output units

[Bruno Collaço]

Hey OQ Users,

All the outputs are in g, right?  or it depends on the chosen GMPE? I'm trying to use Douglas gmpes but getting very high accelerations...

Thanks id advance.

[Marco Pagani]

Output accelerations in g and velocities in cm/s

--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/openquake-users/041fc9a1-c353-4966-8ab2-4254f32e2f4an%40googlegroups.com.

[Hasib Abdulrazaq]

hi Marco 

the output accelerations from the DouglasEtAl2024 gmpe seems to be in m/s2. are all the gmpe's supposed to be in consistent units? I can see this being a risk unless clearly stated in documentation the units. It will be good to fix this to g if all other GMPE outputs are in g.

Best

Hasib

[Christopher Brooks]

Hi Hasib,

Could you give an approximate value range for the high intensity levels you are getting from this model? Also, perhaps some context regarding the magnitude-distance combination and site conditions you are computing for? This way we can check some things and correct the implementation if required.

Thanks,
Christopher

[Hasib Abdulrazaq]

Hi Chris

the following are PGA intensities using GEM implementation of the middle branch of Douglas Et Al. (2024) (+sigma model) for site condition Vs= 900m/s, distance=600km from central England. values are very high (1g for 10,000 return period). I don't see any conversion to g in the gmpe python files. 

Best

[Hasib Abdulrazaq]

[Christopher Brooks]

Hi Hasib,

The units issue has now been corrected in master of the OpenQuake Engine (https://github.com/gem/oq-engine/pull/10854). If you are using master branch, you can just pull from the repo to integrate this fix into your local OpenQuake Engine files. If you are using a fixed version of the engine (e.g. long term support (LTS) version 3.23.3), you can download the master version of the file (https://github.com/gem/oq-engine/blob/master/openquake/hazardlib/gsim/douglas_et_al_2024.py) and replace the equivalent file locally.

This should resolve the disparity between your own implementation and the OQ implementation.

Thanks,

Christopher