Celestia Forums

Look here:

http://science.nasa.gov/headlines/y2005 ... 004mn4.htm

and

http://neo.jpl.nasa.gov/news/news149.html

Fightspit wrote:Look here:

http://science.nasa.gov/headlines/y2005 ... 004mn4.htm

and

http://neo.jpl.nasa.gov/news/news149.html

http://neo.jpl.nasa.gov/risk/

2036 - 2069 99942 Apophis (2004 MN4)

Fightspit wrote:Look here:

http://science.nasa.gov/headlines/y2005 ... 004mn4.htm

and

http://neo.jpl.nasa.gov/news/news149.html

Considering that both articles say and show that the asteroid will come very close to Earth but definitely not crash into it, I'm not entirely sure why you think it will "probably crash with the Earth".

symaski62 wrote:
http://neo.jpl.nasa.gov/risk/

2036 - 2069 99942 Apophis (2004 MN4)

Ok... but you don't talk about the year 2029 (13 April)...

Look:



from :
http://neo.jpl.nasa.gov/news/news149.html

I was talking about that - the articles quite clearly says that it's going to miss the Earth then. Heck, even the diagram shows that. Yes, it'll come very close but it won't hit us.

I know but I am very surprise the distance will be small when 2004 MN4 will be near of the earth:



the distance is almost 3 Earth....

Hi,

Is there anY SSC File available for this one?

Bye,
Brofn

Here's what I use in my asteroids.ssc file...

"2004 MN4" "Sol" # Approaches Earth within 35,000km on 04/13-14/2029
{
Class "asteroid"
Texture "asteroid.jpg"
Mesh "asteroid.cms"
Radius 0.32 # table lookup using H

EllipticalOrbit
{
Epoch 2462240.45833335
SemiMajorAxis 0.922384
Eccentricity 0.191047
Inclination 3.331
AscendingNode 204.472
ArgOfPericenter 126.384
MeanAnomaly 252.098572725691
Period 0.88649041095
}

InfoURL "http://newton.dm.unipi.it/cgi-bin/neodys/neoibo?objects:2004MN4;main"


}

Thanks a lot!
Merry Christmas / Nedeleg laouen!
Brokfn

A Best view:



Good Luck

Hello! It is a nice picture of the near miss, but a bit frightening too!

35000 km is still very close, and the geosynchronous sattelites are in some danger too in 2029, they are at 36000 km high.

Because the asteroid passes close to earth and moon in 2029, as can also be seen in the Celestia simulations, the gravitational interactions make an exact prediction of the next flyby circumstances in 2036 not yet possible. Possibly there will be a spacemission made to get a transponder placed on the asteroid. If new data predict a dangerous passage in 2036 or change the present predictions for 2029, maybe a mission could drop a small rocketmotor on the surface when Apophis comes into our neighbourhood again in 2013. I think the technogy already exists to do all that and it would also make good science. But certainly the asteroid will be closely watched in the next few years! More radar measurements from Arecibo would also be helpful.


I am a new fan of Celestia and very impressed with all its capabilities. I have just been experimenting a bit with what has to go into a .ssc file. Interested in the Near Earth Objects, I downloaded the NEO add-on but discovered that because of missing epochs, the orbits were all out of sync! There seemed to be a close encounter in 2007 with asteroid 1998 SZ27 but on the NASA pages there was no such encounter to be found for this asteroid.. With a new epoch filled in the asteroid appears in a totally different place.

As an experiment I then made a little file with a few asteroids, among them Eros and Apophis, that come near the earth. I made different versions (B,C,D..) with different orbital data as they are available from NASA's JPL pages and from NeoDys among others, and ultimately want to try to find a good orbit approximation for Celestia. JPL and NeoDys also have complete NEO databases online, regularly updated but for converting a whole database to .ssc I'd need to have some script written and I am not a programmer. With lots of data the big .ssc files are maybe not the best approach as can be read in other Asteroid threads in the Celestia developer's section. I'd still have to compute every orbital period from the Kepler elements because Celestia needs to have that. The close encounters are very sensitive to small changes in the period fed into our Celestia program. The only version of Apophis I have made sofar that was really realistic was ApophisE version. Apophis C was way off using the period from the old add-on, Apophis B and D came to about 370000-380000 km using the period data from the JPL page for Apophis at http://neo.jpl.nasa.gov/cgi-bin/db?name=99942, but that only gives four significant digits. It is possible to get a much better fit with the same orbital data elements if you compute the period again, with Kepler's third law or I think preferably with Newton's formula for the two body problem (You have to convert AU to km and get the mass of the sun, I used the data and formula's given by Wikipedia, the Urls are as comments in the ssc file).

Apophis E came not as close as in the above jpg, 35000 km is close to the official predictions. NeoDys gives a close approach of 0.0002441 AU, 36516 km on 2029/04/13.90679. The time I got was close to that in the picture above; on april 14th 2029 at about 7.28 UT "Apophis E" gets to within 48782 km of Earth. Comparing with the official predictions I think Celestia performs pretty well!

Viewing from the surface of Apophis I got approximately the same view of earth as in the picture above but an half hour earlier; in the Celestia FT1.2 simulation I could see the citylights of Sydney below me just switched on in the gathering dusk. After the flyby of earth the asteroid also passes fairly close to the moon: I got to within 73241 km on april 15th around 2.45 UT. The moon appears halflit, I can see several seas, the star Al Nair in the crane is visible just behind the moon.

I assumed I could just use the orbital data as given by JPL or NeoDys, if there is some correction I have to make for Celestia I would very much appreciate some comments! Because the events happen close to earth, any correction for light travel time in Celestia is not necessary I think.

Below is my little NEO2.ssc, it is very much a work in progress!

Eelco

As you mention, simple Keplerian orbits are valid only at the times of their Epochs. If you look at the bodies they describe at times more and more distant from that Epoch, they will be further and further away from where they should be. This is because Celestia cannot calculate the gravitational effects that change the Keplerian orbital parameters.

JPL's Horizons ephemeris service can be used to calculate xyz trajectories which may be more accurate.

An other dangerous asteroid,1950 DA:

When high-precision radar meaurements were included in a new orbit solution, a potentially very close approach to the Earth on March 16, 2880 was discovered to exist

From: http://neo.jpl.nasa.gov/1950da/



From: http://neo.jpl.nasa.gov/1950da/animations.html

Selden, thanks for the comments! Fightspit luckily that is still far off in the future! By the way, as I'm on dial-up link, this animation is okay but unfortunately some of the threads with a lot of images don't always load. But that is just minor detail. I should report there is small bug in my Neo2.ssc: in computing the orbital periods I used a tropical year of 365.242191 days to convert seconds to years. But I just found in this thread http://celestiaproject.net/forum/viewtopic.php ... t=tropical
that I should have used a sideral year of 365.25636 days which is about 20 minutes longer. My value for a tropical year comes from my copy of Peter Duffet-Smith's 'Easy PC Astronmy' which, although not entirely up to date to modern PCs (The programs that come with it are DOS and the Turbo Pascal bug has to be corrected for modern clockspeeds) is still a very valuable reference and introduction to Astronomy.

I will work a little more on my .ssc files, I see that JPL has updated a solution #119 for Apophis, I will try that out.

Modeling gravitational interactions in Celestia in the future is something of a dream, but I realize it conflicts a bit with the whole architecture of the Celestia program. More custom orbits, as an alternative to XYZ orbits, like the VSOP solutions, is possible I think but would require a lot of work and expertise; I believe making them is something of an art! Been looking into the latest solutions for the moon orbit a bit... Maybe for another thread one day!

Thanks, Eelco

A postscript: Strange, when using a sidereal year of 356.25636 days to compute an orbital period out of the SemiMajorAxis, (for "Apophis G"), the perceived accuracy of the flyby in 2029 is more or less destroyed using the JPL data. It could be a coincidence, the osculating elements are not for 2029 but for 2006, but it is still strange that a tropical year gave a much better fit (used in "Apophis E") and also matched the JPL value for the orbital period in four decimal digits. This is a bit confusing. Could it be that the VSOP orbit of the earth actually uses the tropical year?

Eelco

Apophis postscript #2: Curious as it may sound, the close fit achieved by the "ApophisE" orbit was probably pure coincidence of picking exactly the right wrong parameter! Tropical year should have been sidereal and the calculation of the period was wrong. The elements were for epoch 2006, not 2029 but somehow these three errors cancelled out including any other inaccuracies.

But in the mean time some new official data: three new datapoints lead JPL to do a new JPL solution ID#120 and with a telnet session with JPL Horizons I managed to get osculating elements of this solution for an epoch two months before the flyby. Now the accuracy of the simulation I think should depend mostly on the precise placing of Earth in Celestia at february 13th 00:00 hrs in 2029, two months before close approach. If I haven't made lots of new typo's or interpreted symbols and orbital element designations wrong... Sorry, I could have picked a later date, but this should work too, next version I make some day should be epoch april 13th but I did not yet get that data from JPL Horizons. I also found some data on the rotation period of the asteroid. I haven't yet checked what the simulation will do: so this is totally new data if you want to try this! The latest ApophisJ.ssc version, I hope it will be a miss..



http://neo.jpl.nasa.gov/cgi-bin/db?name=99942
http://neo.jpl.nasa.gov/cgi-bin/da_legend
http://en.wikipedia.org/wiki/Sidereal_year

Raw JPL telnet data:

2462180.500000000 = A.D. 2029-Feb-13 00:00:00.0000 (CT)
EC= 1.911748408391438E-01 QR= 7.460277553477852E-01 IN= 3.341614784786085E+00
OM= 2.038689088263874E+02 W = 1.266945746950668E+02 Tp= 2462337.140250629280
N = 1.112637193884422E+00 MA= 1.857162310906158E+02 TA= 1.839558229883659E+02
A = 9.223597299097097E-01 AD= 1.098691704471634E+00 PR= 3.235556046290107E+02

EC=eccentricity
QR=perihelion distance (not used in Celestia)
IN=inclination
OM=node=Longitude of the Ascending Node=AscendingNode
W=w=Argument of Perihelion=ArgOfPericenter
Tp=Time since Perihelion(=periapsis?) Passage (not used)
N=mean motion (deg/d) (not used)
MA=M=MeanAnomaly (deg)
TA=?
A=a=Semi-major Axis
AD=? probably Q=Aphelion Distance (not used)
PR=Period (days) I assume mean solar days

This is the version I actually had meant to give, Apophis K with orbital elements valid about 24 hours before the close approach (to take into consideration more of the gravitational attraction of the earth)

This version is more accurate close to earth but not very accurate in the period preceding it, also not of course after the approach, after the big bend in the actual orbit caused by the earth's gravity.
The approach margins shown by Celestia are I think fairly realistic but I have not yet compared it to the official JPL ones for this solution ID JPL#120.

Celestia gives:

Apophis J, Epoch February 13th 00:00, Close approach 41250 km at 2029-April-14 00:32 UTC
Apophis K, Epoch April 13th 00:00, Close approach 40626 km at 2029-April-13 22:21 UTC

If you view this tracking the Earth with night side lights enabled, the effect is fairly spectacular, (I'm seeing it in Celestia FT1.2), in these orbit approximations it is mainly Europe and North and South-America that you see close by below, the continents lighting up by all the electric lights just after sunset.