After the success of DART, we should fund a DART variant to sit in cold storage for a quick asteroid response launch.
6 years isn’t necessarily a long time to develop and build a program like this from scratch. And deciding on every potential impact whether to start funding a deflector will cut even further into development time.
> Find out why the impact probability is changing over time, and why it is likely to fall to zero:
If we can make (without new data) predictions about how future predictions (once we have more data) are likely to differ from the current prediction ... then why can't we systematically incorporate that into the current prediction?
I understand that the size/shape of the 'risk corridor' is related to our current uncertainty given the observations so far. But if we expect the corridor to both shrink and shift to exclude earth, that implies we have a rather informative prior about future observations, which isn't "baked in" to the current corridor estimate. Right?
The 2.2% is not a "flipping a coin" style probability. It is a "we lack complete knowledge" style probability. So the odds that it drops to 0 with better observations are 100%-2.2% = 97.8%, which is pretty likely. These odds are the odds systematically incorporating all knowledge.
I think the line about changing probabilities is there because us humans are bad at understanding probability.
"The chances that Lucifer's Hammer would hit the Earth head-on were one in a million. Then one in a thousand. Then one in a hundred. And then ... even less."
6 years isn’t necessarily a long time to develop and build a program like this from scratch. And deciding on every potential impact whether to start funding a deflector will cut even further into development time.