[Mesa-users] Question about RSP and equilibrium quantities

Tue Jan 21 16:22:14 EST 2020

How to correctly compare observations to models is interesting and
difficult problem, and I don't have an easy answer. Radius for pulsating
star can be determined observationally by various methods and for the same
star you may get differing results as various methods may probe different
depths of the atmosphere. To make your comparison meaningful, you have to
know how the observational determination was done and then try to do the
same kind of averaging with the model. Simple averaging over data points
doesn't seem the best possibility anyway. You may try fitting Fourier
series and taking zero point as an average. This method should also give
you consistent cycle-to-cycle results even for sparse time sampling.
Radek

wt., 21 sty 2020 o 21:52 Jeremy Sakstein <sakstein at hawaii.edu> napisał(a):

> Thanks Radek,
>
> That's very useful! Would you say that the average that I calculated by
> averaging over the data points is a good representation of the "equilibrium
> radius" or is such a quantity not meaningful?
>
> I ask because papers purport to have measured this radius using other
> means e.g. V-K calibrations but it is not clear to me what such a quantity
> corresponds to if the radius is varying by 4 solar radii.
>
> On Tue, Jan 21, 2020 at 10:44 AM Radek Smolec <smolec at camk.edu.pl> wrote:
>
>> Such procedure would work provided light/radius curve is very densely
>> sampled. Here it is not the case.  RSP time steps are distributed
>> regularly, but not synchronised with pulsation phase (i.e. we don't have
>> time stamps at the same pulsation phases of the consecutive pulsation
>> cycles). It is  visible in your plots with colored points. If I see it
>> correctly in the (strong) zoom, black point overlap with blue. Hence you
>> have two representations for consecutive pulsation cycles with differently
>> distributed time stamps. It may be enough to produce small, alternating
>> difference in the average.
>> Radek
>>
>>
>> wt., 21 sty 2020 o 21:11 Jeremy Sakstein <sakstein at hawaii.edu>
>> napisał(a):
>>
>>> To average I just pick a pulsation cycle, add up the radius over that
>>> cycle and divide by the number of data points.
>>>
>>> On Tue, Jan 21, 2020 at 10:09 AM Radek Smolec <smolec at camk.edu.pl>
>>> wrote:
>>>
>>>> So I'm confused; you wrote:
>>>>
>>>>> I am taking a single pulsation cycle and averaging over the radius
>>>>> throughout the cycle
>>>>>
>>>> how do you average? It seems the pulsations nicely  repeat and problem
>>>> is in the averaging procedure.
>>>> If you are not averaging, but want to look at radius at some specific
>>>> pulsation phase, then you need to interpolate, as RSP time steps are not
>>>> synchronised with pulsation phase.
>>>> Radek
>>>>
>>>>
>>>> wt., 21 sty 2020 o 21:01 Jeremy Sakstein <sakstein at hawaii.edu>
>>>> napisał(a):
>>>>
>>>>> I think it does alternate. Here is a zoom in over 20 cycles:
>>>>>
>>>>> [image: image.png]
>>>>>
>>>>> Here is a plot with 3 alternating cycles:
>>>>>
>>>>> [image: image.png]
>>>>>
>>>>> The three colours correspond to the different cycles so with the
>>>>> exception of the number/position of data points it appears the plots are
>>>>> identical. Here's the radius, which also looks identical:
>>>>>
>>>>> [image: image.png]
>>>>>
>>>>> I'm not using the time-steps at all. I made these points by picking
>>>>> out the data points where the pulsation cycle (rsp_num_periods) is fixed to
>>>>> some value.
>>>>>
>>>>>
>>>>> On Tue, Jan 21, 2020 at 9:47 AM Radek Smolec <smolec at camk.edu.pl>
>>>>> wrote:
>>>>>
>>>>>> So does it alterante between two values?
>>>>>> Can you attach a plot of a few consecutive pulsation cycles, maybe
>>>>>> with zoom at radii maxima or minima?
>>>>>> If you get alteranating values of average radius the consecutive
>>>>>> pulsation cycles must differ.
>>>>>>
>>>>>> One other idea, do you interpolate for start of the pulsation cycle?
>>>>>> If you use the time steps as used by rsp, without interpolation, then some
>>>>>> scatter in the averaged radius is possible (but likely not alternations).
>>>>>> Radek
>>>>>>
>>>>>>
>>>>>> wt., 21 sty 2020 o 20:39 Jeremy Sakstein <sakstein at hawaii.edu>
>>>>>> napisał(a):
>>>>>>
>>>>>>> Hi Both,
>>>>>>>
>>>>>>> I don't think it's periodu doubling because I just see a single
>>>>>>> harmonic when I plot the light curves.
>>>>>>>
>>>>>>> Sorry to be unclear. I am plotting the average stellar (surface)
>>>>>>> radius for each pulsation cycle.
>>>>>>>
>>>>>>> On Tue, Jan 21, 2020 at 9:37 AM Radek Smolec via Mesa-users <
>>>>>>> mesa-users at lists.mesastar.org> wrote:
>>>>>>>
>>>>>>>> Hi Jeremy,
>>>>>>>>
>>>>>>>> It is difficult to guess what is happening without more detailed
>>>>>>>> information. The effect you describe may be physical: for example
>>>>>>>> low-amplitude period doubling phenomenon may be present in your model and
>>>>>>>> then averages for even and odd pulsation cycles will differ.
>>>>>>>> With such large difference you should be able to figure it out
>>>>>>>> simply looking at the light or radius variation curve.
>>>>>>>>
>>>>>>>> For the general point; what you consider as "true equilibriu
>>>>>>>> model"? RSP starts from equilibrium model, but it doesn't mean that the
>>>>>>>> linear properties of the model, such as pulsation period, will not change
>>>>>>>> when model pulsates at full amplitude.
>>>>>>>> It is best to develop own procedures for computing such quantities
>>>>>>>> as mean magnitude, radius, etc.
>>>>>>>> Cheers,
>>>>>>>> Radek
>>>>>>>>
>>>>>>>>
>>>>>>>> wt., 21 sty 2020 o 20:20 Jeremy Sakstein <sakstein at sas.upenn.edu>
>>>>>>>> napisał(a):
>>>>>>>>
>>>>>>>>> Hi Everyone,
>>>>>>>>>
>>>>>>>>> I have a quick question about finding the equilibrium values of
>>>>>>>>> the stellar parameters when using RSP Cepheid.
>>>>>>>>>
>>>>>>>>> At the moment, I am taking a single pulsation cycle and averaging
>>>>>>>>> over the radius throughout the cycle but this seems to give me two (very
>>>>>>>>> closely separated values).
>>>>>>>>>
>>>>>>>>> Below I am plotting radius vs pulsation number. As you can see, I
>>>>>>>>> end up with two very closely separated values for the average. Is this
>>>>>>>>> because the model is changing ever so slightly between pulsations? If so,
>>>>>>>>> why does it converge to two values and not a small scatter about the mean?
>>>>>>>>>
>>>>>>>>> [image: image.png]
>>>>>>>>>
>>>>>>>>> On a more general point, is it possible to extract the true
>>>>>>>>> equilibrium model from the output rather than taking an average?
>>>>>>>>>
>>>>>>>>> Cheers,
>>>>>>>>>
>>>>>>>>> Jeremy
>>>>>>>>> _______________________________________________
>>>>>>>>> mesa-users at lists.mesastar.org
>>>>>>>>> https://lists.mesastar.org/mailman/listinfo/mesa-users
>>>>>>>>>
>>>>>>>>> _______________________________________________
>>>>>>>> mesa-users at lists.mesastar.org
>>>>>>>> https://lists.mesastar.org/mailman/listinfo/mesa-users
>>>>>>>>
>>>>>>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200121/6f48b0ce/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 7667 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200121/6f48b0ce/attachment-0004.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 19402 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200121/6f48b0ce/attachment-0005.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 31932 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200121/6f48b0ce/attachment-0006.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 11526 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200121/6f48b0ce/attachment-0007.png>