Can LISA Hear Black Holes If Its Software Disagrees on a Minus Sign?

Imagine three spacecraft in a constellation, exchanging laser beams. Their mission is to feel gravitational waves. Before data from such a mission can be analyzed consistently, scientists have to agree on what their numbers mean.

A new technical document, the SGS Conventions Document by Quentin Baghi and colleagues, lays out that shared language for LISA, the Laser Interferometer Space Antenna, a space-based gravitational-wave mission. It is not a detection paper. It is a rulebook for the Distributed Data Processing Centre, the distributed centre that will simulate, process, and analyze LISA data. The paper is an arXiv preprint in astro-ph.IM.

A Rosetta Stone for LISA Data

LISA's science targets include merging massive black holes, extreme-mass-ratio inspirals in which a small object spirals into a much larger black hole, and galactic white dwarf binaries. It may also search for a stochastic gravitational-wave background. The DDPC spans many institutions. The document aims to provide standards for exchanging information among coordination units and sub-groups. Without such standards, one team's waveform could use a different sign for a signal, another a different Fourier convention, and a third a different definition of a binary's orientation. The result could be inconsistent data products.

One basic choice is the Fourier transform, the operation that turns a time series into a spectrum of frequencies. The document defines a transform using the exponential e to the minus i2πft. It specifies a one-sided power spectral density with a factor of one-half, and an amplitude spectral density as its square root. LALSuite follows the same convention, but numpy and Matlab differ by normalization, and FFTW differs in the inverse transform. LISA picks one path.

Other choices are equally basic. The metric signature is mostly plus, written as minus, plus, plus, plus. Spacecraft are indexed clockwise as seen from ecliptic north, with spacecraft 1 as the reference spacecraft; the document notes this reference-spacecraft definition still requires confirmation. The phase of a beatnote, the oscillating signal produced when laser beams interfere, is defined as remote minus local. In the science interferometer, that sign choice is corrected during early L0-L1 processing. A flipped sign would make an increase in optical path length look like a decrease, reversing the inferred length change.

LISA data will be described in several reference frames. The document adopts the International Celestial Reference Frame, fixed by distant radio sources, for equatorial sky coordinates: right ascension and declination. It also defines the barycentric mean ecliptic frame at J2000, whose axes are tied to Earth's mean orbital plane at the year 2000. Transformations connect the two. The paper also sets the cosmological backdrop: Planck 2015 parameters implemented through astropy, with distances given as luminosity distances.

Frames, Eccentricity, and When Parameters Were Measured

More subtly, the paper defines several source frames for binary systems. One frame aligns with the Newtonian orbital angular momentum, another with the total orbital angular momentum, another with the total angular momentum, and another with the spin of the most massive object. There are also frames used by existing software: LAL and FEW. Each is convenient for different sources. But the document is candid: these frames depend on the binary's trajectory and are gauge-dependent. In general relativity, gauge means tied to coordinate choices, not directly observable. Efforts are underway to define frames from the waveform alone, using only gauge-independent observables.

Eccentricity is simple for a planet around the Sun. It is less simple for two black holes spiraling together in general relativity. The document adopts a definition based on the gravitational-wave signal itself, using the frequency of the dominant 2,2 mode. It defines an eccentricity and a mean anomaly that reduce to Newtonian definitions when the bodies are far apart. This avoids some coordinate ambiguity.

The paper also recommends reference times for different sources. For extreme-mass-ratio inspirals, the reference time can be when the orbit reaches minimum eccentricity. For massive black hole binaries, parameters are specified at a reference frequency typically chosen during the inspiral; the paper warns that very high-mass systems can locate that frequency close to merger or ringdown, so the user may need to adapt it. For stellar-mass black hole binaries, which can be long-lived in LISA's band, the start of LISA observations is a natural reference, although the paper suggests remaining flexible. For galactic white dwarf binaries, the orbital frequency and its derivative at the start of observations are the natural handle. These choices matter because a source frame is fully defined only with a reference time.

Why TDI Needs a Dictionary—and What the Document Does Not Claim

LISA's lasers have intrinsic frequency noise that would otherwise overwhelm a gravitational-wave signal. Time-delay interferometry, or TDI, solves this by combining measurements taken at different times along the constellation's arms. The combinations are designed so that laser noise cancels while gravitational-wave signals survive. The document defines delay operators and several families of combinations: Sagnac, Michelson, and orthogonal channels named A, E, and T. It also warns that the so-called null channel T has suppressed gravitational-wave sensitivity at low frequencies only under idealized assumptions. Small mismatches in the light travel times can spoil that nulling, and the paper notes the property does not hold in a realistic instrumental setup with the above definitions.

These definitions are not academic housekeeping. They determine how data from three spacecraft are turned into something a scientist can compare with a waveform. The paper is meant to be a living document, updated as the DDPC project evolves.

What does the document not claim? It does not report a detection. It does not prove that any particular waveform model is correct. It is a preprint and a technical reference, not a mission result. Some conventions still require confirmation, such as the definition of the reference spacecraft. Gauge ambiguities remain, and future work will need to extend some definitions to higher-order models. The document also sets a reference epoch for time stamps: January 1, 2035, at midnight in Barycentric Coordinate Time. It notes this will be aligned with the official European Space Agency epoch when decided. That date is a time-stamp reference for the data grid.

When LISA observes, it may see black holes merging, white dwarf binaries, or a stochastic gravitational-wave hum. The document's point is that such analyses will depend not only on hardware and theory but also on the unglamorous work of agreeing what a minus sign means, which frame an angle is measured in, and which Fourier convention is in force. A shared language comes first. This document is part of building it.