By Steven M. Tilley
One of the programs available to the amateur observers of asteroids and comets is Find_Orb.[By Bill Gray] It is useful for calculating approximate ephemeris, determining approximate orbits, generating virtual asteroids, and virtual impactors, predicting impact locations, and many other things.
This guide walks you through using Find_Orb to perform a Monte Carlo simulation for predicting asteroid impact locations, specifically for 2024 YR4 and the Moon.
1. Install and learn Find_Orb if needed.
For the most up-to-date version of Find_Orb and information on how to use Find_Orb, go to https://www.projectpluto.com/find_orb.htm.
Access (Beta)MPC Explorer and download the asteroid 2024 YR4 observations in ADES XML format. Note there are three observation formats (80-column MPC format, ADES, or the AstDyS/NEODys .rwo format). ADES (Astrometry Data Exchange Standard) is the newest format and has more data.
Open Find_Orb, load the downloaded observations, and add in planetary perturbation by toggle on perturbers by hitting the number keys for the Planets; the "0" key is for the Moon, and the "a" key is for major asteroids. Turning on all will slow down Find_Orb and may not always be needed.
Hit the "f" key a few times using the least squares method.
Use Find_Orb filters to remove any outliers and improve accuracy
Hit "{" to filter all observations above to the preferred residual ceiling. Hit enter then take a few "Full Steps."
Hit the "w" will find and highlight the worst (highest-residuals) observation, and the "x" will toggle it off. Like JPL, I used the 441 best observation. Then take a few "Full Steps"
Hit the "e" key and enter epoch 2032 Dec 22.635417 [2032 Dec 22 15:15:00 (3:15 PM) UTC] right before the possible impactor hit enter Then take a few "Full Steps."
Given the fact there is NO such thing as a "perfect" observation, a Monte Carlo simulation adjusts each observation randomly with a defined limit, and variant orbits are calculated for the adjusted observations this is done many times. To run a Monte Carlo simulation for an asteroid. Hit "|" and enter the Gaussian noise level [The Gaussian noise level should be selected based on a judgment of the quality of the observation used. Given I set the residual ceiling as 0.6, I used a Monte Carlo noise factor of 0.6.] Hit enter, and Find_Orb will generate variant orbits and make entries into the files mpcorb.dat, virtual.txt, and state.txt. Until a hit of another key stops it. The impact will be noted in the files mpcorb.dat, virtual.txt
The longer Find_Orb runs, the more variant orbits there are, and there could be more impacts. This will help fill in gaps in any asteroid impact risk corridor and give a better assessment of the impact probability. There were times when the first impact was at variant orbit number 15. Stopping Find_Orb at this point will give an impact probability of 6.6%, which would be wrong. I let Find_Orb run for 6 to 24 hours, so I have 10,000 to 20,000 variant orbits.
The first impactor may be listed after more than 50 entries in the file, so open the virtual.txt or mpcorb.dat file in Notepad and ensure an impactor is listed on the first line (this helps with the Excel import). Also note that virtual.txt may have entries for "Center: 3"[Earth] and "Center: 10"[Moon]. If this is the case, do a find and replace search"Center: 3" with "Center: 03" so the columns line up.
Load the edited virtual.txt or mpcorb.dat file into Excel.
Step 9.1 Open virtual.txt with Excel pick "Fixed width"
Step 9.2 Keep the packed provisional designation and virtual asteroid ID number in the same column.
Step 9.3 Keep the date and time in the same column.
Adjust the time formatting settings in Excel to correctly interpret and display the impact times[yyyy-mmm-dd HH:MM:SS.00].
Sort the data in Excel based on impact time.
Find_Orb uses 0 to 359 format for Longitude; this is useful for Guide 9.1. However, it will not work for Google Earth. Convert longitude values from [0 to 359] to [-180 to 180] using:
=IF([raw longitude] > 180, [raw Longitude] - 360, [raw Longitude]) [=IF(W2>180, W2-360, W2)]
Use Excel to create Placemarks (pins) for a KML file to represent impact locations.
Replace the Placemarks in an existing KML file with newly generated pins.
Open the updated KML file in Google Earth Pro.
Switch from Earth to the Moon in Google Earth Pro to analyze the risk corridor.
|
||
| Here is a risk corridor comprising 272 virtual impactors out of 19636 virtual asteroids. The time of the first is 2032-Dec-22 15:17:34.18, and the time of the last is 2032-Dec-22 15:20:48.55. This risk corridor was created on 2025-02-28 using 441 of 468 observations at the epoch Dec 22.635417 [2032 Dec 22 15:15:00 (3:15 PM) UTC] Monte Carlo noise factor of 0.6. using Find_Orb by Steven M. Tilley |
Check to see if the risk corridor is reasonable. If you pick an epoch right after impactors, some of the impactors may look "odd."
Click this link for Bill Gray's "Risk Corridor for Gibbous Moon"
Click this link for Bill Gray's "Risk Corridor for Gibbous Moon"(text).
Monte Carlo simulations, when used for finding the impact probability, may contain inaccuracies, especially if they are run only for a short time. Too much or too little Monte Carlo noise is used, so the impact probability should be considered an approximation for risk assessment.
