One of the most important qualities of science is its openness and transparency: the fact that researchers are obliged to provide access to their data and share resources-- reagents, samples, cultures, etc.. (There are, of course, discipline- and specialty-specific rules that define how much you have to share, when, and with whom; and such gifts also oblige the recipient to reciprocate.)
But openness can sometimes come as an unpleasant surprise, as this post to the Risks list shows:
Planetary Astronomer Michael Brown, one of the co-discoverers of various Kuiper Belt Objects, including Sedna, the really distant one, recently announced the discovery of a Kuiper Belt Object even larger than Pluto. His web-page indicates why he released the information about the discovery earlier than planned....
He became concerned late in July, after he had learned that the computers that controlled the telescopes his team used for their observations kept publicly searchable logs of where the telescopes had been pointed. (From his description it sounds to me as if these logs must also contain a code for what they were looking at.) Brown also realized that they had used some of their codenames in the publicly available abstracts for some upcoming talks. A call to the Minor Planet Centre revealed that someone had recently used a tool the MPC provides to plot the location of his team's tenth planet for that very night! A hurried press conference followed.
Brown himself elaborates on his Web site:
As has been widely reported in the press, the announcement of the new planet was made in a rather hasty manner because of fears that our discovery was going to be made public by someone who had hacked a web site and gained access to information about where the object is. The details are a little more complicated than this, the terminology can be debated ("hacked?" "sleuthed?" "stole?" "stumbled across?") and not all are 100% clear to me, but here is a reconstruction of the events that lead to the announcement as best I can discern them. Some aspects remain mysterious.
In mid-July short abstracts of scientific talks to be given at a meeting in September became available on the web (for example, here). We intended to talk about the object now known as 2003 EL61, which we had discovered around Christmas of 2004, and the abstracts were designed to whet the appetite of the scientists who were attending the meeting. In these abstracts we call the object a name that our software automatically assigned is, K40506A (the first Kuiper belt object we discovered in data from 2004/05/06, May 6th). Using this name was a very very bad idea on our part! Unbeknownst to us, some of the telescopes that we had been using to study this object keep open logs of who has been observing, where they have been observing, and what they have been observing. A two-second Google search of "K40506A" immediately reveals these observing logs. Ouch. Bad news for us. From the moment the abstracts became public anyone on the planet with a web connection and a little curiosity about this "K40506A" object could have found out where it was. Anyone on the planet with even a modest-sized telescope could then go find the object and claim a discovery as their own....
By Friday morning it occurred to me that once someone knew about the web site where the information on where the telescopes we had been using had been pointing it would take only a little more effort to carefully peruse this web site to see if we had been looking at anything else moving in the sky. At this point I contacted Brian Marsden at the International Astronomical Union's Minor Planet Center (MPC) by email, told him confidentially about the two objects that we had not yet announced (now known as 2003 UB313 and 2005 FY9), expressed my concerns that someone may be able to nefariously find our data and attempt to claim credit for discovering these objects, and sought his advice. His chilling response came less than an hour later: someone had already used a web service of the MPC to use past observations of an object to predict locations for tonight. The past observations were precisely the logs from the telescope we had used! The culprit and not even bothered to change the names that we used (K31021C for 2003 UB313 and K50331A for 2005 FY9). At this point we had no choice but to hastily pull together a press conference which was held at 4pm on the last Friday in July, perhaps the single best time to announce news that you want no one to hear.
All of this came about because of the perfect confluence of three factors: we used our actual code name in publicly available abstracts (dumb on our part), we assumed that no one would piece together information from the internet and figure things out (naive on our part), someone with astronomical knowledge was willing to go to some effort to obtain our data (unethical on their part). It's true that the information was available without breaking into any sites. It's also true that sometimes I don't lock the door to my house. I hope that people don't think it's therefore OK to come in and take my stuff.
Of course, there's a long history of priority disputes in planetary discoveries: claims for priority of the discovery of Neptune were hotly debated in the early 19th century, pitting French and British (actually, mainly a circle of Cambridge-trained and -based) astronomers.
Technorati Tags: astronomy, science, security
Recent Comments