# # # Celestia Impactors Add-On -- About the data sources used # # 0. Preface 1. Tunguska event of 1908 2. Sikhote Alin Meteoroide of 1947 3. Jackson Lake Bolide of 1972 4. Shoemaker-Levy 9 and Jupiter 1994 5. Additional information 0. Preface ---------- The intention of this add-on was to give some *qualitative* models of catastrophic encounters between a planet and a minor celestial body. I chose what in my view were the most impressive events of the 20th century. In particular we have examples for the following: -straight crash of a small body into a planet (Sikhote-Alin) -nearly miss resulting in a very small horizon angle of impacting body (Tunguska) -small body misses planet but pierces through atmosphere (Jackson-Lake) -body misses (at first try) but was torn apart by gravitational forces of planet (Shoemaker-Levy) Note that I had to do some modifications of the original data (which in no case was accurate enough) to hit the right point in space and time. I feel satisfied if the body hits +/- 1 hour and +/- 100 km which is enough to show the overall view. This sounds much but is a *very* close hit with respect to the dimensions of the solar system. 1. Tunguska event of 1908 ------------------------- The data of a possible Tunguska impactor is taken from the paper of Farinella et al, Astronomy & Astrophysics 337(2001) p. 1081 (No. 444 in the notation of the paper). It is not a body of the group the authors state to be the most probable origin of Tunguska. I made this particular choice for two reasons. 1. I don't see why the authors take so much 3 deg horizon angle samples into their survey. Most other authors give 15 deg, some up to 30 deg (which was not included as samples). However, if we drop all of the low-angle samples, the chance of the body's origin asteroid vs. comet still is 2:1. But most of the remaining trajectories are quite similar (highly excentric). 2. *I* am convinced the body was an asteroid but maybe you're a fan of the cometary hypothesis. So the orbit data should fit to both possibilities, which is well done by object No. 444 of the S4 group of the cited paper. Its bodies have an equal chance to come from the Jupiter Family of Comets as well as from the outer asteroid belt. As an addition, I put in an Atmosphere (presently uncommented, remove the #'s in "tunguska.ssc" if you want to use it) which can serve as the faint coma of a nearly exhausted comet if desired. 2. Sikhote Alin Meteoroide of 1947 ---------------------------------- There are a number of papers about the Sikhote-Alin meteroroide, I took the data from Fesenkov, Meteoritika 9(1951) p.27. It has been noted elsewhere (see the articles of R. Gallant in Sky & Telescope) that today's estimate of the impact velocity is somewhat lower which would make the orbit less excentric. Sadly I could not find an example of this newer orbit data so I used the older Fesenkov values. 3. Jackson Lake Bolide of 1972 ------------------------------ The orbit of the Jackson Lake Bolide has been calculated by Ceplecha, Astronomy & Astrophysics 283(1994) p.287. However, there are some difficulties. It is next to impossible to fit his pre- and post-encounter orbits together (remember we talk about an accuracy in the meters range here - in a coordinate system that deals with au's!) so I had to decide which of the orbits I put in. Doing it for both, it showed up that the pre-encounter data of the paper caused the body to impact into (or scratch above) the pacific area. Obviously earth's gravitation has "bent" the real trajectory substantially. Celestia cannot simulate that. Fortunately both orbits aren't that distinct (they differ mainly in the orbital inclination) so I decided to blend the data of post-encounter with pre-encounter inclination. This made a good fit. It also leaves the possibility to watch the close encounter to earth in 1997, for which inclination plays a minor role. For "cosmetics" I set up a texture file for post-encounter which has a badly burned front face showing some rhegmaglyphic structures. 4. Shoemaker-Levy 9 and Jupiter 1994 ------------------------------------ It was not my intention to replace one of the existing SL9 packages for Celestia. If you're interested in exact models of impact phenomena, the work of Jack Higgins (see forum) may be more suitable for you; if you want to have highly precise final orbits of all 21 fragments, Seldens's package (at http://www.lns.cornell.edu/~seb/celestia/index.html#3.3) is the better choice. I wanted to have a *qualitative* simple model of the last three orbits of the body before and after its breakup. For this, I had to simplify the data provided by Selden substantially. However, the motions of a body like SL9 cannot be described in a simple way (it is a three-body problem which has no analytical solution). Moreover, it seems if there are two groups of authors, some describe SL9 to be caught by Jupiter from a heliocentric orbit immediately before impact, some other state it should have been in an (irregular) orbit around Jupiter for at least some decades before (I believe in the latter). So all "orbits" are only partial approximations. To get an overall view, a trajectory has to be fitted from such orbit pieces if you're not able to get .xyz data somewhere (which I'm not). It is possible to fit a set of orbit ellipses of same orientation in a plane by variation of semimajor axis and excentricity only. So I took the SL9-L fragment (largest) out of Selden's data as a reference and set up two other orbits which touch it at apocenter (where the distortions of the orbit(s) by sun's gravitation are greatest). By sort of variation I was able to get orbits for 12 other fragments which all touch my orbit at the point of pericenter (where breakup occured). In view of space navigation, this is a Hohmann transfer orbit for each fragment. By a mathematical analysis all orbits can be made precisely fitting. That means, in Celestia you can sit at the SL9 mother body approaching its last pericenter and see it literally break apart under your feet - and none of the fragments does a step or hop when the orbit is switched. Of course this has a cost. Keeping most of the orbital parameters fixed, I had to set another parameter free to adjust time. This was done by orbital period. By the standpoint of physics, that's not correct but the derivation is not that big -- all orbits are similar extremely elliptic with semimajor axes of approx 26 Mio. km. For the last trajectory segment I adjusted by this trick the impact time to the observed values. A note about some "cosmetics": I put faint coma's to all bodies by the "Atmosphere" trick mentioned above. Celestia's comet feature isn't satisfying for me to model such small exhausted comets - the tail is by far too large and can't be adjusted in size (maybe a request for a future version?). Moreover, in my Celestia 1.3.2 installation the tail of a comet orbiting a planet points away from the *planet* not the sun. 5. Additional information ------------------------- Additional information at least about the Tunguska impact and Sikhote Alin I have put on my homepage: http://www.geocities.com/diane_va/sikhote-alin (german) http://www.geocities.com/diane_va/tunguska http://www.geocities.com/diane_va/sikhote-alin/index_E.html (english) http://www.geocities.com/diane_va/tunguska/index_E.html October 9, 2004 Diane Neisius (~Medusa)