Special about 2002 CU11
The problem with 2002 CU11
by Andrea Milani, Giovanni Valsecchi and Maria Eugenia Sansaturio -
Copyright Tumbling Stone 2002

The asteroid 2002 CU11 was discovered on 7 February 2002; after the data were made public by the MPC, the CLOMON software robot monitored it and since 14 February identified it as a possible impactor; this information was posted on the "risk page" of the Near Earth Objects Dynamic Site (NEODyS) (http://newton.dm.unipi.it/neodys/).  Successive recomputations, using later observations, confirmed the existence of an impact risk, in particular for the year 2049 (impacts in other years are also possible). According to the last computation (based on observations up to 23 march 2002) and to our best estimate of the impact probability, done with the experimental CLOMON2 software, the 2049 impact has a probability of about 1 in 9,300.

2002 CU11 has an estimated absolute magnitude (dict.) 18.3 thus, using the average albedo and density of the Near Earth Asteroids (NEA, see dict.) and the properties of its orbital intersection with Earth, the best estimate of the energy which would be released in case of an impact is 53,000 Megatons (MT, see dict.).

We have recently developed a metric to be used to assess the relevance of such a risk in comparison with the background risk by all Near Earth Objects (NEO, see dict.), both known and unknown, the Palermo Technical Scale, PTS ( see T.S. issue number 11 to know more about the Palermo Technical Scale , also see the scientific article by Chesley et al., 2002).
According to our standard model the risk of impact by 2002 CU11 in 2049 amounts to 45% of the risk, between now and 2049, for an impact with energy 53,000 MT or larger; this corresponds to a value -0.35 in the logarithmic PTS. The PTS does not replace, but complements the simpler (and thus easier to use in communicating with a larger public) Torino Scale (http://impact.arc.nasa.gov/torino/index.html), which is, in this case, 1.

A scientific breakthrough

by Andrea Milani, Giovanni Valsecchi and Maria Eugenia Sansaturio -
Copyright Tumbling Stone 2002

To put the detection of the 2002 CU11 impact possibility in context,we need some historical perspective.

Almost exactly 4 years ago an ambiguous announcement of a possible asteroid impact (by 1997 XF11 in 2028) raised the issue that the scientific community lacked the know how necessary to assess the impact risk, for low probability but nevertheless serious cases. This challenge was taken by some as an opportunity for controversy, by others as a job which needed to be done.

Almost exactly three years ago the announcement by Milani, Chesley and Valsecchi of a possible impact (by 1999 AN10 in 2039) with a probability of 1 in 1,000,000,000 (one in a billion) was met with an enormous interest, some disbelief and very aggressive polemics. We think that the time has shown the relevance of the conceptual breakthough on which that announcement was based... click here for the full article

Sentry: a monitoring system from NASA to the web
by Livia Giacomini - Copyright Tumbling Stone 2002

It took two years of hard work, but finally, on March the 12th, NASA announced that Sentry, its new automatic asteroid impact monitoring system, was beginning to be operated out of Jet Propulsion Laboratory. Sentry was built largely by Drs. Steve Chesley and Alan Chamberlin with technical help from Paul Chodas.
To be more precise, Sentry is a highly automated system, designed to help scientists better communicate about the discoveries of new, potentially threatening Near Earth asteroids (NEAs) and their follow-up observations. While completely independent from other scienitific teams, it is in constant communication with the NEODyS CLOMON impact monitoring system, operated in Pisa, and researchers from the two systems are cooperating to check and improve their results.

But how does this new tool work? First of all, it is important to say that, as it usually is today in the scientific world, Sentry is a technological tool which implements web based technology.
In fact, all the data calculated is posted daily on the web, in the public Sentry risk page (http://neo.jpl.nasa.gov/risk/) which constitutes a very important tool for research teams around the world as well as for amateurs.

The procedure to determine these parameters is quite clear: data about NEAs is drawn each day from the Minor Planet Center in Cambridge, and it is used to update the orbits, Earth approaches, and last but not least, Earth impact probabilities.
Based on these calculations, several asteroids are added monthly to the web page which can be considered as an updated list of "dangerous" objects which must be followed up. These objects are in fact characterized by two aspects: first, they have orbits that can bring them close to Earth, and second, only a limited number of observations are available for each of them, so that their trajectories are not well-enough defined.
Normally, these NEAs are listed on the Sentry Impact Risk page only to be removed to a second no-risk page soon afterward. But pay attention, this procedure doesn't mean that there was an error while calculating the risk. In fact, as new observations become available, the knowledge of the object's orbit is improved (its region of uncertainty becomes smaller) and its risk promptly recalculated. The most likely outcome of the all procedure is that the object becomes harmless and is therefore removed from the risk page.

But how does the public outreach of the Sentry system work? A first characteristic of the Sentry risk page is that the risk is presented using both the usual Torino Scale and the new, more technical, Palermo Technical Scale (click here to go to T.S. issue number 11, to know more about the two scales).
Thanks to the two scales, the risk of a single object is compared to the so-called background level (which is the average risk from the entire NEO population). A Palermo Technical Scale value less than zero and, in most cases, a Torino Scale value of zero, indicates that this risk is below this background level, and that the event can be considered only of academic interest, and not deserving public concern.
On the other hand, on the very rare cases when events have a Palermo Technical Scale value greater than zero, a Technical Review is requested to verify the calculations before the prediction is placed on the Risk Page.
But Sentry is also a scientific instrument, meant to give to scientists all data and information about NEAs. For this reason, independently from the associated risk, for each object of the Risk Page there is a separate page providing more detailed technical information.

But let's come down to mathematics. The real question is: how are the Earth impact probabilities for near-Earth objects calculated? Every day, observations and orbit solutions are received from the Minor Planet Center and once an object has been classified as a NEA, and as soon as enough observations have been collected, an orbit determination process is used to find the orbit which best fits all the observations.
But how is this done? The main idea is that an object's orbit follows some equations that take into account the gravitational attraction of the Sun, the planets, the Moon, and the three largest asteroids, Ceres, Pallas, and Vesta. Given an observed, initial position of an asteroid, its further positions can be computed solving these equations of motion. The difference between these computed values and the actually measured ones are called observation residuals. The overall orbit of the object (and therefore the six orbital parameters that characterize it) is determined iteratively adjusting the calculated and the observed positions until the sum of squares of all the observations residuals reaches a minimum value (this mathematical procedure is called minimum squares fitting, see this issue of T.S. to know more).
The final result of the orbit determination process is called the nominal solution. Of course, slightly different orbits may still fit the observations and this set of orbits lies within what is called the uncertainty region, as all the points inside the region are called virtual asteroids.
Whenever new optical or radar observations become available, automatic updates of this orbit are calculated, giving obviously priority to the objects that seem more dangerous. It is therefore clear that, as new observations of the object are made, the nominal orbit can change and its region of uncertainity can become smaller and smaller.

Orbits and regions of uncertainty are fundamental to make the evaluation of the real risk associated with the asteroid, or in other words, the impact probabilites.
In fact, once the nominal orbit and its associated uncertainty region have been estimated, Sentry can simulate the object's motion in time for up to 100 years. This is done to determine its close approaches to the Earth, published in the Earth Close Approach Tables together with the relative impact probability. This projection in the future of the uncertainty region is not a simple task from a mathematical point of view and it is not always possible. A first mean to achieve this goal is to compute these parameters by projecting the uncertainty region to the close approach time via so-called linearized techniques. Since these techniques lose accuracy when the uncertainties become large, close approaches may be calculated up to decades into the future for objects with well-known orbits, but only a few months or years for objects with poorly known orbits.
On the other hand, Sentry also wants to estimate long-term possibilities of impact for objects with poorly known orbits. To estimate this risk it uses more sophisticated non-linear methods, which are integrated to linear means whenever the uncertainties in a close approach prediction are large.

Sentry, together with other scientific services such as NEODyS, is just one example of the great efforts that every day is been made and must be made to improve our knowledge of asteroids and NEOs. In this optic, by the year 2008, NASA has a congressionally mandated goal to find 90 percent of all Near Earth Objects larger than 1 kilometer. Of them, only about 500 have been found. An estimated 500 or so, still remain undiscovered.

Related Links:
To know more about Sentry and how it works:



Back to
current issue