[aspect-devel] Using seismic tomography as an input

Wed Oct 25 12:44:38 PDT 2017

Hi Nuno,

> I would like to use a Cartesian rectangular box 60*60*13 km (0.5 km
> steps) oceanic Lithospheric/crustal P-velocity 3D model of a section of
> an oceanic ridge to calculate
>
>> 1st and mainly
>
> its gravimetric anomaly, to be compared with ship borne recorded
> gravimetric anomaly.

Unfortunately, this will likely also be the most complicated part of 
your model. The currently available 'geoid' postprocessor in ASPECT only 
allows to compute the gravity anomaly (in terms of perturbation of the 
geoid) for a global spherical model, because it relies on a 
decomposition of the gravity and topography contributions into spherical 
harmonics. I am not particularly familiar with gravity fields in 
cartesian models, have you already thought about how that would work, or 
are you looking for advice on the topic?

Best,

Rene

On 10/25/2017 11:49 AM, John Naliboff wrote:
> Hi Nuno,
>
> Thanks for the email and specific details on what you would like to 
> do.  A few comments below.
>
> So the idea is to convert the crustal-scale tomography model 
> (mid-ocean ridge) to a density structure and then input this into 
> Aspect in order to calculate the Geoid, heat flux density and 
> additional gravity anomalies. Certainly feasible and as mentioned 
> earlier ASPECT already has tools to calculate the Geoid and heat flux. 
> You would have to add new tools for calculating additional ‘gravity 
> anomalies’, but it sounds like you already have experience with this 
> type of calculation.
>
> A few things to consider:
>
> 1. Converting seismic velocity to density: If heat flux is a desired 
> output, then you will be running a thermal-mechanical simulation. In 
> this case, you would need to break the seismic velocity into 
> compositional and thermal contributions. These separate contributions 
> would then serve as the ‘initial conditions’ in Aspect. My 
> recommendation would be to input the initial composition and 
> temperature field through ASCII data files, which you prepare 
> independently. It sounds like you already have a clear idea of how to 
> convert the seismic velocities into a density structure:
>> I extract the contours that define
>> the velocity ranges that define discrete rheological layers using a
>> seismic velocity-density coefficient
>
> 2. You mentioned that
>> a big portion of its signal is caused by the interface between the 
>> water column and the seafloor
> You can include initial topography (bathymetry) on the top boundary 
> (free surface) in ASPECT. An alternative approach is to have a 
> free-slip top boundary (flat), but include a layer for water 
> (low-density and low-viscosity). I’m not sure if either approach is 
> advantageous for your intended purposes. If deeper lithospheric and 
> upper mantle densities anomalies are not included, you may get large 
> isostatic adjustments?
>
>> I would like to use a Cartesian rectangular box 60*60*13 km (0.5 km
>> steps) oceanic Lithospheric/crustal P-velocity 3D 
>
> 3. Assuming you have a similar grid spacing (0.5 km) in your numerical 
> model, this will give 374400 elements. The exact model size will 
> depend on the type of element you choose and number of compositional 
> fields, but a model of this size will likely need to be run across a 
> few hundred cores.
>
> I recommend starting out with a coarser grid (1-2 km) before jumping 
> to higher resolution models. You can also start out by only modeling a 
> small portion of the domain (ex: 10*10*13 km). AMR will also help 
> reduce your total model size, but these are all steps that that should 
> be taken one at a time. In other words, always best to slowly build 
> complexity before jumping straight into large simulations!
>
> I hope this helps and others may chime in with more helpful advice as 
> well. Please let us know as additional questions arise!
>
> Cheers,
> John
>
>
> *************************************************
> John Naliboff
> Assistant Project Scientist, CIG
> Earth & Planetary Sciences Dept., UC Davis
>
>
>
>
>
>
>> On Oct 25, 2017, at 1:50 AM, Nuno Mendes Simão 
>> <nuno.m.simao at durham.ac.uk <mailto:nuno.m.simao at durham.ac.uk>> wrote:
>>
>> Hi Everyone
>>
>> To start a thank you to Phil Heron to introduce me to the exciting world
>> of aspect and to you all for any help.
>>
>> So...
>>
>> I would like to use a Cartesian rectangular box 60*60*13 km (0.5 km
>> steps) oceanic Lithospheric/crustal P-velocity 3D model of a section of
>> an oceanic ridge to calculate
>>
>>> 1st and mainly
>>
>> its gravimetric anomaly, to be compared with ship borne recorded
>> gravimetric anomaly.
>>
>> I understand that the bathymetry is needed to calculate the full
>> gravimetric anomaly, as a big portion of its signal is caused by the
>> interface between the water column and the seafloor. If the use of
>> bathymetry makes things difficult it would still be great to have a tool
>> that only calculates the geopotential anomaly caused by the tomographic
>> Cartesian box defined above. It would be straightforward for me later to
>> subtract the effect of the water column - seafloor from the gravimetric
>> observations to make the comparisons.
>>
>>> and 2nd if possible
>> its Heat flux density
>>
>>> As a note.
>>
>> Normally when I do this in 2d cases I extract the contours that define
>> the velocity ranges that define discrete rheological layers using a
>> seismic velocity-density coefficient for each layer and use the method
>> of Talwani to calculate the geopotential anomalies.
>>
>> Nuno
>>
>>
>>
>>
>> -------- Forwarded Message --------
>> Subject: Fw: [aspect-devel] Using seismic tomography as an input
>> Date: Fri, 20 Oct 2017 22:36:00 +0100
>> From: HERON, PHILIP J. <philip.j.heron at durham.ac.uk 
>> <mailto:philip.j.heron at durham.ac.uk>>
>> To: MENDES-SIMAO, NUNO MIGUEL <nuno.m.simao at durham.ac.uk 
>> <mailto:nuno.m.simao at durham.ac.uk>>
>>
>>
>>
>>
>>
>> Philip J. Heron
>> Junior Research Fellow
>> Dept. of Earth Sciences
>> Durham University
>> web: http://philheron.com
>>
>> ------------------------------------------------------------------------
>> *From:* Aspect-devel <aspect-devel-bounces at geodynamics.org 
>> <mailto:aspect-devel-bounces at geodynamics.org>> on behalf of
>> John Naliboff <jbnaliboff at ucdavis.edu <mailto:jbnaliboff at ucdavis.edu>>
>> *Sent:* 20 October 2017 17:53:23
>> *To:* aspect-devel at geodynamics.org <mailto:aspect-devel at geodynamics.org>
>> *Subject:* Re: [aspect-devel] Using seismic tomography as an input
>>
>> Hi Phil,
>>
>> There are currently postprocessors available for computing the geoid and
>> heat flux. Additional postprocessors for gravity are in the works.
>>
>> Working examples - yes, please have your colleague send an email when
>> they have a concrete idea of what they would like to do. If they simply
>> want to map velocity variations into temperature anomalies that are used
>> as an initial condition, this should be fairly straightforward.
>>
>> Cheers,
>> John
>>
>> *************************************************
>> John Naliboff
>> Assistant Project Scientist, CIG
>> Earth & Planetary Sciences Dept., UC Davis
>>
>> On 10/20/2017 01:37 AM, HERON, PHILIP J. wrote:
>>>
>>> Hi Paul and John,
>>>
>>>
>>> Thanks for the replies!
>>>
>>>
>>> "are you hoping to input seismic tomography models for the express
>>> purpose of obtaining a density and gravity model?"
>>>
>>>
>>> Yes, this is what he is looking to do. Heat flux would be a bonus, too.
>>>
>>>
>>> I've passed on the information and he is going to download Aspect this
>>> afternoon. I think this would be really neat if we can get a working
>>> example up and running from this. I'll keep you posted and I (or more
>>> than likely he) will drop back in with some questions.
>>>
>>>
>>> Thanks again,
>>>
>>>
>>> Phil
>>>
>>> ------------------------------------------------------------------------
>>> *From:* Aspect-devel <aspect-devel-bounces at geodynamics.org 
>>> <mailto:aspect-devel-bounces at geodynamics.org>> on behalf
>>> of Bremner,Paul M <pbremner at ufl.edu <mailto:pbremner at ufl.edu>>
>>> *Sent:* 19 October 2017 22:04:30
>>> *To:* aspect-devel at geodynamics.org <mailto:aspect-devel at geodynamics.org>
>>> *Subject:* Re: [aspect-devel] Using seismic tomography as an input
>>>
>>>
>>> Hi Phil,
>>>
>>>
>>> I believe a 3D Cartesian model can be input into the ascii plug-in.
>>>
>>>
>>> I also just want to make sure I understand your purpose correctly, are
>>> you hoping to input seismic tomography models for the express purpose
>>> of obtaining a density and gravity model? Or are you wanting to simply
>>> use the tomography models to feed a density model to ASPECT?
>>>
>>>
>>> I believe that both the S40RTS and ascii plug-ins assign the
>>> temperature field in ASPECT, and not the density. For example, in
>>> S40RTS the velocity perturbation at a particular point is converted to
>>> density perturbation by multiplying a constant scale factor. That
>>> value is then multiplied by 1/alpha to get temperature perturbation,
>>> which is added to a background temperature for that point.
>>>
>>>
>>> Cheers,
>>> Paul
>>> ------------------------------
>>>
>>> Subject: Digest Footer
>>>
>>> _______________________________________________
>>> Aspect-devel mailing list
>>> Aspect-devel at geodynamics.org <mailto:Aspect-devel at geodynamics.org>
>>> https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.geodynamics.org_cgi-2Dbin_mailman_listinfo_aspect-2Ddevel&d=DwIGaQ&c=pZJPUDQ3SB9JplYbifm4nt2lEVG5pWx2KikqINpWlZM&r=Wl19MSBMb4DZXdORUmp5jA&m=IlyCRz3zvOcycd_IMQiBKL7HsdUkdvb1HQwsaicq2-Q&s=iQC22OjEmDV_vdq0wXaZKucakl7GNkgA7aySC5tDJaE&e=
>>>
>>>
>>> ------------------------------
>>>
>>> End of Aspect-devel Digest, Vol 71, Issue 4
>>> *******************************************
>>>
>>>
>>> _______________________________________________
>>> Aspect-devel mailing list
>>> Aspect-devel at geodynamics.org
>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>>
>> _______________________________________________
>> Aspect-devel mailing list
>> Aspect-devel at geodynamics.org <mailto:Aspect-devel at geodynamics.org>
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel_______________________________________________
>> Aspect-devel mailing list
>> Aspect-devel at geodynamics.org
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>
>
>
> _______________________________________________
> Aspect-devel mailing list
> Aspect-devel at geodynamics.org
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel

-- 
Rene Gassmoeller
http://www.math.colostate.edu/~gassmoel/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geodynamics.org/pipermail/aspect-devel/attachments/20171025/0f381847/attachment-0001.html>