Recent commissioning progress has brought the LIGO Hanford 4-kilometer interferometer (H1) close to the project’s long-standing sensitivity goal, a milestone that will further expand LIGO’s view of the universe as the search for gravitational waves continues.

The frequency spectra in the left figure show improvements in the strain sensitivity of the H1 interferometer from year 2002 to present. The curve from August, 2004 rests slightly above the solid trace that represents LIGO’s design sensitivity. This is a logarithmic plot, so moving one vertical division down on the plot corresponds to a tenfold improvement in strain sensitivity. Since the gravitational strain from an astrophysical event is inversely proportional to the event’s distance from the detector, an improvement of one vertical increment on the plot corresponds to a thousand-fold increase in the volume of space that the detector surveys. The expected rate for a measurable astrophysical event then rises by roughly the same factor of 1000.

A gravitational wave signal must rise above the noise floor that is shown by the traces in the figure. Lowering the floor to its present level has been a several-year challenge, requiring LIGO scientists and engineers to develop extremely careful control of the positions and angles of the mirrors and the power and frequency of the laser. This occurs through the mastery of a complicated set of about 100 control systems consisting of high performance electronics and embedded computing power. Subtle interactions often occur among the control systems. For instance, high laser power makes the systems very sensitive to mirror positions, but elevated light power applies extra radiation pressure to the mirrors. This pressure will torque the 10-kg mirrors if the beams aren’t centered on them exactly. Similarly, laser power needs to be high for adequate position control, but increasing the power before the mirror positions are stable will result in saturation of the detector. Understanding these types of interactions in greater clarity and detail has enabled LIGO personnel to improve the overall control of the interferometer, producing a more stable and more sensitive instrument.

The contributions of several noise sources for the H1 interferometer are shown in the figure on the right. Here the noise is given as displacement rather than strain. Above 200 Hz the main source is quantum shot noise. Below 200 Hz, the control systems that are designed to keep a suspended optic in place in both position and angle will also impart small levels of vibration to that optic. Systems that perform position control on the beam splitter and the recycling mirror (auxiliary length loops), angular control on all the mirrors (angular control loops), and actuation on the end mirrors are currently limiting our sensitivity at lower frequencies. These systems provide clear targets for further performance improvements.

LIGO’s noise reduction scheme is similar in principle to what retail consumers experience through the use of noise cancellation headphones. Perhaps the headphone user’s background noise might consist of shop machinery or a passenger jet engine in flight, and the device might lower the audio pressure of the noise by a factor of ten. LIGO contends instead with persistent vibrations of the earth that exceed the effects of gravitational waves by many orders of magnitude, and the attainment of design sensitivity requires displacement reduction factors that range from ten million to ten billion (this is part of the reason that LIGO will not appear as an item in your holiday catalogue).

Lessons learned on H1 now need to be applied to the Hanford 2-kilometer and Livingston 4-kilometer interferometers. LIGO’s improved instrumentation will be deployed for the next science run (S4) early in 2005. Researchers are trusting that the current round of commissioning will result in an S4 data set of unprecedented quality and quantity.