From lcyxj at mail.ustc.edu.cn Sat Aug 5 18:26:31 2017 From: lcyxj at mail.ustc.edu.cn (lcyxj at mail.ustc.edu.cn) Date: Sun, 6 Aug 2017 09:26:31 +0800 (GMT+08:00) Subject: [CIG-SEISMO] Question for using tomographic model in Latitude/Longitude coordinates Message-ID: <4da509f1.10e80.15db52696aa.Coremail.lcyxj@mail.ustc.edu.cn> Hello developers, I am a student from China and now want to use Specfem3d Cartesian to simulate the waveforms by using an existed tomographic model in Latitude/Longitude coordinates. First,I want to make sure if the Specfem3d Cartesian could work in this situation. If so , could you send me a simple example,since the existed examples in the package do not cover this situation except for a example in Cartesian coordinates.Thanks so much! Wish for you reply! Best! Cheng Li University of Technology and Science of China， Hefei,Anhui,China. -------------- next part -------------- An HTML attachment was scrubbed... URL: From komatitsch at lma.cnrs-mrs.fr Sun Aug 6 11:07:21 2017 From: komatitsch at lma.cnrs-mrs.fr (Dimitri Komatitsch) Date: Sun, 6 Aug 2017 20:07:21 +0200 Subject: [CIG-SEISMO] Question for using tomographic model in Latitude/Longitude coordinates In-Reply-To: <4da509f1.10e80.15db52696aa.Coremail.lcyxj@mail.ustc.edu.cn> References: <4da509f1.10e80.15db52696aa.Coremail.lcyxj@mail.ustc.edu.cn> Message-ID: <2344c22f-a57e-afdd-8205-ab1a45b4499e@lma.cnrs-mrs.fr> Dear Cheng Li, Thank you for your message. I do not think that can be done by default, but it is easy to do, using one of the following two options: - if your mesh honors the curvature of the Earth, then just modify the routine that reads the tomographic model in order to import it in latitude/longitude instead of UTM_x and UTM_y - if your mesh does not honor the curvature of the Earth (which is the classical case), i.e. your mesh corresponds to a UTM-projected grid, just use the routine provided in src/shared/utm_geo.f90 to convert your current latitude/longitude file to a UTM-projected version. Best regards, Dimitri. On 08/06/2017 03:26 AM, lcyxj at mail.ustc.edu.cn wrote: > Hello developers, > I am a student from China and now want to use Specfem3d Cartesian to > simulate the waveforms by using an existed tomographic model in > Latitude/Longitude coordinates. First,I want to make sure if the > Specfem3d Cartesian could work in this situation. If so , could you send > me a simple example,since the existed examples in the package do not > cover this situation except for a example in Cartesian > coordinates.Thanks so much! > Wish for you reply! > Best! > Cheng Li > > University of Technology and Science of China， > Hefei,Anhui,China. > > > > _______________________________________________ > CIG-SEISMO mailing list > CIG-SEISMO at geodynamics.org > http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-seismo > -- Dimitri Komatitsch, CNRS Research Director (DR CNRS) Laboratory of Mechanics and Acoustics, Marseille, France http://komatitsch.free.fr From shengfantasy at gmail.com Thu Aug 10 02:39:54 2017 From: shengfantasy at gmail.com (=?UTF-8?B?5a2r6YOB5Yud?=) Date: Thu, 10 Aug 2017 17:39:54 +0800 Subject: [CIG-SEISMO] Problem in Using Gmsh to build the mesh for SPECFEM3D Cartesian Message-ID: Dear CIG-SEISMO members, We want to use Gmsh to build the mesh in Taiwan region for SPECFEM3D Cartesian (version 2.0). First, we had done some simple tests with a cuboid mesh. (1) The first problem is how to make the mesh with the topography in Gmsh. (2) We cannot find "constants.h" in package. The file is mentioned at topography section in the manual. (3) How to change the horizontal grid size with depth changing? The grid size is 40mx40m at depth of 100m, and then moves to 2 km, the grid size transforms to 250mx250m something like that. Thank you very much Sincerely, Yu-Sheng Sun Department of Earth Sciences, NCU, Taiwan Special Assistant TEL: (03)422-7151 #65614 email: shengfantasy at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From shengfantasy at gmail.com Thu Aug 10 02:28:04 2017 From: shengfantasy at gmail.com (=?UTF-8?B?5a2r6YOB5Yud?=) Date: Thu, 10 Aug 2017 17:28:04 +0800 Subject: [CIG-SEISMO] Problem in Using Gmesh to build the mesh for SPECFEM3D Cartesian Message-ID: Dear CIG-SEISMO members, We want to use Gmesh to build the mesh in Taiwan region for SPECFEM3D Cartesian (version 2.0). First, we had done some simple tests with a cuboid mesh. (1) The first problem is how to make the mesh with the topography in Gmesh. (2) We cannot find "constants.h" in package. The file is mentioned at topography section in the manual. (3) How to change the horizontal grid size with depth changing? The grid size is 40mx40m at depth of 100m, and then moves to 2 km, the grid size transforms to 250mx250m something like that. Thank you very much Sincerely, Yu-Sheng Sun Department of Earth Sciences, NCU, Taiwan Special Assistant TEL: (03)422-7151 #65614 email: shengfantasy at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From csuzmc at 163.com Fri Aug 11 09:01:00 2017 From: csuzmc at 163.com (=?GBK?B?1cXD97LG?=) Date: Sat, 12 Aug 2017 00:01:00 +0800 (CST) Subject: [CIG-SEISMO] =?gbk?q?specfem2d_MPI_unstable=A3=BF?= Message-ID: <6b056c5f.db5e.15dd207004e.Coremail.csuzmc@163.com> Hi Dimitri I test specfem2d on serial simulation，it works well, When i change the parameter "nproc=4" in par_file, MPI occur error and exit , there are many NAN in the result. how can i fix this problem？ Best regards Mc -------------- next part -------------- An HTML attachment was scrubbed... URL: From komatitsch at lma.cnrs-mrs.fr Fri Aug 11 09:55:23 2017 From: komatitsch at lma.cnrs-mrs.fr (Dimitri Komatitsch) Date: Fri, 11 Aug 2017 18:55:23 +0200 Subject: [CIG-SEISMO] =?utf-8?q?specfem2d_MPI_unstable=EF=BC=9F?= In-Reply-To: <6b056c5f.db5e.15dd207004e.Coremail.csuzmc@163.com> References: <6b056c5f.db5e.15dd207004e.Coremail.csuzmc@163.com> Message-ID: Hi, It is probably MPI that is not correctly installed on your machine. You can try one of the examples in the "EXAMPLES" directory, they all work fine and thus if they do not work on your machine this will confirm that the problem comes from your local MPI installation. Otherwise you can also configure the code with "configure --enable-debug" to see if you get any useful message. Best regards, Dimitri. On 08/11/2017 06:01 PM, 张明财 wrote: > Hi Dimitri > > I test specfem2d on serial simulation，it works well, > > When i change the parameter "nproc=4" in par_file, MPI occur error and > exit , there are many NAN in the result. > > how can i fix this problem？ > > Best regards > > Mc > > > > > _______________________________________________ > CIG-SEISMO mailing list > CIG-SEISMO at geodynamics.org > http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-seismo > -- Dimitri Komatitsch, CNRS Research Director (DR CNRS) Laboratory of Mechanics and Acoustics, Marseille, France http://komatitsch.free.fr From justfornameonly at gmail.com Tue Aug 15 06:03:15 2017 From: justfornameonly at gmail.com (Gary) Date: Tue, 15 Aug 2017 09:03:15 -0400 Subject: [CIG-SEISMO] SPECFEM3D: ADJOINT example not working + missing files Message-ID: Dear developers, I am new to the specfem3d code and I am trying to get familiar with it by looking at the examples. I am interested in simulating a synthetic seismic attribute (e.g. impedance, RMS amplitude, instantaneous frequency, or time-shift) for an oil and gas reservoir, so I was looking at the BENCHMARK_CLAERBOUT_ADJOINT example, which I thought would be the closest to what I am interested in. However, I am not able to run this example for both acoustic and elastic cases, which give the following error when ran using $ ./run_inhomogeneous.bash > ./run_inhomogeneous.bash: line 42: $1: unbound variable Also, some of the python mesh creation files produce an error in Cubit because it needs cubit2specfem3d.py, and I believe this file is missing from the specfem3d repository. Can you please help? I have been looking for one example to simulate a seismic response of a synthetic oil and gas reservoir, so I will really appreciate your kind help. Thanks! Best regards, Gary -------------- next part -------------- An HTML attachment was scrubbed... URL: From komatitsch at lma.cnrs-mrs.fr Tue Aug 15 16:16:43 2017 From: komatitsch at lma.cnrs-mrs.fr (Dimitri Komatitsch) Date: Wed, 16 Aug 2017 01:16:43 +0200 Subject: [CIG-SEISMO] SPECFEM3D: ADJOINT example not working + missing files In-Reply-To: References: Message-ID: Dear Gary, Thanks! Let me cc Daniel Peter, since the problem you mention at line 42 of the scripts comes from this, which contains a comment from Daniel: #daniel do_setup=$1 The problem comes from the fact that $1 may be undefined when no command line argument is given to the script (i.e. in most cases); however, even when I remove that line, the rest of the script does not work either, it seems that the paths have changed. Best regards, Dimitri. On 08/15/2017 03:03 PM, Gary wrote: > Dear developers, > > I am new to the specfem3d code and I am trying to get familiar with it > by looking at the examples. I am interested in simulating a synthetic > seismic attribute (e.g. impedance, RMS amplitude, instantaneous > frequency, or time-shift) for an oil and gas reservoir, so I was looking > at the BENCHMARK_CLAERBOUT_ADJOINT example, which I thought would be the > closest to what I am interested in. However, I am not able to run this > example for both acoustic and elastic cases, which give the following > error when ran using $ ./run_inhomogeneous.bash > > ./run_inhomogeneous.bash: line 42: $1: unbound variable > > > Also, some of the python mesh creation files produce an error in Cubit > because it needs cubit2specfem3d.py, and I believe this file is missing > from the specfem3d repository. > > Can you please help? I have been looking for one example to simulate a > seismic response of a synthetic oil and gas reservoir, so I will really > appreciate your kind help. Thanks! > > Best regards, > Gary > > > _______________________________________________ > CIG-SEISMO mailing list > CIG-SEISMO at geodynamics.org > http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-seismo > -- Dimitri Komatitsch, CNRS Research Director (DR CNRS) Laboratory of Mechanics and Acoustics, Marseille, France http://komatitsch.free.fr