I just learned the term "third generation gravitational wave detector" in answers to What would a kHz gravitational wave detector look like? (mountains on millisecond magnetars) How would it differ from LIGO/Virgo?

Wikipedia's Gravitational wave observatory; Interferometers lists LIGO and Virgo as 1st generation, and Advanced LIGO and Advanced Virgo as second generation.

Is the difference between the originals and the "Advanced" versions really generational or were these just planned incremental upgrades that were just waiting for funding once the basic systems had demonstrated they worked as planned?

  • $\begingroup$ What does "really generational" mean? And "the plans" for how LIGO works were constantly being updated. It is expected to reach "design sensitivity" sometime soon journals.aps.org/prd/pdf/10.1103/PhysRevD.102.062003 $\endgroup$ Commented Apr 26, 2021 at 12:57
  • $\begingroup$ @DaddyKropotkin I didn't invent the term "generation" in terms of GW detection technology nor assign 1st, 2nd or 3rd to different designs, these classifications are somewhat established. My question is about the appropriateness of saying that the difference between these detectors original configurations and their advanced configurations are really a generational change or not. The two answers posted so far seem to accept "how generational" as a viable question, though interestingly have different positions. Anyway, I've just added the terminology tag to the question to highlight this. $\endgroup$
    – uhoh
    Commented Apr 27, 2021 at 9:26
  • $\begingroup$ Okay, so then the answer is by definition yes, since "generations" is defined as successive upgrades from the original construction to the design construction. That's why i asked for you to clarify, since you're question seemed, as is, trivial. $\endgroup$ Commented Apr 27, 2021 at 12:54
  • 1
    $\begingroup$ I'm just trying to arrive at a better posed question because I don't understand the OP qusetion. A "generation" refers to a suite of upgrades/improvements, I think. dcc.ligo.org/public/0006/P0900255/005/P0900255-v5.pdf $\endgroup$ Commented Apr 27, 2021 at 23:58
  • 1
    $\begingroup$ Cool! Yes, indeed. They basically kept the same chassis of the old tunnels but improved essentially all the technical aspects. $\endgroup$ Commented Apr 28, 2021 at 18:38

2 Answers 2


I think this is largely a matter of what you decide is incremental as opposed to a generational change.

The basic location and overall size and detection principle of the interferometers did not change. An important change to the interferometer configuration was the addition of the signal recycling mirror that allows a certain amount of "tuning" of the frequency response. A homodyne readout scheme was employed instead of heterodyne - and the interferometer is run slightly off the dark fringe as a result. The test mass suspension system changed from being a single to a quadruple pendulum and the suspension wires are now silica rather than metal. The test masses themselves were increased from 11 kg to 40 kg which reduces the low frequency noise significantly.

I believe most of the other changes were essentially upgrades to the materials and technology being used. New mirrors, better lasers, improved vacuum, better seismic isolation etc.

I don't think it is anywhere near the generational change between aLIGO and the proposed Einstein Telescope, which will be in a different location, have a different size, different number of arms and geometry and will be operated cryogenically etc.

However, as pointed out by @mmeent, the improvement in sensitivity is about an order of magnitude in both cases, which is probably what justifies the "next generation" tag for going from LIGO to aLIGO and from aLIGO to the 3G detectors.

  • $\begingroup$ The improvement in strain sensitivity from initial LIGO to advanced LIGO is similar as the projected improvement from aLIGO to 3g detectors. $\endgroup$
    – TimRias
    Commented Apr 24, 2021 at 17:42

I don't feel that ProfRob's answer does the step to advanced LIGO justice.

First, it is important to recognize that LIGO, from its inception, was envisioned as a two-stage project. From Caltech`s LIGO lab:

... and in 1989, Vogt, Drever, Fred Raab, Thorne and Weiss submitted a joint Caltech/MIT proposal for LIGO construction to the NSF.The proposal envisioned building LIGO facilities at two sites, and then operating in them a pair of initial interferometers based on proven technology, with a sensitivity where gravitational waves might be detected, followed by advanced interferometers based on more advanced technology, with a high probability of detecting waves. This two-stage approach has been essential to LIGO’s 2016 success. The technological leap from prototypes to advanced interferometers was too great to be carried out in a single step.

Building an Michelson interferometer isn't hard, it is literally done by first year physics students around the world. The hard part is sufficiently isolating the interferometer against environmental noise that the minute signal produced by a gravitational wave can be isolated.

By 2010, initial LIGO had reached the absolute limit of what was possible with the technology upon which it was built, reaching a strain noise of $1.5\times 10^{-22} Hz^{-1/2}$. To do better would require an almost entirely new instrument. Advanced LIGO was designed to give an order of magnitude improvement in strain sensitivity over intial LIGO. Achieving this required changing almost every part of the instrument, including heavier test masses, more powerful lasers, and a completely redesigned suspension system. Essentially the only parts that carried over from initial LIGO were the facilities in which the instruments were installed, and parts of the vacuum system (which were designed with the ultimate goal of building advanced LIGO in the first place. Construction of advanced LIGO started in 2008 and took over 6 years to complete.

In this sense the step between initial and advanced LIGO can be best compared to the relationship between the LHC and its predecessor LEP, which had occupied the same tunnel at CERN before it.

As work on advanced LIGO started, people also start thinking about would be necessary to make another order of magnitude step in strain sensitivity. As the goal was to make a similar step from advanced LIGO (and advanced Virgo in Europe) as had been made from initial LIGO and Virgo to their advanced versions, these detectors where dubbed "third generation" (3G) detectors.

There is currently two 3G detectors proposed. The Einstein Telescope in Europe and Cosmic Vision in the US. Einstein Telescope is currently the furthest in planning and is envision as a triangular interferometer with 10km arms build underground. Cosmic Vision is closer to a scaled up version of LIGO with an L-shaped layout with 40km arms.

The realization of these 3G detectors is still very far off. In the meantime more incremental improvements to the current ground based detectors are planned. First of all, advanced LIGO is not yet at its full design sensitivity which it is expected to reach in its 4th observation run. Beyond that further upgrades to LIGO labelled "A+" have been approved.

People have also considered what would be the maximum sensitivity that can be squeezed out of the existing LIGO facilities, if one were to build an almost entirely new instrument (again) using all the technology that is foreseen for 3G detectors, including moving to cryogenics to reduce thermal noise (as is currently being applied in the Japanese KAGRA detector). This hypothetical design is called "LIGO Voyager". It is sometimes also referred to as a third generation detector, although since it wouldn't be quite at the level of Cosmic Explorer and the Einstein Telescope it is now more commonly referred to as a "2.5G" instrument.

  • $\begingroup$ Isn't aLIGO expected to achieve design sensitivity sometime this year or next? $\endgroup$ Commented Apr 27, 2021 at 12:55
  • $\begingroup$ @DaddyKropotkin It should indeed be O4 (O5 will be the A+ upgrades). Unfortunately, O4 seems to be pushed back to the second half of 2022 (at least). $\endgroup$
    – TimRias
    Commented Apr 27, 2021 at 13:03
  • $\begingroup$ Neat! Also, is LISA not a third gen detector? And what about the many decihertz detectors being planned such as TianGO? $\endgroup$ Commented Apr 27, 2021 at 13:10
  • $\begingroup$ No, normally the term "3G" is reserved for ground based interferometers. See also: gwic.ligo.org/3Gsubcomm/documents/… $\endgroup$
    – TimRias
    Commented Apr 27, 2021 at 13:12
  • $\begingroup$ I listed all the improvements that had been made between aLIGO from LIGO. I don't see that I missed anything significant out other than the length of time it took. I initially voted to close this question as a matter of opinion. So it has proved. $\endgroup$
    – ProfRob
    Commented Apr 27, 2021 at 14:14

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .