Fading Comet Tails in Multiple Star Systems

[t00fri]

Hi,

I thought some of you might be interested in my latest comet coding for the forthcoming Celestia 1.4.0pre8 version:

+++++++++++++++++++++
Let me discuss a plausible scheme generalizing the distance
x0, where comet tails start fading, to systems with possibly several suns and luminosities different from the solar one.
++++++++++++++++++++

Consider first a system with only /1 sun/, but arbitrary luminosity L

The relevant quantity for the comet tail to become visible, is the sun's light pressure, i.e. its luminosity/area, also called light irradiance Irr =L/(d^2), with d being the distance of the comet from the sun.

For our solar system, the threshold irradiance where the tail starts fading, is (empirically) known to be
Irr_crit =L_sol/x0_sol^2, with x0_sol ~ 5 AU and L_sol being the known solar luminosity.

It is now very sensible to assume that comet tails start vanishing at a universal threshold irradiance
Irr_crit that we know already from our solar system.

Hence for any other system (with 1 sun), we immediately obtain the scaling law for the fading start x0

+++++++++++++++
x0 = x0_sol sqrt(L/L_sol).
+++++++++++++++

Since we all know the familiar relation, expressing the star's luminosity in terms of its radius R (and temperature T),

L ~ R^2 * T^4 (black body),

we confirm x0/x_sol ~ R/R_sol, as I proposed already some time ago.

(Chris proposed x0 ~ L ~ R^2 which is thus disfavoured!)

Next, I generalized this concept to multiple systems i.e. with several suns! Along the comet's trajectory, one simply has to pick that sun which produces the maximum light irradiance as function of the comet's distance! The rest is straightforward.

This is now all coded and works very neat, e.g. for the tau Boo system, having two suns! Below is a graphical display of a ficticous comet orbit passing close to the two stars tau Boo A,B. Depending on the comets distance, it's tail points to tau Boo A if the distance is smaller than 9.1 AU or to tau Boo B if the respective distance is smaller than 0.43 AU. Indeed it all works as calculated

Bye Fridger

[Malenfant]

So presumably, if more than one star can create enough light pressure at the comet's current location to create a visible tail, you can get two comet tails pointing away from each star?

[t00fri]

So presumably, if more than one star can create enough light pressure at the comet's current location to create a visible tail, you can get two comet tails pointing away from each star?

Indeed...

To the present approximation some inaccuracies may arise around comet positions where the light irradiance from several suns is about comparable. This "higher order" effect may, however, be easily accounted for in a next "round" of tuning...

Bye Fridger

[Malenfant]

Can you vary the brightness of the tails? So that maybe the tail pointing away from the star with the higher light pressure would be brighter than the tail pointing away from the star with less light pressure?

[t00fri]

Can you vary the brightness of the tails? So that maybe the tail pointing away from the star with the higher light pressure would be brighter than the tail pointing away from the star with less light pressure?

Yes, easily, but this is then getting us into a "deeply speculative" regime. How should I scale the brightness with the light pressure? I'd prefer to keep things transparent, until we find a good ansatz how to do it well. The problem is that for this task, one has to map in the physiological response of the human eye and a generic monitor. This is all VERY soft ground

Bye Fridger

[Malenfant]

The problem is that for this task, one has to map in the physiological response of the human eye and a generic monitor. This is all VERY soft ground

Bye Fridger

Well did you see my post in the bugs forum about multiple light sources? I think that this 'human eye response' is on very shaky ground as it is. I got some numbers to back up what I said there too, just trying to figure out how to generalise it more...

[jdou]

Comet can have 2 tails ?
Why not only one, in direction to the combination of the 2 sunlights pressure vectors ?

[BrainDead]

So presumably, if more than one star can create enough light pressure at the comet's current location to create a visible tail, you can get two comet tails pointing away from each star?

Not to nit-pick here, but to truly understand what I think I know about
space...

When you refer to "light-pressure" you are really referring to "Solar-Wind."

Yes?

In other words, it's not photons from the Sun which cause the tail on the
comet. It's the particles emitted from Coronal Massive Ejections and
other solar eruptions. Is this not correct?

Thanks, Bob

[GlobeMaker]

Clarification is needed for the two tails comment.
If two stars are near a comet, there will be two directions of
lighting (photons). The solar winds from the two stars will
combine to form one solar wind. That is like two water rivers
converging to make one big river. The solar wind is made
from protons and other particles. This single solar wind sometimes
only makes one gas tail on a comet.

A comet can have two tails : a dust tail in one direction and a
gas tail may be in a different direction. See comet West photos.
If two stars are present, the single wind can make two tails on
a comet : dust and gas tails. That is also true when only one
star is nearby.

http://antwrp.gsfc.nasa.gov/apod/ap950826.html

This link shows two tails on Comet West, and explains the ion tail
and the dust tail.

This link explains more

http://www.solarviews.com/eng/comet.htm

The light pressure contributes to the tail. So two stars have two light
pressure directions, and two tails could be seen due to two stars'
light pressures.

[jdou]

Thanks GM for your explanations

[t00fri]

Certainly,

the points made about 2 tails are generally correct, BUT...

for now I have not done any modifications to the (old) basic comet rendering code by Chris. One reason is that presently we experiment with quite different rendering techniques both for stars and DSO's (point sprites)...So before rewriting the comet code, it's better to gain more experience here.

In addition, the physical origins of the two tails are quite different, hence a generalization to non-Solar systems less straightforward. We know from our solar system already that not all comets develop both a ion and a dust tail...so criteria are definitely more complex for adaptation to other systems. This part needs more preparatory work before coding can start.

As to picking for now the star with the maximum light irradiance along the comet's trajectory, this is to be considered a typical theorist's /first/ approximation to a more complex problem. Clearly there may be significant corrections both to the luminosity scaling law for the tail fading point and to the tail direction vector in the region where several suns contribute comparable irradiance. Yet for the (large) parts of the comet's trajectory where the irradiance from one sun dominates, the present approximation is fast and satisfactory.

The next "iteration" may be starting from the following procedure, very familiar in theoretical physics:

Let L_i , vec_i(x) , d_i(x) =sqrt(vec_i(x)^2) be the luminosities, vectors and distances from the comet to the i th sun, respectively. Then we may construct a smoothely changing, irradiance weighted resulting orientation vector as follows:

Code: Select all

light irradiance "probability" weight  from i=1..N suns:
--------------------------------------------
P_i = Li/d_i^2 /sum_i(Li/d_i^2)  > 0 , such that

sum_i (P_i) = 1  (proper "probability" normalization!)


The effective direction for the tail, would then be given as

vec(x)_eff := sum_i (P_i * vec_i(x))

Clearly if one irradiance, say P_k, among the set of irradiance weights /dominates/ all others, we immediately get back my present approximation:

vec(x)_eff ~ vec_k(x).


In this more complex case, the scaling law for the tail-fading point will also become more complicated....

We'll see what turns out to be the best overall compromise.


Bye Fridger

[wcomer]

So presumably, if more than one star can create enough light pressure at the comet's current location to create a visible tail, you can get two comet tails pointing away from each star?

Certainly not! A gradient is a gradient, is a gradient. You could, however, have the tail pointing in a direction that is no longer directly away from either star. I'm very excited to see this in affect. But I'll be dissapointed if any one comet is ever shown displaying two ion or two dust tails. Both situations seem unphysical, although things could possibly get more interesting if there were a turbulant boundary layer at the merging of two solar winds fields.

-Walton

[Malenfant]

So presumably, if more than one star can create enough light pressure at the comet's current location to create a visible tail, you can get two comet tails pointing away from each star?

Certainly not! A gradient is a gradient, is a gradient.

I didn't realise it'd be an average of the two directions . Though that makes sense now it's been explained a bit more.

[rthorvald]

I thought some of you might be interested in my latest comet coding for the forthcoming Celestia 1.4.0pre8 version

I have to say it is very, very inspiring, all the new development work that you (and others) are putting out these days. Thank you!

BTW, comet tails suffer from the same "atmosphere gap" bug that "regular" atmospheres do ,but it is more prominent. Is there any hope of getting rid of this, or is that a different "cup of tea" in this context?

(A different question of course is whether the tail should be seen from the comet surface in this fashion at all?).

-rthorvald

[Paolo]

I would like to express some opinions on this issue. Probably are incorrect but... IMHO

Comets with stable closed orbits in multiple star systems should be extremely improbable due to reciprocal gravitational effects. Even planets in some regions of multiple star systems are impossible.

Stable cometary orbits should occur only if a massive star is the barycenter of the system and the comet remains very far from the barycenter itself. So comet tail will be very small all the time.

The density of the solar wind of the stars (except during novae events) is very low so at least solar winds from different stars will compenetrate and interact differently accordingly with the distance from the stars themselves and their mass and age. It is probable that the winds will create turbulence effects at short distances and that won't merge at all over a certain distance. In the last situation a comet should have two or more separate very short and faint tails (almost invisible all the time). In the turbulence area stable comets will be impossible.

The problem seem quite complex to solve. If there is a good reason to solve it. It should have a reason only for single pass non stable comets. Perhaps Frank Gregorio should be interested in developing a lesson about this creating an opened orbit comet realistic addon.

So the separate ion and dust tail for canonical single star systems seems to me a more intriguing issue. But perhaps I'm wrong.

Kind regards.

[Vincent]

I totally agree with you Paolo. (See my post http://www.celestiaproject.net/forum/viewtopic ... 2536eb0d21)

The rendering of the two components of comet tail would be a big improvement. And I would propose then an update of my Deep Impact Mission script...

[wcomer]

Paolo,

I agree with most of your post, the exception being the two or more tails bit.

Given the low density of solar wind, where the streams merge, there should be a region with low mixing such that the particles streams are separate. The mean free path length of solar wind particles is about 1 AU (*). So that's about as large a mixing region as we'd expect. Now, if the dust tail was a result of single ion-dust collisions, then two tails would be possible in this region. However, the tails are the result of multiple collisions, and since the collisions could come from either stream in our mixing region, the net result will be one tail along the sum of the stream vectors (and it should be fatter than usual due to the variance of collision vectors.) As for the ion tail, I believe the relevant length scale is the Debye length, which for solar wind is about 10m (*). So there is really no circumstance for multiple tails, although I'd expect considerable turbulence of the tail in the mixing region.

cheers,
Walton

* - http://www.apl.ucl.ac.uk/lectures/3c37/3c37-2.html

[t00fri]

Hi,

just returned from a conference at CERN/Geneva an hour ago.
While tonight I am too tired for any sensible remarks, I guess, I'll be back here by tomorrow night

Bye Fridger

[maxim]

To my feeling multiple suns should primarily increase the angle that defines the tails width within areas of similar amounts of light pressure.

Starting with a picture of two tails for a binary system, every tail will also be affected by the light pressure of its non-adjacent sun. Thus it will try to blow the tail away. The result then should be a much wider tail with a much less intensity. The width of the tail and its direction should be a function of the angles between the suns as seen from the comet. I'm sure you can find examples of multiple wind sources (turbines i.e.) blowing dust away.

maxim