Recently, I met scientists from Imperial College London to discuss whether or not climate change in the Artic is increasing the risk of landslide-triggered tsunamis in the UK. One of those very low-frequency, high-impact risks. Wait for it…
During the introductions one of the scientists explained his expertise in ‘impact cratering’ and other violent geologic processes. I joked I would keep this in mind when any meteors came close to earth, but we would stick to modelling tsunamis for now.
Just to prove me naive, a few days later a meteor hit a frozen lake in central Russia. Humbled, at least I know who to call.
The meteor in question was a little bigger than initially estimated by the Russian Academy of Sciences – NASA now reports it weighed 10,000 tonnes and had a 17m diameter by the time it landed, which would make it a 1 in 100 year event.
What’s even more unusual is that it fell near a populated area. What is the probability of that?
Spatial Distribution of Meteor Risks
I looked at global maps of known craters and it struck me that their spatial distribution isn’t uniform. It is somewhat correlated with populated areas. I doubt meteors are viciously targeting earthlings. More likely this is simply an observational bias.
The Earth Impact Database at the University of New Brunswick in Canada lists 183 confirmed sites (soon 184 then) but there’s only one impact known in the deep ocean. Many craters have also been lost to erosion, burial or subduction (one tectonic plate moving under another tectonic plate).
To get a better understanding of impact probabilities, scientists also consider craters on the Moon, observations of asteroids and fireballs. But basically, the risk of impact is uniform across the globe, although the poles might be safer for dynamical reasons.
Land covers about 29% of the globe, and inhabited areas a mere fraction of that. Using impervious surface areas (ISA) as a proxy, around 0.5% of this landmass is classified as urbanised. According to my calculations that makes for a roughly 0.15% probability of a direct meteor strike on a human settlement.
This is reassuring, unless you are dealing with a meteor as big as the one which created Chesapeake Bay 25 million years ago. More recently, the 1908 Tunguska event (50m diameter) flattened trees over a 25 km radius (that’s about the radius of the M25 around London).
According to one local account:
Suddenly in the north sky… the sky was split in two, and high above the forest the whole northern part of the sky appeared covered with fire… At that moment there was a bang in the sky and a mighty crash… The crash was followed by a noise like stones falling from the sky, or of guns firing. The earth trembled.
Fortunately, this was in the middle of the Siberian forest, and nobody was hurt. If something that big happened over a big urban area (or within 25km of it) it wouldn’t go unnoticed.
Can we predict when/where meteors might strike? There may be periods of higher risk, but cycles aren’t well understood. The passage of asteroid 2012 DA14 (50m diameter at just 27,700km from Earth) on the same day as the Russian metor seems to be a coincidence.
There are several asteroid monitoring programmes (amateur astronomers also often find new ones). The orbit of anything big is tracked until it can be predicted accurately. If there is no risk of collision with Earth they move to the next one. But with limited resources, choices have to made.
While 99% of objects bigger than 1 km in size are tracked, about 1% of objects less than 100m in size are known. Ironically, we know more about objects with cataclysmic consequences, but not enough about impacts that could be mitigated by simple warning or evacuation.
As a report by the Near Earth Object Media/Risk Communications Working Group pointed out: “Today, no worldwide disaster-notification protocol of any kind exists: this group is on a mission to change that.”