Does Nemesis Exist? Has it Ever?

[mlorrey]

Mikemmert has a good idea here: I'd really like to be able to break the solar system up into different zones, each operating at different time steps, communicating only to transfer ownership of an object that is traveling from zone to zone, along with timing info as to the time stamp that it translates the boundary space. Thus, I could have the terrestrial part of the solar system running on one machine with a very small time step, the jovian zone out to 50 AU running another, and 50 AU to 3 ly on another, with the outer zones treating the inner zones as an average gravitational field that integrates the sun with the planets of the inner zones.<br /><br />Appreciate the feedback on why integration doesn't work. I am wondering if every object can't simply have its path integrated as if it were an N=3 body problem, given the assumption that for most all conditions, the influence on any given body is overwhelmingly dominated by two other bodies and all other bodies, together offer miniscule influence that for all intents and purposes could be averaged out.

 

[mlorrey]

Any numerically integrated solution can be solved in a few ways: given a finite number of steps, the path becomes a sawtooth rather than a curve. How the sawtooth lines up to the curve is the major source of math errors that pile up: the sawtooth rides either above, below, or sits its mean upon the curve. The process you describe appears to me that the sawtooth rides below the curve, so the average for each time step is actually slightly less than real. The smaller the time step, the smaller this underestimation is. <br /><br />However, if the sawtooth sits astride the curve, then the area above the curve should, on average, cancel out the area below the curve, and should result in the most accurate path estimation.

 

[tony873004]

As to why integration doesn't work, google for n-body problem. It seems to have something to do with the fact that there are more unknowns than knowns in the equations.<br /><br />Simply integrating a body as if N=2 (not 3) is the way a lot of the planetarium programs operate. VSOP87 is the main engine behind their computations. And it is very accurate for hundreds of years into the future.<br /><br />But these formulas are just a representation of what we've already observed. We've observed that Mars follows an elliptical path around the Sun. We've observed that its eccentricity changes as a function of time. We've observed that its orbital nodes precess as a function of time. So we can write a formula to predict where it will be, or where it has been. <br /><br />But is the rate of change of these pertubations from an ellipse constant with time? No. And if we don't have enough observations to know how they change long-term, then the analytic methods will fail.<br /><br />Using numerical integration, you don't need to know the pertabutions. They are a natural consequence of the numerical integration.<br /><br />For example, consider the simulation "saros.gsim" on the Gravity Simulator web site. It shows how the Moon's nodes (points where the plane of its orbit intersect the plane of the Earth's orbit) complete 1 cycle in an 18 year period. But there is nothing in the source code of the program to tell it to do this. It is just a natural consequence of numerical integration. A planetarium program will hard-code this into the program. And if this motion were not known, the planetarium program will not help you discover it.<br /><br />The main advantage of doing things the way the planetarium programs do it is so you can instantly jump to any date. In a numerical integration, you've got to chug your way there, step by step, which can be time consuming.<br /><br />So for things such as predicting how Nemesis would interact with the Oort Cloud, where we have no obser

 

[mikeemmert]

<blockquote><font class="small">In reply to:</font><hr /><p>... predicting how Nemesis would interact with the Oort Cloud<p><hr /></p></p></blockquote>If Muller's lunar spherule model is right, (and I have expressed some reservations about that) and this thing has only been going for 400 million years with a 26 million year period, then the comets could well have been shaken loose from the Kuiper belt rather than the Oort cloud. In fact, it could be that the Oort cloud is too thin and spread out to be the source of these comets.<br /><br />In fact, the Kuiper Belt was in the plane of the solar system long ago, probably. This would give a much higher probability of a comet hitting the Earth. The Kuiper belt is more concentrated, with more comets per cubic kilometer. <br /><br />mlorrey, are you still trying to do 26 MY orbits? I think you can represent the situation with a much shorter period orbit. Some times you have to use tricks like that to fool the machine into giving you more accurate answers in a shorter time frame. The question is, where are these comets getting shaken loose from, and does the shaking object leave some kind of trace on the Kuiper belt, Oort cloud, short or long period comets, or any other phenomenon. The action is happening at perihelion, I believe.<br /><br />I have what I call "base simulations". When I'm doing binary flybys, I get everything set up except one factor which I am going to vary. For instance, I wanted to have a slightly different angle at closest approach between Triton at the point of the angle and Xena and Neptune on the two lines. To do this, I would do a simultation and then increase the distance between Triton and Xena by one kilometer increments. This rotated the angle through 360 degrees in about 15 or 16 tries, if the separation between the objects was 20,000 km. All I had to do was change one number and erase all the dispersions created for making groups of objects. (Actually I blanked longitude of the ascending node

 

[mikeemmert]

Here's a screenshot of a Solar system (with the inner terrestrial planets deleted to speed things up) with a Nemesiss (in addition to deliberate misspelling to denote imaginary object, there is a <i>real</i> typo) clone of about 13 Jupiter masses and an eccentric orbit inclined 30 degrees. Notice that the inner Kuiper belt is not that much disturbed, but the outer belt is wildly chaotic. There are objects way off the screen.<br /><br />Here's the link, I hope the moderators approve the image:<br /><br />http://orbitsimulator.com/gravity/simulations/users/mikeemmert/NmeisissII.gsim<br /><br />and here's another one from wikipedia showing the orbit of 2004 XR190 "Buffy", and several other orbits. Notice how much it looks like the simulation:<br /><br />http://en.wikipedia.org/wiki/Image:XR190orb_side.gif<br />

 

[mlorrey]

http://www.space.com/scienceastronomy/060207_habitable_zone.html<br /><br />Latest news shows scientists are agreeing with many of my own points about habitability zones and red dwarves...

 

[mlorrey]

The Pan-STARRS first of four telescopes is nearing "first light". Among its goals will be detecting possible large companions, from Earth mass up to possible brown or red dwarfs. This four telescope system is truly daunting in its power. It will have CCDs of over a billion pixels.<br /><br />Here are the distances at which various sized objects could be detected with this device, once all four telescopes are online:<br /><br />Table 3: Detectability of Distant Planets<br /><table><tr><td><br />Planet

 

[mikeemmert]

Thanks for the alert, mlorrey. Here is the official Pan-STARRS website from the University of Hawaii.<br /><br />Some quotes:<br /><br /><i>Pan-STARRS -- the Panoramic Survey Telescope & Rapid Response System -- is an innovative design for a wide-field imaging facility being developed at the University of Hawaii's Institute for Astronomy. <br /><br />By using four comparatively small telescopes, each with a 3-degree diameter field of view, we will be able to develop and deploy an economical observing system that will be able to observe the entire available sky several times each month. <br /><br />The immediate goal of Pan-STARRS is to discover and characterize Earth-approaching objects, both asteroids & comets, that might pose a danger to our planet. <br /><br />The huge volume of images produced by this system will provide valuable data for many other kinds of scientific programs.</i>.<br /><br />The real advance of the Pan-STARRS is the enormous number of pixels available. The detectors have an enormous area. The optical systems are deliberately small to give a wider field of view:<br /><br /><i>How big are the CCD cameras?<br />The four PanSTARRS cameras will each be the largest digital cameras ever built. Each camera will have about 1.4 billion pixels spread over an area about 40 centimeters square. For comparison, a typical domestic digital camera contains about 3 million pixels on a chip a few millimeters across.<br /><br /><br />How many CCDs are there in the focal plane?<br />The focal plane of each camera contains a 64 x 64 array of CCD devices, each containing approximately 600 x 600 pixels, for a total of about 1.4 gigapixels. The CCDs themselves employ the innovative technology called "orthogonal transfer", which is described below. <br /><br />The individual CCD cells are grouped in 8 x 8 arrays on a single silicon chip called an orthogonal transfer array (OTA) , which measures about 5 cm squar</i>

 

[mlorrey]

Charting it out, I'm expecting they'd be able to detect a 10 Jupiter mass object at 5,000-10,000 AU, and a 50 Jupiter object at 20,000-30,000 AU. Once an object is above the 15-20 J mass level, its own internal emissions should improve its ability to be detected at further distance.

 

[mikeemmert]

I found some cool <img src="/images/icons/cool.gif" /> pictures today while I was working on a paper for my Xena project. This was in Wikipedia, as the next couple of posts will be (one image at a time):<blockquote><font class="small">In reply to:</font><hr /><p><i>The diagram on the right illustrates the orbits of all known scattered disk objects up to 100AU together with Kuiper belt objects (in grey) and resonant objects (in green). The eccentricity of the orbits is represented by segments (extending from the perihelion to the aphelion) with the inclination represented on Y axis.</i><br /><br /><br /><br /></p></blockquote>

 

[mlorrey]

They plan on taking parallax images 3-12 weeks apart. Should be plenty of time to detect movement even for objects moving directly away from us.

 

[mlorrey]

I find it interesting that KBOs are exhibiting a higher D signal than the inner solar system. This could be indicative of a brown dwarf heliopause passing through the Oort/Kuiper region, or at least that one once orbited out there.

 

[mikeemmert]

Y'know, mlorrey, I hope you're right, I really do. It's obviously vital to find this thing.<br /><br />And you might be right. So far, I have no real idea just how big this thing is. It could very well be a red dwarf. That would account for the damage to the Kuiper belt.<br /><br />But I fear you are wrong. The damage to the Kuiper belt could easily be accounted for by a much smaller object. The geological evidence you have submitted says nothing about the size of the perturber, it only indicates that the perturber was perturbed somewhere between 400 and 600 million years ago.<br /><br />I'm going to make another attempt at posting my GravSim of the damage a 13 Jupiter mass object can have on a simulated solar system which started out with all KBO's at 0 degrees inclination. I just recently imported some software that converts the .gsim images to .jpeg. I remember trying to post something like this on this thread. This is not the best image, even of that series. This was dug out of an old file.<br /><br />I also had trouble posting the material from Wikipedia. I don't understand why one image went through and two others did not. It has something to do with the kind of files Uplink will accept. This post is kind of a guinea pig, you won't see it if there is an obvious failure to post and might see the post and not the simulation anyway. Cross your fingers. <br /><br />The simulated KBO's were knocked out of their zero inclination orbits by the red object, which is the simulated 13 Jupiter mass Nemesis. I did this simulation a long time ago and have frankly forgotten a lot of details. I think the semi-major axis of "Nemesis" is 600 AU. As I said, "Nemesis" is red, did somewhere around 15 or 20 passes, and the "KBO's" are various colors.<br /><br />Cross your fingers or whatever superstitious thing you do to keep psychlopropane from blowing up ( <img src="/images/icons/smile.gif" /> )

 

[thepiper]

<font color="yellow">They plan on taking parallax images 3-12 weeks apart. Should be plenty of time to detect movement even for objects moving directly away from us.</font><br /><br />I hope they factor in the Sun's movement through the galaxy during this time for their measurements.

 

[mlorrey]

I think they are competent to do whatever they need to do to make accurate measurements.

 

[mlorrey]

The problem is that even a dozen Xena type bodies could not clear out the KB like it has been. Observations could not see brown dwarfs in that region. I've posted the sensitivity range of the Pan-STARRS, and that is the most sensitive telescope to date for this type of survey. A small brown dwarf of 20 J masses could be out there past 10k or 20k AU, and we still couldn't see it, even with Pan-STARRS, unless we already knew exactly where it was and were looking for it.. I also note from the sensitivity range of my prior post that the gravitational influence range is significantly less than the visible sensitivity range.