generating scenario-based map realizations with spatial and cross correlation

[Anne Hulsey]

Hello,

I want to generate shaking intensity map realizations for a single scenario, including both spatial correlation and cross correlation between periods. It looks like previous efforts to add cross correlation have been closed (https://github.com/gem/oq-engine/issues/3607).

Is it possible to retrieve the median and inter/intra event standard deviations for each GMPE in the logic tree so that I can generate the correlated map realizations in post-processing? I know how to get the medians (set the truncation level=0) and found a non-site-specific inter-event standard deviation for each IM in the .hdf5's sigma_epsilon dataset. However, I haven't found anything for the intra-event standard deviations.

Thanks,

Anne

[Erika Sch]

Hello,

I have the same issue. 

Any suggestion about the intra-event std?

Thanks,

Erika

[Marco Pagani]

What you aim to do - at the moment - it can be done only with some python code and not directly with the OQ Engine.

--
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/1b688e71-07f3-4f4e-8a40-1ea2155d2aa4n%40googlegroups.com.

MARCO PAGANI | Seismic Hazard Team Lead | Skype mm.pagani | +39-0382-5169863

GLOBAL EARTHQUAKE MODEL | working together to assess risk

GEM - globalquakemodel.org | T - @GEMwrld | F - GEMwrld

[Anne Hulsey]

Thank you for the response, Marco. Just to confirm - you mean that both the ultimate aim of generating the cross-correlated gm fields and the intermediate step of retrieving the standard deviations cannot be done with the engine, right?

[Marco Pagani]

Regarding cross-correlated gm fields, not at the moment Anne. If some of you are willing to work on this we can help but at present we are focused on a number of improvements defined within a project e.g. CS spectrum, VPSHA.

Regarding the possibility of retrieving standard deviations, I will send you an example of python script that does it.

Marco

To view this discussion on the web visit https://groups.google.com/d/msgid/openquake-users/e57c3ccd-9a40-4bf2-b078-1cf63875ecf0n%40googlegroups.com.

[Anne Hulsey]

Thanks, Marco, that will be perfect. I already have my own post-processing script for generating the cross-correlated gm fields from previous work with OpenSHA.

Anne

[Anne Hulsey]

Hello Marco,

You had mentioned an example python script for retrieving standard deviations. I have again found myself needing them for an unrelated task. Could you pass on the example script, if it is available?

Thanks,

Anne

[Marco Pagani]

Hi Anne, sorry it slipped through the cracks. Terrible period in terms of workload.

Attached you find a jupyter python notebook and a script. They show how you can get total and between event std for one example GMM (Boore Atkinson, 2008). In order to use it for other GMMs please make sure that all the explanatory variables requested (as described in the header of each model implemented in the OQ Engine see for example https://github.com/gem/oq-engine/blob/master/openquake/hazardlib/gsim/boore_atkinson_2008.py#L59) are defined. For example, if a GMM uses a rrup site-rupture metric you must add it to the DistanceContext in the same way in the script it’s done for rjb. For a nomenclature of distance parameters you can take look here https://github.com/gem/oq-engine/blob/master/openquake/hazardlib/contexts.py#L80 and here for rupture https://github.com/gem/oq-engine/blob/master/openquake/hazardlib/contexts.py#L319.

If you are experiencing problems or there’s something not clear please just write.

Marco

To view this discussion on the web visit https://groups.google.com/d/msgid/openquake-users/3d11f9de-24a2-40a0-9ea2-493c9f175c33n%40googlegroups.com.

[Anne Hulsey]

Many thanks, Marco, that worked perfectly.