Nodal plane distributions

[ozka...@gmail.com]

Dear Developers,

I have small question about the logic of nodal plane distributions. OQ allows to define nodal planes as distributions. For example, I can model an area source with rake angle 0 degree with probability of 0.8 and another rake angle 90 degree with probability of 0.2. Let's assume I obtain the hazard result for this source.

In another calculation, I model the same source as splitting into two different sources. First one is modeled by rake angle 0 degree with probability of 1.0 and second one is modeled by another rake angle 90 degree with probability of 1.0. I calculate the hazard also for this case but I combine the hazard results of these two sources with probabilities of 0.80 and 0.20.

Should the results of these two analysis be same?

Best regards.

[Marco Pagani]

Dear Ozkan,

Can you specify which results you expect should or should not be the same?

Best,

Marco

--
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.

[ozka...@gmail.com]

Dear Marco,

The resultant hazard curves for the first (one source with Pr's of 0.80 and 0.20) and the second (two source with Pr's of 1.0 and 1.0 but, in this case, the resultant hazard curve is obtained from combining the hazard curves with Pr's 0.80 and 0.20) calculations.

Best regards.

[Marco Pagani]

If I understand correctly, in the second case you are modelling uncertainty as epistemic while in the first one as aleatory. In the second case are you using a logic tree structure (i.e. is the OQ-engine the one computing the final results)?

[ozka...@gmail.com]

Yes Marco. You have understood correctly.