Surface does not conform with Aki & Richards convention~
225 views
Skip to first unread message
Rita
Sep 24, 2016, 7:56:23 PM9/24/16
to OpenQuake Users
Hi, all:
I used the oq-engine v2.0 to calcualte hazard wiht complex fault setitng. The engine show that "Surface does not conform with Aki & Richards convention".I cecked the fault geomerty which didn't show any unusal. My friend run the case with oq-engine v1.2, and finish running successfully. The case is attached. Is there any mistake I made in inputs for oq v2.0?
Rita
Graeme Weatherill
Sep 27, 2016, 3:41:50 AM9/27/16
to openqua...@googlegroups.com
Dear Rita,
Looking at the error and the configuration file you sent it seems as though the fault you are trying to build is dipping to the north, yet the order in which you input the nodes on each edge is going from west to east.
The "Aki & Richards convention" requires that the order in which the nodes on an edge must be read should be such that if I were to walk a path along those nodes the fault should be dipping to my right. The complex fault source has a further check that each edge going down dip should be placed so that it is on the down-dip side of the previous edge. Basically, you need to reverse the order in which the nodes are input for each edge (i.e. enter them from west to east in your example) and the problem should go away (hopefully).
Much earlier versions of the engine did not have this check, and although it was possible to build and float the ruptures on the surface the resulting distance calculations would be erroneous if the user did not follow the convention, which could be a tricky bug for the user to track down! Now we enforce the correct configuration.
I hope that solves the problem, but let us know if you encounter any further issues.
Thanks,
Graeme
Looking at the error and the configuration file you sent it seems as though the fault you are trying to build is dipping to the north, yet the order in which you input the nodes on each edge is going from west to east.
The "Aki & Richards convention" requires that the order in which the nodes on an edge must be read should be such that if I were to walk a path along those nodes the fault should be dipping to my right. The complex fault source has a further check that each edge going down dip should be placed so that it is on the down-dip side of the previous edge. Basically, you need to reverse the order in which the nodes are input for each edge (i.e. enter them from west to east in your example) and the problem should go away (hopefully).
Much earlier versions of the engine did not have this check, and although it was possible to build and float the ruptures on the surface the resulting distance calculations would be erroneous if the user did not follow the convention, which could be a tricky bug for the user to track down! Now we enforce the correct configuration.
I hope that solves the problem, but let us know if you encounter any further issues.
Thanks,
Graeme
--
You received this message because you are subscribed to the Google Groups "OpenQuake Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openquake-use...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Rita
Sep 28, 2016, 8:31:06 PM9/28/16
to OpenQuake Users
Dear Graeme,
Graeme Weatherill於 2016年9月27日星期二 UTC+8下午3時41分50秒寫道：
Thank you for your answer.
The complex fault depicts the whole fault geometry, I supposed that I don't need to consider the "Aki & Richards convention" in addition.
Now I have considered the convention in complex fault setting. The calculation works well.
Thank you very much!!
Rita
Graeme Weatherill於 2016年9月27日星期二 UTC+8下午3時41分50秒寫道：
Reply all
Reply to author
Forward
0 new messages