Help on Gutenberg-Richter distribution in OpenQuake

[Mehdi Sadeghi]

Dear all,

Hello

I need to know the definition of Gutenberg-Richter distribution in OpenQuake.

Is the number of earthquakes (N) in following equation defined as [Event/Year/(Length or Area)] or [Event/Year]?

Log N=a - bM

 

Best,

Mehdi

[Marlon Pirchiner]

Hi Mehdi,

May someone else could answer you better. Anyway

You can find some MFD used by OQ here:
https://github.com/gem/oq-hazardlib/tree/master/openquake/hazardlib/mfd
Are available a truncated-GR and a discretized MFD, and you need just parametrize yours MFD for each seismic source that you have.

In addition, may you are starting from some catalog, so you can find here:
https://github.com/GEMScienceTools/hmtk/tree/master/hmtk/seismicity/occurrence
some tools to model your seismicity occurrence. For a quickstart, please check this notebook:
http://nbviewer.ipython.org/github/GEMScienceTools/notebooks/blob/master/hmtk/Recurrence%20and%20Mmax.ipynb

I hope it helps,

good luck,

marlon

[Marlon Pirchiner]

Hi Mehdi,

I’m answering here because our discussion could help someone else in the furure.
Even more, as I said, I don’t know a lot OQ and may someone else could explain you better…

On Apr 17, 2014, at 9:48, Mehdi Sadeghi wrote:

Dear Marlon

Thank you for your explanation. I looked at the websites you sent but I need to know if I want to use a classic way to calculate a and b value, I have two options:

1st plot log(Number of earthquakes/Years) vs (Moment Magnitude) then find slope of line as b-value and distance from axis as a-value.

2nd plot log(Number of earthquakes/Years/Length or Area) vs (Moment Magnitude) then find slope of line as b-value and distance from axis as a-value.

Before anything, you need to realize that exist many ways and proposed forms to model the MFD, more than N(m > m_i) = 10**(a - b*m_i). For instance, you can taper or extend the end part distribution for higher magnitudes, or you can truncate your distribution from some M0 to some Mmax that you had estimate.

Which one is defined in OpenQukae?

It's not exactly what you are asking, but could be useful to understand how OQ use its values:
https://groups.google.com/forum/#!searchin/openquake-users/a$20value/openquake-users/BfPPhlcgkwI/hTLWiKT72OQJ

Does OpenQuake automatically calculate length or area of the source or I have to calculate it by myself?

Good question, I don’t know it. It could be related with the kind of source you’re using. Far as I know to area sources you provide the polygon full geometry and its uniform MFD, so the calculation have all needed to compute area. Even more, area sources are discretised and considered as a source point set (each one if their own MFD).
It’s a good question to know if the area MFD is 'distributed' to each discretised point or just 'replicated'.

Regarding to my catalogue I don’t know which sources cause specific earthquake so I have to consider all the catalogue data for each sources, in this case what should I do?

This seems to me like an modelling issue, that you need to decide by yourself. OQ allows you to represent your seismic sources in a couple ways… if you have an well known fault, for instance, you can use an ‘simple' or a more 'complex fault' seismic source and give it their known parametrisation. If you don’t known any specific fault, you can define by your own criteria some 'area sources’ which will be consider as having an uniform seismicity characterised by its parameters, one of them will have some MFD  as I did mention. But, in addition, you can use some smoothing method to define a grid of ‘point sources’ with their MFD as well.

Thank you for your time in advance,

Mehdi

Do not worry, I hope that helps.

Good luck at all!
marlon


                                                                                             

Mehdi Sadeghi

Faculty Member

Department of Civil Engineering

Fars Science and Research Branch, IAU

Shiraz, Iran

Ph.D. Student, Earthquake and Structural Engineering

Department of Structural Engineering

Science and Research Branch, IAU

Tehran, Iran

On Wednesday, April 16, 2014 8:00:31 AM UTC-3, Mehdi Sadeghi wrote:

[Mehdi Sadeghi]

Dear Marlon,

Thanks again for your time and explanation and I hope someone else can help!

 

As I mentioned, I don’t have any problem in the concept of calculating a and b values, I just need to know if OpenQuake engine consider the length/area of line/area sources in its calculation or not?

 

Best,

Mehdi

[Marlon Pirchiner]

Mehdi,

I read a bit of hazardlib code, which transform the sources geometries and parameters into ruptures surfaces to perform the probabilistic calculations and there I could check that YES, area, points and simple faults (I guess that is the same for complex faults) consider their geometries in order to calculate the ruptures occurrence rates.

For point sources, the rate_scale_factor is 1. For area, it is divided by the number of point sources generated by discretisation, and for the simple faults "The occurrence rate of each of those ruptures is the magnitude occurrence rate divided by the number of ruptures that can be placed in a fault." in the same words of iter_ruptures method.

I hope this help you, if you still do not solve this question.

Cheers,

Marlon

[Mehdi Sadeghi]

Dear Marlon and Community,

 

With having one earthquake catalogue for all region we should define the same a and b value for all faults and OpenQuake will consider their geometrics in order to calculate the ruptures occurrence rates.

But now, there is another problem, with same a and b values I got very large numbers for PGA and maybe there is something wrong somewhere else which I could not find it.

Would you please take a look at my source file and job.ini.

 

Best Regards,

Mehdi

                                                                                              

Mehdi Sadeghi

Faculty Member

Department of Civil Engineering 

Fars Science and Research Branch, IAU

Shiraz, Iran

Ph.D. Student, Earthquake and Structural Engineering

Department of Structural Engineering

Science and Research Branch, IAU

Tehran, Iran

--
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.