[aspect-devel] Progress in writing the mantle convection code Aspect

[Eric Heien]

We're quantifying them for an AGU poster so we'll have more information then.  If I recall correctly, they are on the order of 5-10% of the gradient (so for a gradient of 10^4, the overshoot would be on the order of 10^2 to 10^3), versus up to 20-30% for CitcomS.  Finer resolution will make them more localized but generally worse in magnitude.

Their affect on the simulation will depend on the material properties.  For the simple material model in Aspect the only difference is a change in density and thus evolution of the system.  For a more complex rheology I'm not sure what the effect would be.

-Eric

On Oct 16, 2013, at 11:57 AM, Magali Billen wrote:

> Hmm? Are they improved with better resolution? What is there magnitude?
> 
> For a density or compositional field, its okay to have very local smearing between fields, as this is probably
> more realistic anyway, but if you have over/undershoots that become unstable  (for temperature in particular)
> then that will ruin the simulation.
> -Magali
> 
> 
> On Oct 16, 2013, at 11:50 AM, Eric Heien wrote:
> 
>> FYI, the Aspect advection scheme still has over/undershoots.  Not as severe as CitcomS but they definitely appear and affect the simulation.
>> 
>> -Eric
>> 
>> On Oct 16, 2013, at 11:43 AM, Magali Billen wrote:
>> 
>>> Hi Katrina,
>>> 
>>> I agree with Wolfgang that using active fields is a better choice for this (sorry - one of the
>>> many things we need to be talking about as you get more into ASPECT). I talked with
>>> Wolfgang about this at some meeting (?) a couple years ago.  Using active fields will
>>> be much less costly computationally than using tracer particles.  In CitcomS, we can't
>>> using active fields because the advection scheme that is there for temperature (which
>>> one might copy for active fields) leads to over/undershoots at the field boundaries 
>>> when you have a non-diffusive field.  The scheme in ASPECT is quite different and my
>>> understanding is that it doesn't have this problem.
>>> 
>>> Magali
>>> 
>>> On Oct 15, 2013, at 11:13 AM, Wolfgang Bangerth wrote:
>>> 
>>>> 
>>>> Hi Katrina,
>>>> 
>>>>> I'm hoping to have active particles (or fields?) that define layers
>>>>> within a plate, with assigned densities that affect the model. In
>>>>> CitcomS, I model bouyant basaltic crust, harzburgite and dense eclogite
>>>>> layers that move with the subducting plate and affect its total bouyancy
>>>>> over periods of >30 Ma. I hope that helps.
>>>> 
>>>> Interesting.
>>>> 
>>>> This is already possible using active fields. Take a look at the cookbook section in the model -- it has an example of an active field that is used to model the density as a function of composition.
>>>> 
>>>> Best
>>>> W.
>>>> 
>>>> 
>>>> -- 
>>>> ------------------------------------------------------------------------
>>>> Wolfgang Bangerth               email:            bangerth at math.tamu.edu
>>>>                             www: http://www.math.tamu.edu/~bangerth/
>>>> 
>>>> _______________________________________________
>>>> Aspect-devel mailing list
>>>> Aspect-devel at geodynamics.org
>>>> http://geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>>> 
>>> _______________________________________________
>>> Aspect-devel mailing list
>>> Aspect-devel at geodynamics.org
>>> http://geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>> 
>