I heard that it's somewhat normal for the probably to rise before dropping to zero. This makes mathematical sense, since decreasing the size of the probable zone makes the earth a larger part of that zone until the zone fully excludes the earth.
My understanding is that it was due to the moon being full which caused them to not have as much of an ability to observe it during the past few estimations. In an article[0] linked from the above link they state:
> Now that skies are darker after a week of limited visibility around full moon, astronomers have resumed observations
The estimates changed a lot because they were being very transparent about the information as they were discovering it.
Not "worrying" in a rational sense (i.e. we should put more thought into the possibility of this collision), but in a human emotional sense ("if NASA engineers can change their minds this quickly, surely we don't know nearly as much as we think, and the universe feels like a scarier place again").
The current odds (0.28% as gnabgid@ cites) are still uncomfortable. But also, most of the Earth's surface isn't a city, so maybe we should be even less worried.
But on the other hand, I do wonder if a city-killer blast in the wrong spot could trigger some kind of nuclear response in a jumpy geopolitical climate. I keep thinking about the time the USSR caught some weird atmospheric effect and nearly launched, save for the skepticism of a lone individual who chose to break with protocol...
I am curious how to map the calculation onto a human impact number. 0.28% chance to hit the entire planet, most of which would have limited impact on humanity. What are the odds for example that this kills 1000 people?
You'd need to start with GIS population data for the planet, a first order answer would look at the surface area of the planet divided out by the principal destructive area size of (say) the Tunguska event, each area weighted by population.
Complications over that, once sorted out, include:
* are whether impact sites are evenly distributed across the planet or, perhaps, weight to more common near equatorial regions .. this would relate to the expected incoming trajectories of chuncky rocks which are maybe more likely in the solar plane.
* chance of impacts in non populated (or low populated) areas having a disproportionate secondary effect that kills people.
eg: Will hitting shallow ocean cause a wave that travels and kills?
Will hitting and destroying the massive water pipes that bring drinking water to New York City kill very few directly but indirectly cause the death and suffering of a million+ ?
I didn't even consider that different parts of the planet may be more or less likely to be hit. Are most of these dangerous asteroids in the normal orbital plane or could something come straight down onto the North Pole from 90 degrees? I assume yes possible, but much less likely.
If you're interested in such things you might enjoy joining your local chapter of the Global Fireball Observatory, an expansion the original Desert Fireball Network.
Scott Manley explains it well in his video: https://youtu.be/Esk1hg2knno
NASA Downgrades the Risk of 2024 YR4 to Below 1% (45 points, 11 hours ago, 54 comments) https://news.ycombinator.com/item?id=43148259
> Now that skies are darker after a week of limited visibility around full moon, astronomers have resumed observations
The estimates changed a lot because they were being very transparent about the information as they were discovering it.
[0]: https://blogs.nasa.gov/planetarydefense/2025/02/19/dark-skie...
Every observation narrows the range of possible trajectories. After the initial discovery, a lot of observations were scheduled.
But on the other hand, I do wonder if a city-killer blast in the wrong spot could trigger some kind of nuclear response in a jumpy geopolitical climate. I keep thinking about the time the USSR caught some weird atmospheric effect and nearly launched, save for the skepticism of a lone individual who chose to break with protocol...
Complications over that, once sorted out, include:
* are whether impact sites are evenly distributed across the planet or, perhaps, weight to more common near equatorial regions .. this would relate to the expected incoming trajectories of chuncky rocks which are maybe more likely in the solar plane.
* chance of impacts in non populated (or low populated) areas having a disproportionate secondary effect that kills people.
eg: Will hitting shallow ocean cause a wave that travels and kills?
Will hitting and destroying the massive water pipes that bring drinking water to New York City kill very few directly but indirectly cause the death and suffering of a million+ ?
All in all though it is extremely complex.
https://gfo.rocks/partners.html https://dfn.gfo.rocks/
They've got camera kits, forums, data, nerds, actual recovered rocks, . . .