NEO News (11/9/01) Impact frequency again

Dear friends & students of NEOs:

One of the persistent topics in NEO studies is the frequency or rate
of impacts of differing sizes. This impact rate depends on the
population of near-earth asteroids (NEAs) and occasional comets, and
the dynamics of their obits that ultimately may bring them into
collision course with our planet. This edition of NEO News notes a
recent paper from the Sloan Digital Sky Survey team at Princeton,
which reports on the implications that the Sloan Survey data may have
for NEA impact rates.

Before discussing these results, it may be worth noting that the
exact statistical frequency, while of interest to asteroid
scientists, is not directly relevant to the present impact hazard.
Impacts by NEAs of any size are exceedingly rare, from the 5-megaton
limit of atmospheric shielding up to the hundreds of millions of
megatons associated with mass extinctions. Statistically, no impact
is to be expected within a human lifetime. The real issue, therefore,
is whether against the odds there is an NEA on an impact trajectory.
If there is such a threat, we want to identify it and deal with it.
That is the purpose of the Spaceguard Survey. The important issue is
not the average frequency of impacts but whether we can expect an
impact within our lifetime, or the lifetime of our grandchildren.

Still, the total population of NEAs and their average impact
frequency are often discussed. Sometimes these issue are used as a
“hook” to attract attention to work that might otherwise be
overlooked by the media and the general public. This week there has
been considerable press interest in the conclusion that the new Sloan
Survey results have reduced the estimated frequency of impacts and
therefore that we were safer than had been previously thought. It is
true that the estimated frequency for impacts by 1 km and larger NEAs
has dropped in recent years, but not by more than the probable
uncertainties, which have been given by most scientists studying the
problem as about a factor of 2 or 3. Since the mid-1990s, the most
common estimate has been that the Earth is hit by a “civilization
threatening” impact (by a 1.5-km-diameter asteroid) about twice per
million years, which is equivalent to a 1-in-5000 chance per century.
But it is hard to tie down such estimates, in part because there is
also a range of uncertainty as to what constitutes a civilization
threatening impact, spreading over at least a factor of two in
asteroid size (from 1 km to 2 km diameter).

Alan Harris of JPL has provided several estimates of the impact
frequency for NEAs with absolute magnitude H less than 18
(approximately corresponding to a 1-km asteroid with average
reflectivity or albedo). For these estimates he used various recent
values of both the population of NEAs and their dynamical lifetime
(how long they orbit the Sun, on average, before they hit the Earth).
These estimates for the chances of an impact in the next century are:
1 in 8600, 1 in 7100, 1 in 4800, and 1 in 4000. Expressed in terms of
NEAs 1.5 km in diameter, the estimated impact frequencies are 1 in
12000, 1 in 14000, 1 in 10500, and 1 in 8800. The range in these
values, which us about a factor 3, is probably representative of our
uncertainty in the impact frequency.

It should be noted that all of Harris’s numbers are based on the
directly observed population of NEAs, of which nearly 1500 are now
known. In contrast, the Sloan Survey measured main belt asteroids and
assumed that the main belt population had the same size dependence as
the NEAs. It is by no means clear that this assumption is correct, or
that main belt asteroids are a satisfactory surrogate for NEAs.

I would also like to comment on the statement in the Sloan news
release that “the new impact risk estimate, like most previous ones,
relies on assumptions about a single event 65 million years ago when
a 10-kilometer asteroid collided with earth and killed the dinosaurs.
The researchers assumed that such impacts occur on roughly 100
million-year intervals and used that statistic to calculate the
impact odds for the more common asteroids of smaller sizes.” As one
of the first to publish NEA impact frequencies (with Clark Chapman,
see our Nature paper in 1994), I note that we made no such
assumption. Nor, to my knowledge, have any previous estimates
involved any assumption about the frequency of KT-size impacts.
Indeed, we have no way from a single example 65 million years ago of
estimating the average frequency of such impacts. To my knowledge,
this Sloan impact frequency estimate is the first to depend on such
an assumption.

The Sloan Survey is an exciting astrophysics project with great
promise in a number of fields of astronomy. It has already
contributed to our knowledge of the main belt asteroids. But it has
not measured NEAs, and it has not made an independent estimate of the
impact frequency. It is therefore probably premature to conclude from
these observations of the main belt asteroids that we should revise
our estimates of the impact frequency of NEAs or of the magnitude of
the impact hazard.

David Morrison