
As of February 2025, prior to any trajectory-altering events, is asteroid 2024-YR4 currently on a trajectory that will result in it impacting Earth’s atmosphere?
Update 2025-02-19 (PST) (AI summary of creator comment): Clarification to Resolution Criteria:
Scenario 1: If no attempt is made to redirect the asteroid and no evidence of any trajectory change is uncovered, resolution will be based on the actual impact outcome.
Scenario 2: If the asteroid’s trajectory is known with confidence but changes (e.g., due to a redirection mission) before a projected impact, resolution will be based on the expected trajectory prior to any changes.
Scenario 3: If the asteroid’s current trajectory is not known with confidence and it is unclear whether a redirection occurs, resolution will be based on the best available information.
@creator Will this be resolved as NO if the asteroid misses the Earth on the first pass in 2032, or will it wait until all possible impacts have passed (the last is in 2047). The first impact was the one which previously had a >1% chance.
Sentry (https://cneos.jpl.nasa.gov/sentry/details.html#?des=2024%20YR4) currently shows a 1.1e-5 chance in 2032, and a 2.2e-5 chance in total.
@DanielTilkin the intention is for this market to resolve based on the anticipated Dec 2032 encounter. After that it will resolve
@DanielTilkin the intention is for this to resolve based on the anticipated December 2032 encounter. After that it will resolve.
To clarify a bit:
There are three possible scenarios:
1) no attempt is made to redirect the asteroid and no evidence of any trajectory change is uncovered. Resolution is based on actual impact
2) asteroid’s trajectory is known with confidence, but changes (for whatever reason, but most likely a redirection mission) before the projected impact. Resolution is based on the expected trajectory before any changes
3) asteroid’s current trajectory is never known with confidence, and is maybe or maybe not redirected somehow. In this very unlikely case I will resolve based on the best information available.