The LIGO
Web Newsletter
Front Page Hanford Livingston MIT

LIGO Caltech News

Weaving A Network

Weaving A Network

- Contributed by Gary Sanders

One by one the gravity wave antennas dotted about the globe are blinking on. One by one. In Tokyo, the 300 meter TAMA has been running for two years. GEO, the 600 meter interferometer outside Hannover, Germany, has now bounced light up and down a double pass along one of its arms, forming a 2400 meter optical cavity. The American LIGO has run its 2000 meter interferometer in Hanford, Washington, in its final optical configuration, although at low laser power in this early stage. And the second LIGO interferometer, the 4000 meter system in Livingston, Louisiana, achieved servo-control of the recombined system only last month. The sister 4000 meter detector at the Hanford site is in a late stage of installation, due to be operated later this year. And then there is the Virgo interferometer outside Pisa, Italy, running its laser and input optics, preparing for a short arm interferometer test this year, while construction of its 3000 meter arms is proceeding apace.

One by one these delicate machines are stirring to life. Mirrors are swinging on slender wires. The subtle control algorithms to hold them "floating" in inertial reference frames are being tuned, tested and refined in the quiet of late-night engineering runs. Sample data, harboring no hope at this stage of uncovering gravity wave signals, is being collected to gauge the antenna noise signals as well as to train an entire community of apprentice gravity wave hunters. A few years from now these antennas will be on guard around the clock, around the globe, listening for faint signals from around the universe.

A signal sensed by one antenna would likely appear in the data from another. The LIGO, Virgo, GEO, and TAMA teams could sift their data and reveal any observed gravitational wave candidates to the other teams for confirmation by each antenna's data set. In effect, any team that believes it has sensed the subtle waves would turn and ask the others the gravity wave equivalent of "Did you hear that?" With so many sensitive detectors, what could be better than training all these ears on the sky?

There is a way to use these antennas even more effectively. We could combine the data sets and analyze the information together. Use the full, discriminating power of each set in correlation with the others at an early and fundamental level. Instead of listening separately with each detector, we could operate the combined array of detectors in a single phased array. To phrase it visually, the Earth could be like a globe with six ears, all trained on a quiet field of sky.

We can test for ourselves how we hear better with two ears working together rather than with only one at a time. Plug up one ear and then listen for a very faint sound, perhaps your thumb and fingers rubbing together. Listen again with only the other ear. Now listen to the same sound with both ears clear. Ah, that last way is surely the best. It is obvious a faint sound is better detected with both ears operating together. Our brain provides a great advantage, correlating the signals into a stronger, more coherent whole.

Similarly, we could hook all these antennas together via cables (the internet) to a combined analysis computer system. The computer would provide the analogue to our brain's processing of binaural signals. But we do not need to do a unified correlation simultaneously as we observe. Humans correlate in real time, on the fly. Then again, our species' survival often hinges on good hearing. Interferometers, our gravity wave "ears," are not listening for any preying tiger sneaking in the jungle. We have the freedom to form the combined correlations later, in a more relaxed manner.

How then do we convince the teams to pool their data, to work elbow to elbow and, to quote Adalberto Giazotto, Virgo's project leader, to "operate their machines as a single machine." Aren't these teams in a bit of a race? Aren't they overworked and struggling to commission these machines of unprecedented complexity with scant time for bonding and collaboration? Aren't the data sets from each detector written in different formats, making correlated data analysis akin to a choir singing at the tower of Babel, from music sheets in entirely different languages? Briefly, the answers are: In a Race? "Yes, but they had better get smart and realize that this difficult effort could only profit from collaboration." Overworked? "Yes, but whoever said this prize would be won easily!" And, separate formats? "No, we had the foresight to adopt a common data format years ago." That last point especially shows sound judgement. It shows we were prescient enough not to make this correlation harder than it needs be.

So we have realized for some years that this opportunity would arrive. It was fine to work separately while we were each constructing. But the data is not too far over the horizon. And a few years from now, all the interferometers will be operating. And the biggest of them, LIGO and Virgo, will arrive separately at roughly comparable sensitivity. And the next step will be...

A couple of years ago in this newsletter, we described a meeting of the Gravitational Wave International Committee (GWIC) in which project leaders from all the teams agreed to form a working group to study the technical issues involved in combined data analysis. Virgo chief Giazotto has since organized two working group meetings, and progress has been made.

The teams have also engaged more broadly in research and development of technical improvements to the interferometers, anticipating the time when more sensitive instruments can be fielded. There has been considerable inter-project collaboration in these endeavors.

GEO and LIGO have collaborated in detector R&D and in proposing the Advanced LIGO detector. These two teams have also drafted an agreement and formed joint teams in which GEO and LIGO members work together on data analysis for LIGO engineering runs later this year. We hope the GEO instrument will also run in coincidence with our upcoming fourth engineering run this autumn.

Additionally, Virgo and LIGO over the last several months have discussed moving toward a closer collaboration in detector R&D, and in data sharing and joint analysis. Meetings have been held in Lyon, France, and here in Pasadena to discuss these topics. A particular R&D focus is the development of fine mirror coating technology in Lyon, by the SMA/IPNL Laboratory coating the Virgo mirrors. The development is needed for Advanced LIGO, and would be useful in a next generation European detector as well.

On Easter Sunday, I found myself at the Los Angeles airport bound for Pisa and the nearby suburb of Cascina, where the Virgo instrument is rousing to life. David Shoemaker of the MIT/LIGO group, and a leader of our R&D program, joined me on this trip to conclude a coating development agreement. Armed with a technical plan and the draft agreement from an earlier Caltech meeting with Jean-Marie Mackowski of the Lyon laboratory, David and I were traveling to work out any fine details, and to proceed to draft the first text of an agreement on joint Virgo-LIGO data sharing and combined data analysis.

Clearly the time was right for this arrangement. A good omen peered out at me from a magazine on the news rack in Rome's Leonardo da Vinci Airport. (A great scientist, that Leonardo!) The Italian version of Scientific American, "Le Scienze," announced with dramatic cover art, "Cacciatori di ondi gravitazionali," or "Hunters of gravitational waves." Inside, I found myself quoted--well, translated--as saying, "Con un solo strumento e difficile convincere la gente che e stato veramente rilevato un segnale". Even here in the Italian press, there was notice that more than one detector was needed to accomplish our goal. At least I hope that is what I was saying.

On I flew to Pisa, historic site of Galileo's reputed gravity experiments at the celebrated Leaning Tower. I landed at Pisa's Galilei International Airport. The Italians, with scientists on their airports, with scientists, artists and teachers on their currency, set quite an example for us all. It was no wonder this was my destination.

The Virgo Observatory in Cascina, Italy. Fetched at the airport by David and Virgo's other project leader, Alain Brillet, of the observatory in Nice, we traveled out to the Virgo observatory site, pictured at right. I had not been to Cascina in several years, so much of what I saw was new to me. The flat, green alluvial plain held two, long, blue metal covered arms, three kilometers long. Mid stations were evident, where sections of the beam tube are being brought into the tunnels for welding to the growing arms. Only the previous week, Virgo had vacuum tested the first 300 meter section of the tube, finding no leaks and surpassing their vacuum pressure goals. Another 600 meter was installed, about to be joined to the assembly.

At right: The Virgo detector takes shape in Cascina, near Pisa, Italy.

We viewed the Virgo laser and the long mirror assembly dubbed a Mode Cleaner. This optical assembly cleans and prepares the very pure and stable laser beam needed for the interferometer. We viewed the large mirrors installed in the vacuum chambers through a video link peeking through small windows.

Above all, David and I enjoyed the familiar feeling that we knew from LIGO's Hanford and Livingston sites. There is a palpable sense of a powerful tool coming to life. There is still so much to be installed, much not yet working. But there is the shape, the arc, the breath of life, the sense of emergence. We both felt very lucky to be present at this moment.

The next morning we met in the Virgo conference room with representatives of the funding agencies: of Virgo, of the host European Gravitational Observatory (yes, it is abbreviated EGO), and of the Lyon laboratory. After a day discussing agreement text and editing, we arrived at an agreement on common work for the mirror coating, as well as on principles to include in the agreement on joint data sharing and analysis. A bit of homework and and a few email exchanges should lead to final agreements quite soon.

The author stands with his Italian colleagues.

Above: The author stands with his European colleagues. From left to right are Jean-Marie Mackowski of Lyon; the author; Angelo Scribano, Director of EGO; Alain Brillet from Nice; and Adalberto Giazotto of Pisa. David Shoemaker of LIGO/MIT is behind the camera.

In about two months, LIGO will send carefully prepared samples of both fused silica and sapphire mirrors to Lyon for the first test coatings of a new generation of ultra-fine mirrors. The development program will last a couple of years and should lead to knowledge of how to produce the large mirrors needed for the future.

In roughly the same period, scientists from LIGO and Virgo should begin the first exchanges of data from the environmental monitor instruments at the several LIGO and Virgo observatories. Measurements of seismic vibrations, electrical and magnetic fields, and a myriad of other influences affecting these instruments will be exchanged and correlated by data analysts. Later this year, when LIGO carries out additional engineering runs in our series of progressively more complete interferometer operations, we will schedule a period when parts of the Virgo system are also operating, and collect data in time coincidence for the first LIGO-Virgo combined data study.

Our meeting in Cascina marked a great step in building a global array of gravitational antennas. Of course there is much yet to learn. How do we operate the antennas with the needed sensitivity? What is the optimum way of pooling the data and sifting for the signal? How do we build a unified team of scientists from the various projects to jointly agree on the results and to publish together? These are a few of the challenges ahead. But the vision is present and inspiring, and the first steps are being taken. One by One.