Special Issue : 10/01/2004
A scientific publication by SGF and NEODyS

The history of Hermes

An interview to Andrea Boattini (Spaceguard Foundation)

Andrea Boattini has been looking for Hermes for a long time. He was among the first researchers to think that Hermes was the unusual object recorded on Skiff's images

Question: Can you tell us about your scientific interest for Hermes and how were you involved in the search for Hermes over the years?

A. Boattini: My personal interest for Hermes materialized in February of 2000 when I started a dedicated search effort with Germano D'Abramo who has worked with me in the past few years at the National Research Council in Rome and with Giuseppe Forti, an expert in orbit calculation, who works at the Arcetri Observatory in Florence.
Since we had no access to a suitable telescope to look for Hermes, we decided to search for it on some astronomical archives, namely The Palomar Observatory Sky Survey (POSS) I and II, the UK Schmidt Telescope archive (UKST) and some plates taken with the ESO Schmidt telescope in Chile. Most of this material is available directly on the web as part of the Digital Sky Survey. The work on the archives was not a new one, but it was conducted as part of ANEOPP (Arcetri NEO Precovery Program), an initiative which has led to the precovery of more than 100 NEAs since 1999.

Since Hermes was badly lost, we had to sample its confidence region with about 10,000 orbits using the multiple solution method. To do so we used the Orbfit software which was made by a Group of scientists led by Andrea Milani at the University of Pisa. This number of orbits was reputed a good compromise between good sampling (necessary to deal with non-linear effects caused by close encounters with Mars, Venus and the Earth) and work load: for a proper job, about 1 million orbits would have been necessary.

In order to identify good archive images we cross-checked each orbit with each archive catalogue. The result was a list of images where Hermes, if present on the part of the sky, would have been bright enough to be detected and recognized.

We knew from the start that this would have been a huge work and that it may have not led to success since only about 25% of the orbital solutions had at least one good plate where it could be found. But it was a nice and useful exercise to do during our spare time: a significant amount of the work was done in Edinburgh in Sept. 2000 checking basically all useful plates from the UKST collection which are available at their plate library.

By early 2002 about 95% of all the available plates/films from these collections were checked but with no results. Maura Tombelli, an active amateur astronomer and one of the women involved in minor planet astrometry, provided some help at Arcetri Observatory in 2000 and 2001. In Sept. 2002 the search continued with the help of Fabrizio Bernardi who helped me to continue the search with the Schmidt telescope at Campo Imperatore in central Italy, where we had started a small NEO survey called CINEOS. A month later I was visiting T. Bowell in Flagstaff checking a few films from the Shoemaker collection, but again with no luck.

Question: Can you tell us Hermes' rediscovery story from your point of view?

A. Boattini: On October 15, 2003, I arrived at work in the early afternoon: after checking the NEOCP I noticed 5AF003, a bright target from LONEOS and immediately thought it could be Hermes simply by comparing the location of the new object with Hermes orbital plane, its motion vector and its expected magnitude for this date. Since the new object was particularly bright (at magnitude 14.5), the chances it would have been another asteroid became very remote.

This very simple method for proposing identifications (IDs) is not new: for example, we used it in the recent past to identify two other lost NEAs, 1991 NT3 and 1997 NJ6. And it is, essentially, is the only option available when the astrometric data is not available yet. However, I could not figure out why it was not noticed during the previous lunation. At that point I informed Germano D'Abramo and Giuseppe Forti about the new object. It became obvious that 5AF003 (the original designation) was moving in the sky like a main-belt asteroid during the previous two lunations.

At that point we sent a message to the Minor Planet Center (MPC) and a few other astronomers. In the meantime, Tim Spahr at the MPC had already figured out that it was Hermes at that time (read Tim Spahr's interview). Regretfully, MPEC 2003-T74 (the official announcement by the MPC), which came out only hours after our message, did not report our independent suggestion. But the real bad news came out after S. Chesley and P. Chodas at JPL made the linkage with the 1937 UB: it turned out that Hermes was missed by a previosuly checked UKST plate (J 548) in 1974 by about 10' when it was of mag. 12.7 (read Steve Chesley's interview).

Question: Why was Hermes so difficult to find? Why was there a doubt that 2002SY50 and Hermes could be the same object?

A. Boattini: Hermes was so difficult to find for two reasons: the ephemeris uncertainty became so large after all these decades that it could be located anywhere along its orbit and the detemination of its orbital plane was not too accurate; second, it was very complicated to use the results of a negative search on some images/archives for the benefit of other searches.

The uncertainty in the determination of Hermes orbital plane was the main reason that caused confusion with 2002 SY50. It turned out that the two orbits were very similar as well as their absolute magnitude. The discrimination between these two objects could not fully be clarified until the orbit of 2002 SY50 became so well established to exclude that it could be located on the part of the sky were Hermes was discovered in 1937.

Question: From the technical point of view of an astronomer making observations, why are archive images important?

A. Boattini: Archives are important because they may contain one or more unnoticed images of newly discovered NEAs as well as undiscovered objects, for several reasons:

1) It is the most rapid method for obtaining very good orbits soon after discovery; the orbital improvement could be of several orders of magnitude. This is particularly relevant when we deal with potential impactors because the collision orbital solutions, technically called "Virtual Impactors" can be eliminated by adding a couple of astrometric positions from an old image.
2) it saves a great deal of telescope time, as well as money;
3) it can be a day-time activity, unaffected by the vagaries of the weather.

Question:  Can you tell us about SGF and the SCN? What is your role in discovery, follow up, rediscovery?

A. Boattini: The SCN (Spaceguard Central Node) has served the NEO community for about 4 years providing suggestions about which objects should received highest priority at any time. This service has contributed significantly to reduce the number of NEOs that became lost and to make sure that NEOs with remote collision possibilities received the necessary attention from observers.

Question: Are there other "famous lost asteroids" that you would like or are trying to recover?

A. Boattini: Yes, there is one that has interested me for quite some time: 1984 QY1. This is one of the largest unknown NEOs (with an absolute magnitude of
14.0, about 5 km in size): its orbit is extremely underterminate with an arc of only three days. We did some searching on the old photographic archives, but did not find anything so far. It is very likely that it will be recovered accidentally.