and consequences
Replies: 7
Views: 9135

Re: ROMS grid stiffnesses and consequences

No, I do not agree. The factor rx1 can be lower than 1. If the bathymetries are z_w(e,k) with e a wet cell and k a vertical level then rx1=max (z_w(e,k)-z_w(e',k) + z_w(e,k-1)-z_w(e',k-1))/(z_w(e,k)+z_w(e',k) - z_w(e,k-1)-z_w(e',k-1)) with e and e' two adjacent wet cells. If the bathymetry of the we...
by mathieu
Fri Aug 29, 2008 11:49 am
Forum: ROMS Discussion
Topic: ROMS grid stiffnesses and consequences
Replies: 7
Views: 9135

Re: ROMS grid stiffnesses and consequences

Suppose that you do not have any thermocline ( hc=0 ). Then the vertical levels of the models are h(e, k)=h(e) phi(k) with h(e) the bathymetry at the wet cell e and phi(k) the vertical parametrization. The Beckman and Haidvogel number is defined as rx0=max |h(e)-h(e')|/(h(e)+h(e')) with e and e' two...
by mathieu
Fri Jul 25, 2008 12:24 pm
Forum: ROMS Problems
Topic: SWAN coupling, with output
Replies: 6
Views: 5419

Re: SWAN coupling, with output

Since we are at SWAN and ROMS , let me point out a problem others might have encountered. It is necessary when running SWAN coupled with ROMS to output at least one field from the SWAN program, i.e. to have something like BLOCK 'COMPGRID' NOHEADER 'depth.mat' LAY 4 DEPTH 1. OUTPUT 20000101.000000 1 ...
by mathieu
Thu Jul 17, 2008 12:51 pm
Forum: ROMS Problems
Topic: How does ASSIMILATION_SST work?
Replies: 0
Views: 1931

How does ASSIMILATION_SST work?

The wiki says that the file oi_update.F is where computations of the option ASSIMILATION_SST should take place.
But there is nothing in this file, or any other file, which uses the collected data.
What is the right way to use this option?
by mathieu
Fri Nov 09, 2007 12:59 pm
Forum: ROMS Bugs
Topic: Tangent problem
Replies: 0
Views: 2469

Tangent problem

Dear all, I have a problem with ROMS 3.0 rev 126. When I used OPT_PERTURBATION with mpi , it works (up to the next problem in arpack) only for 1 processor. When NPROC>1 , the following happens: TL_WRT_HIS - error while writing variable: vbar depending on NPROC other variables than vbar show up as pr...
by mathieu
Fri May 25, 2007 12:09 pm
Forum: ROMS Discussion
Topic: Some problem with input files.
Replies: 8
Views: 7885

Solution to problem

ROMS requires the time dimension of the initial file to be unlimited.
The simplest way to create netcdf file with unlimuted dimension is, possibly,

Code: Select all

ncgen -o roms_init.nc roms_init.cdl
Then edit under matlab with

Code: Select all

nc=netcdf('rom_init.nc', 'write');
....
close(nc);
by mathieu
Thu May 24, 2007 6:58 am
Forum: ROMS Discussion
Topic: Some problem with input files.
Replies: 8
Views: 7885

Upwelling forcing

Looking at upwelling.h it appears that all necessary fields are defined by analyticals: #define ANA_GRID ... #define BS_FLUX When fields are available in analyticals and in netcdf files, ROMS takes the analytical. If you want to provide your own forcing fields, then you need to undef the correspondi...
by mathieu
Thu Apr 12, 2007 8:44 am
Forum: ROMS Problems
Topic: automated process
Replies: 2
Views: 3269

To run several command in sequence do ./FirstCommand && ./SecondCommand && ./ThirdCommand The advantage is that if an error occurs, then all stops. When things become more complicate, you create a file, say, ListComm #!/bin/sh ./firstCommand ./SecondCommand you put it as executable w...
by mathieu
Mon Mar 26, 2007 9:07 am
Forum: ROMS Source
Topic: History files
Replies: 2
Views: 6270

Yes this is ROMS rutgers and yes I did not explained myself correctly. What I would like is that the file ocean.in contain: NHISFILES == 2 HISNAMES == ocean_his_complete_6hour.in \ ocean_his_partial_30min.in Then the file ocean_his_complete_6hour.in would contain: NHIS == 54 NDEFHIS == 1728 ! 6 hour...
by mathieu
Fri Mar 23, 2007 9:24 am
Forum: ROMS Source
Topic: History files
Replies: 2
Views: 6270

History files

Dear all, I recently worked on running ROMS coupled with SWAN and I wanted to put the wave informations (Pwave, Dwave, etc.) as output in history file . This faciliity is not available with ROMS 2.2 since those variables were forcing informations put into the input file. What about having the abilit...
by mathieu
Fri Nov 24, 2006 12:50 pm
Forum: ROMS Problems
Topic: Bathymetry smoothing
Replies: 2
Views: 4698

Linear Programming smoothing

Thank you for the above explainations. I have a better understanding of
the problem.

The linear programming smoothing, which I mentionned earlier is available
from http://www.liga.ens.fr/~dutour/Bathymetry/index.html.
All critics are accepted.
by mathieu
Wed Nov 15, 2006 12:12 pm
Forum: ROMS Problems
Topic: Bathymetry smoothing
Replies: 2
Views: 4698

Bathymetry smoothing

Dear all, a rough bathymetry creates unstability of the model because of the terrain following coordinates. One solution is to use smoothing but I have some more questions below. (1) What are the means of measuring the roughness? The roughness factor , i.e. the maximum of |h1-h2| / (h1+h2) over adja...
by mathieu
Thu Nov 03, 2005 3:20 pm
Forum: ROMS Problems
Topic: Problem with mpich
Replies: 0
Views: 2548

Problem with mpich

Dear All, I tried to run ROMS on a HP cluster of 4 computers, every one of them having 2 Xeons in hyperthreading with MPI. The problem happens with the mp_exchange procedure. After doing mpi_irecv and mpi_send the procedure mpi_wait is called and the thread is stopped. However, only the thread of ra...
by mathieu
Mon Oct 24, 2005 2:20 pm
Forum: ROMS Installation
Topic: ROMS/TOMS 2.1 array size limitation problem
Replies: 3
Views: 6385

Solutions to the problem

Dear Lanerolle, I encounter exactly the same problem. I managed to pinpoint the bug to the following minimal program: PROGRAM BUGSEARCH CALL gls_corstep_tile(1044574) CONTAINS SUBROUTINE gls_corstep_tile(Jend) integer, intent(in) :: Jend integer, parameter :: r8 = selected_real_kind(12,300) real(r8)...