OQ model output files -

[Suresh]

Hi,

I have successfully ran OQ classical_psha_based_risk demo and got
results in computed_output folder in NRML format. Is there any
utilities that I can convert them to a readable format? I am a GIS guy
and would like to convert the output in GIS compatible format so that
I can visualize and present the results. Please let me know if you
have any ideas?

Regards,
Suresh

[Damiano Monelli]

Hi Suresh,

currently the only supported format for the OQ output is NRML. We are planning
to provide tools (i.e. scripts) for visualizing NRML output using GMT (http://gmt.soest.hawaii.edu/).
These tools will be available in the OATS accounts (hopefully by end of November).

As you point out, a converter to a GIS format would also be useful. We will work on this.
For the moment you have to produce a script to read the NRML and convert to a GIS format (e.g. shapefile)
by yourself.

I'm sorry we cannot provide you with these tools right now.

Regards,

Damiano

[Suresh]

Hi Damiano,

Thanks for your reply. I am trying to parse the NRML into TXT format,
so that I can migrate it to GIS platform.

I have observed a value1.0 adjacent to every X, Y coordinate when you
define a fault geometry representing linestring Ex:
<faultTrace><gml:LineString><gml:posList>19.5417 40.0925 1.0</
gml:posList></gml:LineString></faultTrace> in dissFaultModel.xml
file.

But it is a different number (in this example it is 15.55) when you
define complex fault geometry Ex:
<complexFaultGeometry><faultEdges><faultTopEdge>
<gml:LineString><gml:posList>-124.0415 40.347
15.55<gml:posList><gml:LineString></faultTopEdge></faultEdges> </
complexFaultGeometry>.

Can you please let me know what is the significance of this value?

Regards,
Suresh

[Damiano Monelli]

Hi Suresh,

in NRML the geometry of a 'simple fault' contains a 'faultTrace' element,
while for a 'complex fault' two elements must be provided: 'faultTopEdge' and 'faultBottomEdge'.

They are all defined as 'LineString'. A 'LineString' is defined as a 3D line, meaning that is
a sequence of points, where each point is defined by a triplet (longitude, latitude, depth).
Longitude and latitude are in decimal degrees, while depth is in km and assumed positive
downwards.

In your examples 1.0 and 15.5 represent depth values.

Hope it helps, let me know if more clarification is needed,

Regards,

Damiano

[Suresh Betapudi]

Hi Damino,

Thanks for the clarification.

Currently I am working on Sumatra faults (1 subduction; 1 crustal) to
generate Hazard curves for Singapore. As I see from the documentation,
GMPEs are hardcoded (user cannot define his own ground motion
attenuation functions), in such case I cannot use them.

If we provide 2 GMPE (from our research published papers ) to
implement in OQ and make available to run model, is it possible? If
so, what are the inputs we need to provide? ELSE is there any way I
can give GMPE inputs as a OQ user?

Regards,
Suresh

[Damiano Monelli]

Hi Suresh,

Currently GMPEs are hardcoded in OQ, so you can use only those that are implemented.
GMPEs that have been implemented can be found here:
https://github.com/gem/OpenSHA/tree/master/java/org/opensha/sha/imr/attenRelImpl

If you want to use a GMPE that is not among those provided, you have two options:

- you implement the GMPE by yourself (in OpenSHA lite, this requires Java programming).
I can give more details on that if this is an option for you.

- you can make a request to the GEM Model Facility to implement the GMPEs
you need by sending an e-mail to in...@openquake.org.

Hope it helps,

Damiano