Re: [swanmodel-users] Questions about the couple model of swan+adcirc
Brought to you by:
mzijlema
Menu
▾
▴
Re: [swanmodel-users] Questions about the couple model of swan+adcirc
|
From: Casey D. <cas...@gm...> - 2012-04-14 02:22:43
|
Du Bingru It is good to hear from you. To your questions: 1 and 2. There doesn't appear to be anything wrong with your mesh file (fort.14). However, it may depend on which version of SWAN you are using. For unstructured meshes in the ADCIRC/SMS format, the boundaries are numbered sequentially as they are read from the mesh file. The open boundaries are numbered starting at 1, and then the land boundaries continue the numbering. (In a parallel run, the global numbering is carried forward automatically to the local mesh files.) So, if you want to apply an incoming wave boundary condition at the first open boundary in the mesh file, then you can use the BOUN SIDE command with K=1. If the open boundary is the fourth boundary listed in the mesh file, then you would use K=4, and so on. The early implementations of SWAN for unstructured meshes required this boundary numbering to be included in the mesh file. This numbering was an additional quantity, beyond what is required by ADCIRC. SWAN has since been changed to do the numbering automatically, so that it can use the same exact meshes as ADCIRC. I think this may be the error that you are seeing. Your version of SWAN is probably expecting the boundaries to be numbered in the mesh file. When I run your files with the latest SWAN+ADCIRC, I do not see the error message. 3. To turn on the coupling from SWAN to ADCIRC, you need to set the hundreds digit of the NWS parameter to 3 in the ADCIRC parameters file (fort.15). For example, if you are reading ATCF-formatted "best track" files to develop the wind field in ADCIRC and share it with SWAN, then NWS would be set to 308. The "3" tells ADCIRC to receive radiation stress forcing from the coupled SWAN, and the "08" tells ADCIRC to read and develop the wind field from this file format. More information is available on the ADCIRC User Manual: http://www.adcirc.org/documentv49/fort_15.html http://www.adcirc.org/documentv49/parameter_defs.html#NWS and in this online guide: http://www.caseydietrich.com/swanadcirc/ http://www.caseydietrich.com/2009/08/26/how-to-compile-and-run-swanadcirc/ 4. Our computational meshes often contain a similar variation in element sizes. With elements as small as 15m, it is not surprising that ADCIRC becomes unstable with a 5s time step. ADCIRC is conditionally stable due to its semi-implicit solution method as well as the propagation speeds of any wet/dry fronts. I recommend decreasing the time step to 1s and trying the simulation again. I hope this helps. Casey Dietrich die...@ic... 2012/4/12 杜冰茹 <159...@qq...> > > Hi, > I am trying to simulate the storm surge using the SWAN+ADCIRC model and it > seems a tough work for me. Here i list my questions and send my grid file > and model control file to you. I really need your help. > 1.Why the swan can't read the fort.14(unstructured grid) file? it > print "Terminating error: error reading data from grid file fort.14" in the > Errfile. > 2. I have edited the swan input file follow yours fort.26 file. But i > confuse that how to impose spectra at boundaries. is it appropriate to set > hs=5m, period=20s for the open boundary lied in deep sea? I use the command > SIDE but don't how to set k=? I seted that below in the input file. > BOU SHAPE JON PEAK DSPR POWER > BOU SIDE 2 CCW CON PAR 5.0 20 270 2 > 3. Don't I need to output any data such as force data (radial stress ) for > the ADCIRC compute when I use the couple model SWAN+ADCIRC? > 4. As you see my grid file fort.14, you will find that it contain a large > domain and the grid step change from 15m to 10000m. So it is not convergent > in the small grid areas when my time step set to 5s. It will take a long > time if the time step is set shorter to simulate pure tidal or storm surge > for more than ten days. Do you have a good way to solve the problem? I try > to nest the domain first but fail because the computational grid points in > the boundary of small domain were not according to the computational grid > points in the big domain. So it can't not impose the water level to the > boundary of small domain. > looking for your reply and thanks again. > Best, > Du Bingru > ________________________________ > 从QQ邮箱发来的超大附件 > fort.15 (144KB, 2012年05月12日 00:43 到期) > > 进入下载页面：http://mail.qq.com/cgi-bin/ftnExs_download?k=5e373233758f9e994b24280147315149530f0557520854034f02070b561c575656511f0a5552014b075205510309020753015751613663000d45461d5004635b&temp=b723a1cf&t=exs_ftn_download&code=b723a1cf&s=email > INPUT (1KB, 2012年05月12日 00:43 到期) > > 进入下载页面：http://mail.qq.com/cgi-bin/ftnExs_download?k=09326634b098c8c81a217c0610630518500404510155000e1e0a5751054e030e520a4b560200551a065707575754510f040a54513666377e7d623360365e&temp=32f46c77&t=exs_ftn_download&code=32f46c77&s=email > fort.14 (43MB, 2012年05月12日 00:56 到期) > > 进入下载页面：http://mail.qq.com/cgi-bin/ftnExs_download?k=0963323427cd7d9e1c7028061f66074c535a540501505752185251035a4b015506551f0d0100004e575b57015d53015703515152396135055a11461a0852355e&temp=5c249f5c&t=exs_ftn_download&code=5c249f5c&s=email > > > ------------------------------------------------------------------------------ > For Developers, A Lot Can Happen In A Second. > Boundary is the first to Know...and Tell You. > Monitor Your Applications in Ultra-Fine Resolution. Try it FREE! > http://p.sf.net/sfu/Boundary-d2dvs2 > _______________________________________________ > swanmodel-users mailing list > swa...@li... > https://lists.sourceforge.net/lists/listinfo/swanmodel-users > |
