WorksPhysics and Astronomy
The new generation lunar gravitational wave detectors: sky map resolution and joint analysis
Abstract
Lunar-based gravitational-wave interferometry is a fascinating endeavor, and was proposed as a promising approach to bridge the observational gap between space-borne and ground-based detectors. In this work, we adopt the Fisher-matrix method to examine the angular-resolution performance of the newly proposed Crater Interferometry Gravitational-wave Observatory (CIGO) on the lunar crater rim near the north pole, together with TianQin and LISA, for monochromatic sources in the 0.1-10 Hz band. We find that above 0.1 Hz, CIGO achieves better localization accuracy than the other two space-based missions and dominates the combined detector network's performance, provided that lunar noise mitigation is achieved in the 0.1-2.87 Hz frequency range. We further explore an upgraded Tetrahedron configuration, TCIGO, with a fourth station at the bottom of a crater, which forms a regular tetrahedral constellation on the lunar surface. The result shows that TCIGO yields a five-fold improvement in angular-resolution capability over CIGO and gets better sky coverage across the target frequency band.
Introduction
Gravitational-wave (GW) detection has become a crucial method of probing the nature of compact astrophysical objects under extreme conditions. Gravitational waves cover a broad frequency range from – Hz [1, 2, 3]. Existing GW detection missions are designed to probe different segments of this spectrum. Ground-based detectors, such as LIGO and Virgo [4, 5, 6, 7, 8, 9, 10, 11], are sensitive to frequencies in the range of 3 – Hz, while space-based interferometers like LISA [12, 13], TianQin [14, 15] and Taiji [16, 17] target the lower-frequency band of – Hz. However, the intermediate deci-hertz range (0.1–10 Hz) lies outside the optimal sensitivity bands of both ground- and space-based detectors.
The deci-hertz window (0.1–10 Hz) is of particular astrophysical interest, as it encompasses numerous rich phenomena [18, 19], such as the mergers of intermediate-mass black holes, compact binary white dwarf systems, and core-collapse supernovae [20, 21, 22]. Moreover, it provides a unique opportunity to test General Relativity [23, 24] and extensions of the Standard Model [25, 26, 27, 28]. Thus, constructing detectors optimized for this frequency range is essential for probing these phenomena.
Several proposals have been put forward for space-based deci-hertz GW observatories to bridge this gap, including the geocentric mission like SAGE [20] and heliocentric concepts such as DECIGO [29, 30], TVLBAI [31] and ALIA [32], and for earth-based detectors including TOBA [33], AION [34] and ZAIGA [35]. Recently, lunar-based GW detection has attracted increasing attention, aiming to fill this observational blind spot [29, 31, 36, 37, 38, 39, 40, 41]. The lunar GW observatory concept envisions the construction of an interferometric detector on the Moon. GLOC (Gravitational-Wave Lunar Observatory for Cosmology) [42] and LILA (Laser interferometer lunar antenna) [43] missions were proposed, aiming to build a 40-km equilateral triangle laser interferometer on the lunar surface as part of NASA’s Artemis program. LGWA (lunar gravitational-wave antenna) [40, 44, 45] deploys an array of high-end seismometers on the Moon to monitor its normal modes in the 1 mHz–1 Hz band excited by GWs. Additionally, Italian researchers proposed the LSGA (Lunar Seismic and Gravitational Antenna) [46] project to detect GW-induced lunar seismic strains.
There are a few advantages to constructing GW detectors on the Moon. Compared with earth-based detectors, firstly, the Moon naturally provides an ultra-high-vacuum environment, with surface pressures as low as Pa at sunrise, comparable to the LIGO vacuum system, significantly reducing construction and maintenance costs [47]. Secondly, at low frequencies (0.1–0.4 Hz), lunar seismic noise is three to four orders of magnitude lower than that on Earth [42, 48]. Thirdly, the absence of atmospheric, human, and tectonic disturbances allows for stable, long-term operation with a high duty cycle [43, 49, 50]. Consequently, gravity-gradient noise is also strongly suppressed, thereby minimizing the need for costly mitigation strategies [31, 37, 38, 39, 40, 41]. Compared with space-based detectors, a lunar-based GW observatory, stationed on the surface, can be serviced directly by astronauts, resulting in far lower maintenance costs than those of deep space missions [51]. Secondly, for low-frequency signals around 0.1 Hz, the data transmission rate required for a lunar observatory is only about one order of magnitude higher than that of space-based missions, making data transfer considerably more manageable [29, 36, 51, 52, 53]. Together, these advantages make the Moon an auspicious site for next-generation gravitational-wave astronomy, offering a stable, low-noise, and serviceable platform that bridges the spectrum gap between ground and space-based detectors.
Several leading research institutions in China, including the Beijing Institute of Technology and the Chinese Academy of Sciences, have jointly proposed a lunar-based gravitational-wave (GW) detection concept known as the Crater Interferometry Gravitational-wave Observatory (CIGO) [54, 55]. CIGO is conceived as a third-generation lunar interferometric detector deployed on the lunar surface, aiming at a target strain sensitivity of , approximately one order of magnitude beyond current second-generation ground-based detectors (). The instrument is designed to operate in the 0.1–10 Hz frequency band, bridging the observational gap between space- and earth-based missions. Unlike space-based missions such as LISA and TianQin, CIGO is not a free-flying constellation. Instead, it consists of three end stations rigidly anchored to the lunar surface, arranged in an equilateral triangular configuration near the lunar north pole (see Figure 8). The nominal arm length adopted in this work is 100 km. The constellation rotates with the Moon. The detector concept assumes a standard equal-arm Michelson interferometric configuration with stabilized laser links established between surface stations. Owing to its rigid geometry and equal-arm design, time-delay interferometry (TDI), which is essential for space-based unequal-arm interferometers, is not required. The interferometric response is therefore described by the standard finite-arm laser interferometer transfer function derived from the light round-trip time delay in the presence of gravitational waves, which is directly applicable to the lunar-based configuration considered here.
Figure 8. Schematic diagram of CIGO’s orbit and coordinate system. The detector is located at the Moon’s north pole. The Moon orbits the Earth with inclination relative to the ecliptic plane, while the Earth–Moon system revolves around the Sun. The inset shows the conceptual design of CIGO: three end stations (SC1–SC3) are placed on the rim of a lunar crater, forming an approximately equilateral triangular configuration with arm lengths of about 100 km. Each station hosts a test mass and three laser links forming the interferometric baselines.
In this work, we focus on the monochromatic gravitational waves, which can be emitted in the early stage of compact binary systems, including intermediate-mass black holes and neutron stars [20, 21, 22, 40], the first-order phase transitions from the early universe [56] and the subsequent cosmic string [57]. As with space-based detectors, we adopt an equilateral triangular configuration (see Figure 8). Using the Fisher Information Matrix (FIM) method, we analyze 3600 potential GW sources distributed across the celestial sphere to evaluate the localization accuracy of the detector. We further generate sky maps of angular resolution for CIGO, LISA, and TianQin at three representative frequencies: 0.1, 1, and 10 Hz. By comparing these results, we investigate the performance of the detectors across the 0.1–10 Hz Hz band. Since the operation of Taiji in space is similar to LISA’s, we will not discuss it separately here. Besides, we compute joint angular-resolution maps for the CIGO, LISA, and TianQin network configurations at the same frequencies, and assess the enhanced localization capability using a lunar–space joint analysis. Additionally, we upgrade the CIGO system by adding a detector at the base of the crater, forming a tetrahedral configuration with the origin plane. We then calculate and compare its angular resolution over the sky region. These results provide valuable insight into the potential of lunar-based interferometers to bridge the existing detection gap between space-based and ground-based GW observatories.
The paper is organized as follows. In the results section, we present the sky-localization performance of CIGO operating both independently and in a network with LISA and TianQin, and estimate the impact of lunar-based noise on CIGO’s localization accuracy. Besides, we investigate the regular tetrahedral configuration of CIGO and compare its angular resolution with that of the original planar configuration to assess potential improvements. We conclude all the results and provide an outlook in the discussion section. In the methods section, we outline the methodology and settings adopted in this work, including the Fisher information matrix formalism for monochromatic gravitational waves and the orbital equation of CIGO detectors. Some detailed derivations and parameter settings are presented in the supplementary information.
Results
Source Localization
The Crater Interferometry Gravitational-wave Observatory (CIGO) is a proposed equilateral-triangle laser interferometer designed to probe mid-frequency gravitational waves in the deci-hertz to 10-hertz range. Each arm has a nominal length of approximately km, aiming for a target strain sensitivity of .
We evaluate the angular-localization accuracy of CIGO and compare it with two representative space-based detectors to assess its performance. LISA and TianQin, which adopt similar triangular configurations. LISA consists of three spacecraft forming a triangle with km arm lengths, trailing Earth by about in the ecliptic plane with an orbital plane inclined at . While TianQin has arm lengths of km and its normal vector of the detector plane points to RX J0806.3+1527 with the latitude and the longitude in heliocentric coordinates. Both detectors operate in an Earth-centered orbit.
The distinct orbital dynamics and orientations of these missions introduce different levels of amplitude and Doppler modulation in their detector responses [58, 59], which, in turn, strongly influence their sky-localization capabilities. By comparing the sky maps of angular resolution across these three detectors in the mid-frequency band, we can clearly identify the optimal operating frequency range and relative advantages of CIGO.
In this section, we adopt the FIM method described above, simulate 3600 gravitational-wave sources uniformly distributed across the sky, and compute angular-resolution sky maps for CIGO, LISA, and TianQin. We set the mission time to one year and the observation coverage to in latitude and in longitude. The gravitational-wave signals are modeled as monochromatic waves, characterized by a simple waveform and a minimal set of parameters. To ensure the detectability of all signal sources, we set the SNR threshold for all sources to 7.
The results of angular resolutions at the frequencies Hz, 1 Hz and 10 Hz for the LISA, Tianqin and CIGO are shown in Figure 1 and Figure 2. We show the sky map of the angular resolution for LISA, Tianqin and CIGO in Figure 1, and the cumulative histograms of the angular resolution in Figure 2. The mean values of angular resolutions are summarized in Table I.

Figure 1. The sky map of angular resolutions of sources from different directions in the unit of steradian (1 steradian is 3000 square degrees) for LISA, TianQin, CIGO, and their network. The horizontal axis represents the longitude and the vertical axis represents the latitude . The frequencies of monochromatic sources are 0.1 Hz, 1 Hz, and 10 Hz.

Figure 2. Cumulative histograms of sky localization estimations for LISA (red), TianQin (blue), CIGO (yellow), and their joint network (black). From left to right and top to bottom the panels correspond to representative frequencies Hz, Hz, Hz, and Hz. Smaller values of indicate better angular resolution.
| (Hz) | LISA | TianQin | CIGO | Network |
| Mean value of angular resolution | ||||
| 0.01 | ||||
| 0.1 | ||||
| 1 | ||||
| 10 |
Table I. Angular resolution (in steradians) comparison for LISA, TianQin, CIGO, and their combined network at different frequencies
Figure 1 presents the sky maps of angular resolution for 3600 monochromatic gravitational-wave sources uniformly distributed over the celestial sphere, computed for LISA, Tian-Qin, CIGO, and their network at frequencies of 0.1 Hz, 1 Hz, and 10 Hz. The horizontal and vertical axes correspond to the source longitude (−180° to 180°) and latitude (−90° to 90°), respectively. The color scale represents the localization uncertainty in steradians (deg), where darker regions indicate better (smaller) angular resolution and lighter areas correspond to poorer performance. From top to bottom, the rows correspond to increasing frequencies (0.1 Hz, 1 Hz, and 10 Hz), while each column corresponds to a specific detector: LISA (left), TianQin (middle), and CIGO (right).
For LISA, we find that at all frequencies, the localization accuracy for sources near the ecliptic plane is typically two orders of magnitude worse than that for sources near the poles. In contrast, TianQin shows degraded angular resolution for sources located along its detector plane (around and ) and near the heliocentric equatorial plane, where is roughly two orders of magnitude larger than in other sky regions. Similarly, for CIGO, the localization accuracy deteriorates by about one order of magnitude for sources located near the longitude and , as well as near the lunar equatorial plane.
These patterns arise from the differences in the orbital geometry and orientation of the detector planes. TianQin’s detector usually continuously points toward its reference source RX J0806.3+1527, located at latitude and longitude . CIGO, in contrast, is fixed on the lunar surface near the north pole, where the lunar spin axis is nearly parallel to the normal of the ecliptic plane, and the inclination between the lunar equator and the ecliptic (the so-called white-plane inclination) is about . Consequently, CIGO’s detector plane faces approximately toward . Along the detector-plane normal vector, the signal-to-noise ratio is significantly reduced, leading to reduced localization accuracy (larger ) and weaker modulation [60]. LISA, however, lacks a fixed reference source; its detector plane rotates around the Sun with an inclination of , inducing strong amplitude modulations that significantly enhance sky-localization performance across most directions. As a result, LISA exhibits a more uniform angular resolution over the sky map. At the same time, TianQin and CIGO show slightly less sky coverage, characterized by high precision over large regions but weaker performance in specific, limited directions.
One point that needs emphasis is the impact of uncertainties in lunar orbital and rotational modeling. Actually, in contrast
to heliocentric free-flying constellations, CIGO is rigidly anchored on the lunar surface. Therefore, the detector geometry is fixed in the lunar body frame, and the only time-dependent modulation relevant to sky localization arises from the rotation and precession of the Moon, which will induce the deflection of the detector plane. Based on the residual levels reported in the JPL ephemerides [61], we adopt a very conservative uncertainty in the lunar libration angles, mas, such an perturbation would induce a shift of angular resolution , and the resulting fractional change is
This level of shift is entirely negligible compared with the intrinsic Fisher-matrix uncertainties and instrumental noise contributions.
Importantly, small perturbations in the lunar orientation do not change the overall scale of the angular resolution. Instead, they induce only a tiny redistribution of the sky-location error pattern. In other words, the global angular resolution capability of the detector remains unaffected, while the directional dependence of the error ellipse may experience an imperceptible shift at the level of .
Figure 2 illustrates the cumulative distribution functions (CDFs) of the angular resolution for the same set of sources across different frequencies, with curves for LISA (red), TianQin (blue), CIGO (yellow), and their combined network (black). The -axis is logarithmic in (deg), and the -axis shows the cumulative probability, i.e., the fraction of sources localized below a given uncertainty. The panels are arranged by frequency: left for 0.1 Hz, middle for 1 Hz, and right for 10 Hz.
At Hz, both CIGO and TianQin exhibit localization accuracies of deg, whereas LISA’s resolution is reduced by roughly one order of magnitude. As the frequency increases to Hz, CIGO’s localization performance surpasses that of the other two detectors by more than two orders of magnitude. However, in our target band (– Hz), the observing frequency satisfies , the angular resolution of a single detector improves with increasing gravitational-wave frequency. In contrast, for LISA ( Hz) and TianQin ( Hz), the target band lies well above their frequencies. Consequently, the transfer function’s frequency dependence deteriorates markedly, causing the responses of LISA and TianQin to saturate.
For the network configuration combining CIGO, TianQin, and LISA, the joint observations provide a clear improvement at low frequencies (– Hz) when compared to any single detector. In this regime, the decreasing regions of sky map are well-complementary, and the cumulative distributions of shift leftward relative to the individual-detector curves, reflecting the benefit of the geometric complementarity among the three non-coplanar constellations.
As the frequency increases into the – Hz range, however,
the network’s performance becomes overwhelmingly dominated by CIGO. As indicated by the black CDF curve in Figure 2, the network curve coincides with that of CIGO at these frequencies, showing that CIGO’s response effectively sets the localization accuracy of the joint observation. In this high-frequency regime, the contributions from LISA and TianQin become negligible, and the network inherits CIGO’s superior angular-resolution capability.
For Michelson interferometers operating in space, such as TianQin and LISA, the analytical form of the single-link noise power spectral density is written as [62, 63, 64]
where denotes the interferometer arm length, and are displacement noise and acceleration noise, respectively. Since there is currently no formula for estimating quantum noise for a CIGO-like detector, which is a Michelson-laser interferometer operating on a planet, in this work, we first naively assume that the noise level of CIGO can be achieved at least at the noise level in TianQin [60] with the same frequency dependence.
One notable feature is that the CIGO detector is designed as a Michelson-laser interferometer deployed on the lunar surface. If we examine the detector’s noise budget more closely, CIGO’s performance can vary significantly with noise level. Notice that the formalism of Michelson-laser power spectral density is largely implementation-agnostic, so it’s also suitable for CIGO [65, 66]. Here, we estimate the noise based on the ground-based program. Given that CIGO is deployed on the moon, it will be affected by the lunar environment, to assess for these effects, we employ the pygwinc package [67]and assume that CIGO adopts detector technologies comparable to those of the Cosmic Explorer Stage 1 (CE1) [68], including the test masses, thermal shielding, and vibration isolation systems,etc. Starting from the CE1 design noise budget, we modify the environmental parameters correspond to lunar conditions and obtain the corresponding lunar environmental strain noise. Following Ref [65, 66], we adopt the relation below to connect the noise power spectral density and the strain :
Substituting into equation (17), we obtain the overall strain-noise budget, shown in Figure 3. For instructions on setting and modifying the relevant parameters, please refer to Supplementary Information B.

Figure 3. The lunar environmental noise level estimation of CIGO. The strain noise amplitude spectral density (in ) is shown as a function of frequency for the dominant lunar noise sources: seismic (brown), Newtonian (green), suspension thermal (blue), coating thermal (red), substrate thermal (orange), and the space-borne noise level (yellow horizontal line). Total noise (black) is the square root of the sum of the squares of all contributions.
As shown in Figure 3, within the target frequency band (0.1–10Hz) of CIGO, if we assume that the overall sensitivity level of CIGO is comparable to that of CE1, then even though the lunar seismic environment is quieter than that on Earth, the total noise curve is still significantly elevated in the frequency range below 2.87 Hz due to seismic noise contributions. Applying power-law fitting and smoothing procedure [42, 69, 70], we obtain an analytical expression for the noise strain amplitude for CIGO below:
By replacing the in Eq. (15) with Eq. (21), we recompute the angular resolution of CIGO, and the cumulative distribution histograms of the angular resolution are shown in Figure 4.

Figure 4. Cumulative histograms of sky localization estimations considering the lunar noise. From left to right and top to bottom the panels correspond to representative frequencies Hz, Hz, Hz, and Hz. Smaller values of indicate better angular resolution.
Fig. 4 shows the cumulative histograms of the angular resolution at 0.1 and 1Hz after accounting for lunar environmental noise. It’s seen that within the seismic-dominated frequency range of 0.1–2.87 Hz, the angular resolution of CIGO is substantially reduced. This indicates that, for a lunar-based implementation of CIGO, more advanced seismic isolation techniques and improved suspension thermal-noise suppression will be required to reduce instrumental noise further and achieve the localization performance demonstrated in Figure 2.
Actually, such low-frequency seismic noise is a critical technical challenge for any lunar-based interferometer. However, several mitigation strategies may be considered. For example, it has been reported in the literature that by using the Suspension Platform Interferometer (SPI) technology, the length fluctuations of the detector in a suspended optical resonator can be reduced by three orders of magnitude [71, 72]. Additionally, installing environmental sensors, such as seismometers and microphones, can reduce gravitational noise [73]. Besides, our estimation of the noise spectrum is intentionally conservative. Compared to other posted proposals, such as GLOC [42], our total noise level is more than three orders of magnitude higher in the relevant frequency band. This difference likely arises because GLOC adopts the default pygwinc quantum noise to estimate the shot noise induced by photon fluctuations in the laser beam, whereas we implement a space-based formalism. The precise noise curve will ultimately depend on the detailed detector configuration of CIGO and can be adjusted accordingly once the instrumental design is specified.
It should be clarified that the above noise modeling is carried out under the assumptions of stationary Gaussian noise and thermal equilibrium. All modeling is conducted in the frequency domain and does not consider time-evolution characteristics.
Tetrahedron Form of CIGO
In this section, we consider an upgraded configuration of the CIGO detector, referred to as the Tetrahedron Crater Interferometry Gravitational-wave Observatory (TCIGO). As illustrated in Figure 5, we augment the original lunar-polar CIGO by adding an interferometric station at the bottom of the crater. The four stations form a regular tetrahedral geometry. Each triangular face functions effectively as two independent laser interferometers, introducing three additional baselines with distinct orientations. Consequently, a single observation can be treated as an eight-interferometer network. This configuration is proposed to improve sky-localization performance for lunar-based detectors.

Figure 5. Conceptual design of TCIGO. Three spacecraft (SC1–SC3) are situated on the rim of a lunar crater, with a fourth (SC4) located at its bottom, forming a regular tetrahedral configuration. Each end station, separated by 100 km, houses a test mass and three laser systems. Image of the lunar surface is adapted from Ref. [42].
We label the four interferometric stations as SC1–SC4. Following the orbital prescription of CIGO and adopting the settings of [74, 75], the position functions of the -th detector are identical to Eqs. (10)–(13), the angle represents the angular displacement of the -th detector in the detector-plane / lunar-polar coordinate system, with , characterizing the relative phase between SC1 and SC4. All other parameters follow the ”Source Localization” section. is the distance from the center of the tetrahedron to each detector. For SC1–SC3, we use km, while for SC4, we take km, such that the four stations form a regular tetrahedron.
However, it is important to note that such a regular tetrahedral configuration is more of an idealized discussion. The purpose of introducing it is to study how adding the fourth station can improve the angular resolution and sky coverage of CIGO. Actually, such deep craters on the moon’s surface are virtually non-existent. In practice, the non-equal-arm configuration may be preferable to the regular tetrahedron.
We refer to the triangular detector formed by SC1–SC2–SC4 as and that formed by SC1–SC3–SC4 as . The original CIGO corresponds to the triangle SC1–SC2–SC3. Figure 6 presents the sky maps of angular-resolution performance for 3600 monochromatic gravitational-wave sources uniformly distributed over the celestial sphere for S1 (left), CIGO (middle), and TCIGO (right), evaluated at frequencies of Hz, Hz, and Hz. Figure 7 shows the cumulative distribution functions (CDFs) of the angular resolution for the same set of sources across all frequencies, with curves for LISA (red), S1 (blue), S2 (pink; nearly overlapping with S1), CIGO (yellow), and TCIGO (purple).

Figure 6. The sky map of angular resolutions of sources from different directions in the unit of steradian (1 steradian is 3000 square degrees) for S1, CIGO, and TCIGO. The horizontal axis represents the longitude and the vertical axis represents the latitude . The left panel shows S1, the middle panel shows CIGO, and the right panel shows TCIGO. From top to bottom, the frequencies of monochromatic sources are 0.1 Hz, 1 Hz, and 10 Hz.

Figure 7. Cumulative histograms of sky localization estimations for LISA (red), TianQin (blue), CIGO (yellow) and TCIGO(purple) . From left to right and top to bottom the panels correspond to representative frequencies Hz, Hz, Hz, and Hz. Smaller values of indicate better angular resolution.
From Figure 6, we observe that for a single detector formed with SC4 (e.g., S1), the sky-localization distribution and precision are comparable to those of CIGO. However, its high-precision region is slightly smaller. However, when all four faces are combined to form the full TCIGO array, the reduction in angular resolution caused by the orientation of the CIGO (latitude ) was compensated. As seen in Figure 7, at Hz, the cumulative localization performance of TCIGO is already comparable to the CIGO+space-based network configuration shown in Figure 2. For Hz, TCIGO surpasses not only CIGO but also all existing or planned space-based interferometers. Overall, within our target frequency band, TCIGO improves the angular resolution by a factor of 5 compared to CIGO.
Discussion
We investigated the angular resolution of LISA, TianQin, and CIGO for monochromatic gravitational waves in the frequency band 0.1–10 Hz. The results show that CIGO exhibits excellent source-localization capability. For Hz, CIGO’s angular resolution surpasses that of TianQin and LISA by several orders of magnitude, and the joint localization performance of the three-detector network is dominated by CIGO. Similar to TianQin, the normal vector to the CIGO detector plane is aligned with the lunar north pole and thus points toward the exact ecliptic longitude on the sky; in this direction, the signal-to-noise ratio is significantly reduced, resulting in reduced localization accuracy (larger ) and weak amplitude/phase modulation. Consequently, in the sky maps, the region around ecliptic longitude exhibits noticeably lower angular resolution than other regions. In contrast, LISA benefits from continuously rotating detector-plane orientation along its orbit, which provides amplitude modulation and yields more uniform sky coverage. In addition, we investigate the impact of the detector-noise formalism on angular-resolution modeling and show that CIGO would achieve better performance if low-frequency lunar seismic noise is further suppressed.
To eliminate the angular-resolution blind spot caused by CIGO’s fixed orientation, we further placed an additional detector at the bottom of the same lunar crater. Together with the original three CIGO detectors, this forms a regular tetrahedral configuration. The tetrahedral geometry introduces three additional oriented planes, yielding six new independent interferometric channels. The results show that, across the observed frequency band, the angular resolution of each new detector alone is slightly lower than that of the original CIGO triad. However, when combined with the original detectors, the overall localization accuracy improves by a factor of approximately 5. Moreover, because the detector orientations now provide complete sky coverage, the sky coverage blind spots previously present in CIGO are effectively eliminated.
The lunar-based CIGO, particularly in its tetrahedral configuration, delivers outstanding localization performance for mid-frequency gravitational-wave sources in the 0.1–10 Hz band. It thereby serves as a vital bridge between space-based detectors (LISA and TianQin), which dominate the low-frequency regime (– Hz), and ground-based observatories (LIGO, Virgo, KAGRA, Einstein Telescope, Cosmic Explorer), which cover the high-frequency band (– Hz). This unique setup makes CIGO an indispensable component of gravitational-wave astronomy in the coming decades
Methods
General GW Signal and Response in GR
In general relativity, a gravitational wave signal is described by the strain tensor
where is the GW strain amplitude in the polarization mode , and is the corresponding polarization tensor. At the quadrupole leading order, the two polarizations of a monochromatic gravitational wave with frequency take the form:
Here denotes the amplitude, determined by the binary masses , the distance between them and the luminosity distance between the GW source and the observer, represents the unit vector for the binary’s orbital angular momentum, and is the initial phase.
Then the output of the detector is
where denotes the plus and cross polarizations, is the instrumental noise, is the doppler phase modulation factor. The response function of detector for the polarization mode can be written as
where is the detector tensor for detector , and is the GW polarization tensor. The detector tensor represents the instrument’s response to signals from specific directions. For a monochromatic gravitational wave with frequency propagating along , in an equal-arm Michelson laser interferometer with single round-trip light travel, is given by
respectively. The transfer function is given by [58, 59, 60, 69, 76]
where and denote the unit vectors along the two arms of the detector , and is the incident direction vector of the gravitational wave. is the transfer frequency, where is the arm length and is the speed of light. For CIGO, which primarily targets the 0.1–10 Hz band, the GW’s wavelength is much larger than the detector arm length (100 km). This implies that the observing frequency , so and the transfer function approaches unity: . Consequently, in low frequency limit, the detector tensor reduces to its long-wavelength, frequency-independent form:
The modulation factor on the detector induced by the Earth-Sun and Moon-Earth motions , can be expressed as follows [60]
where is the GW frequency, is the orbit radius of the detector center in the heliocentric-ecliptic coordinate system, is the spherical coordinates of the gravitational-wave source on the celestial sphere, is the ecliptic longitude of the detector at . The detailed derivation of this formula is presented in Supplementary Information C. For CIGO, we can calculate using the equivalent radius by direct geometric projection:
where and are the radius of Earth and Moon orbit. Due to the long duration and large scale of the CIGO mission, we often set to 1 AU and the observation time to 1 year. is the orbital inclination of the lunar orbit plane in Figure 8. Given the significant gap in their orders of magnitude, (0.00257 AU) can be neglected.
Lunar detector orbit and detector tensor
To calculate the detector tensor, it is necessary to determine the orbital equations of each station in CIGO. We adopt a heliocentric-ecliptic coordinate system (Figure 8), with the x axis pointing toward the vernal equinox, the y axis completing the right-handed triad, and the z axis normal to the ecliptic plane. The n-th detector position can be decomposed as
where represents the position vector of the Earth in the heliocentric frame, represents the position vector of the Moon relative to Earth, and represents the position vector of the detector relative to the Moon, where . is the high order perturbation to the corrections from the orbits of the moon, the Earth, the Sun, other planets (mostly Jupiter), etc [76].
The position vector of the -th detector is given by
where denotes the higher-order infinitesimal to the eccentricity. Here, is the radius of the moon; and are the semi-major axis and eccentricity of the Moon’s orbit around the Earth; and are the semi-major axis and eccentricity of the Earth’s orbit around the Sun, respectively. The angle is the angular displacement of the -th detector in the detector-plane–lunar polar coordinate system after time , where . The term represents the mean lunar longitude of the Moon in the geocentric–lunar coordinate system, where is the modulation frequency due to the Moon’s orbital motion and is the mean lunar longitude at . Similarly, denotes the mean ecliptic longitude of the Earth in the heliocentric–ecliptic coordinate system, where is the orbital frequency of the Earth and is the mean ecliptic longitude measured from the vernal equinox at . , where is the distance from the center of the detector to each detector. At the same time, is the angle between the lunar orbit plane and the ecliptic. The detailed derivation is provided in Supplementary Information A.
Fisher Information Matrix
The monochromatic gravitational waves considered in this work can be probed by interferometers such as CIGO, as well as by other space-based detectors. We adopt seven parameters in our Fisher information matrix (FIM) analysis:
which are source direction ; the direction of the binary’s orbital angular momentum ; The amplitude , the initial phase , and frequency .
Then the FIM takes the form as [58, 59, 60, 69, 76, 77]:
in which denotes the -th parameter and .
Accordingly, the corresponding covariance matrix is given by
The is the single-link noise power spectral density, which we have discussed in ”Result” section. The angular uncertainty in sky localization is quantified as
with and denoting the source sky coordinates. We now have the basic setup for the sky-map resolution, and we have present numerical results in ”Result” section.
References
- [1]Yungui Gong, Jun Luo, and Bin Wang. Concepts and status of chinese space gravitational wave detection projects. Nature Astronomy, 5(9):881–889, 2021.DOI
- [2]Bailes, M., Berger, B. K., Brady, P. R., Branchesi, M., Danzmann, K., Evans, M., Holley-Bockelmann, K., Iyer, B., Kajita, T., Katsanevas, S., & al., . et . (2021). Gravitational-wave physics and astronomy in the 2020s and 2030s. Nature Reviews Physics, 3(5), 344–366.
- [3]Rong-Gen Cai, Zong-Kuan Guo, Bin Hu, Chang Liu, Youjun Lu, Wei-Tou Ni, Wen-Hong Ruan, Naoki Seto, Gang Wang, and Yue-Liang Wu. On networks of space-based gravitational-wave detectors. Fundamental Research, 4(5):1072–1085, 2024.DOI
- [4]Abbott, B. P., Abbott, R., Abbott, T. D., Abernathy, M. R., Acernese, F., Ackley, K., Adams, C., Adams, T., Addesso, P., Adhikari, R. X., & al., . et . (2016). Gw151226: observation of gravitational waves from a 22-solar-mass binary black hole coalescence. Physical Review Letters, 116(24), 241103.
- [5]Abbott, B. P., Abbott, R., Abbott, T. D., Acernese, F., Ackley, K., Adams, C., Adams, T., Addesso, P., Adhikari, R. X., Adya, V. B., & al., . et . (2017). Gw170104: observation of a 50-solar-mass binary black hole coalescence at redshift 0.2. Physical Review Letters, 118(22), 221101.
- [6]Abbott, B. P., Abbott, R., Abbott, T., Acernese, F., Ackley, K., Adams, C., Adams, T., Addesso, P., Adhikari, R. X., Adya, V. B., & al., . et . (2018). Gw170814: a three-detector observation of gravitational waves from a binary black hole coalescence. Physical Review Letters, 119(14), 141101.
- [7]Abbott, B. P., Abbott, R., Abbott, T., Acernese, F., Ackley, K., Adams, C., Adams, T., Addesso, P., Adhikari, R. X., Adya, V. B., & al., . et . (2018). Gw170817: Measurements of neutron star radii and equation of state. Physical Review Letters, 121(16), 161101.
- [8]Abbott, B. P., Abbott, R., Abbott, T. D., Acernese, F., Ackley, K., Adams, C., Adams, T., Addesso, P., Adhikari, R. X., Adya, V. B., & al., . et . (2017). Gw170817: observation of gravitational waves from a binary neutron star inspiral. Physical Review Letters, 119(16), 161101.
- [9]Abbott, B. P., Abbott, R., Abbott, T., Abraham, S., Acernese, F., Ackley, K., Adams, C., Adhikari, R., Adya, V., Affeldt, C., & al., . et . (2019). Gwtc-1: a gravitational-wave transient catalog of compact binary mergers observed by ligo and virgo during the first and second observing runs. Physical Review X, 9(3), 031040.
- [10]Benjamin P Abbott, R Abbott, R Adhikari, P Ajith, Bruce Allen, G Allen, RS Amin, SB Anderson, WG Anderson, MA Arain, et al. Ligo: the laser interferometer gravitational-wave observatory. Reports on Progress in physics, 72(7):076901, 2009.DOI
- [11]Abbott, B. P., Abbott, R., Abbott, T. D., Abernathy, M. R., Acernese, F., Ackley, K., Adams, C., Adams, T., Addesso, P., Adhikari, R. X., & al., . et . (2016). Observation of gravitational waves from a binary black hole merger. Physical Review Letters, 116(6), 061102.
- [12]Amaro-Seoane, P., Audley, H., Babak, S., Baker, J., Barausse, E., Bender, P., Binetruy, P., Born, M., Bortoluzzi, D., & al., . et . (2017). Laser interferometer space antenna. In arXiv preprint arXiv:1702.00786. 1702.00786
- [13]Agency, E. S. (2025). Summary of the laser interferometer space antenna (lisa) mission. In ESA Science and Technology Summary. <a href="https://doi.org/https://sci.esa.int/web/lisa/-/31380-summary" rel="noopener noreferrer" target="_blank">https://doi.org/https://sci.esa.int/web/lisa/-/31380-summary</a>
- [14]Luo, J., Chen, L.-S., Duan, H.-Z., Gong, Y.-G., Hu, S., Ji, J., Liu, Q., Mei, W., Milyukov, Y., Sazhin, M., & al., . et . (2016). Tianqin: a space-borne gravitational wave detector. Classical and Quantum Gravity, 33(3), 035010.
- [15]Luo, J., Bai, S., Bai, Y., Cai, L., Dang, H., Dong, Q., Duan, H.-Z., Du, Y., Fan, L., Fu, X., & al., . et . (2025). Progress of the tianqin project. Classical and Quantum Gravity, 42(17), 173001.
- [16]Wen-Rui Hu and Yue-Liang Wu. The taiji program in space for gravitational wave physics and the nature of gravity, 2017.DOI
- [17]Ziren Luo, Yan Wang, Yueliang Wu, Wenrui Hu, and Gang Jin. The taiji program: A concise overview. Progress of Theoretical and Experimental Physics, 2021(5):05A108, 2021.DOI
- [18]Kazuaki Kuroda, Wei-Tou Ni, and Wei-Ping Pan. Gravitational waves: Classification, methods of detection, sensitivities and sources. International Journal of Modern Physics D, 24(14):1530031, 2015.DOI
- [19]Dongfeng Gao, Wei-Tou Ni, Jin Wang, Ming-Sheng Zhan, and Lin Zhou. Mid-frequency gravitational waves (0.1-10 hz): Sources and detection methods summary of the parallel session gw2 of mg16 meeting. arXiv preprint arXiv:2111.09715, 2021.arxiv.org/abs/2111.09715
- [20]Manoel Felipe Sousa, Tabata Aira Ferreira, and Massimo Tinto. Parameter estimation precision with geocentric gravitational wave interferometers: Monochromatic signals. Universe, 11(4):122, 2025.DOI
- [21]Tong Jiang, Yungui Gong, and Xuchen Lu. Sky localization of space-based detectors with time-delay interferometry. Journal of Cosmology and Astroparticle Physics, 2024(10):067, 2024.DOI
- [22]Ivo R Seitenzahl, Matthias Herzog, Ashley J Ruiter, Kai Marquardt, Sebastian T Ohlmann, and Friedrich K Ropke. Neutrino and gravitational wave signal of a delayed-detonation model of type ia supernovae. Physical Review D, 92(12):124013, 2015.DOI
- [23]Arca Sedda, M., Berry, C. P., Jani, K., Amaro-Seoane, P., Auclair, P., Baird, J., Baker, T., Berti, E., Breivik, K., Burrows, A., & al., . et . (2020). The missing link in gravitational-wave astronomy: discoveries waiting in the decihertz range. Classical and Quantum Gravity, 37(21), 215011.
- [24]Karan Jani, Deirdre Shoemaker, and Curt Cutler. Detectability of intermediate-mass black holes in multiband gravitational wave astronomy. Nature Astronomy, 4(3):260–265, 2020.DOI
- [25]Yao, R.-M., Jin, X.-J., Yin, P.-F., & Huang, Q.-G. (2025). Axion-like dark matter search with space-based gravitational wave detectors. https://doi.org/arXiv:2504.10083
- [26]Heng-Tao Xu, Yue-Hui Yao, Yong Tang, and Yue-Liang Wu. Distinguishing monochromatic signals in lisa and taiji: Ultralight dark matter versus gravitational waves. Physical Review D, 112(9):095021, 2025.DOI
- [27]Ke Wang and Yin Zhong. Frequency modulation of gravitational waves by ultralight scalar dark matter. Physical Review D, 108(12):123531, 2023.arxiv.org/abs/2306.10732
- [28]Figlilia, M., Grippa, F., Lambiase, G., & Visinelli, L. (2026). Gravitational signatures of axion dark matter via parity-violating interactions. Physical Review D, 113(2), 023012.
- [29]Kawamura, S., Nakamura, T., Ando, M., Seto, N., Akutsu, T., Funaki, I., Ioka, K., Kanda, N., Kawano, I., Musha, M., & al., . et . (2019). Space gravitational-wave antennas decigo and b-decigo. International Journal of Modern Physics D, 28(12), 1845001.
- [30]Kawamura, S., Ando, M., Seto, N., Sato, S., Musha, M., Kawano, I., Yokoyama, J., Tanaka, T., Ioka, K., Akutsu, T., & al., . et . (2021). Current status of space gravitational wave antenna decigo and b-decigo. Progress of Theoretical and Experimental Physics, 2021(5), 05A105.
- [31]Abend, S., Allard, B., Alonso, I., Antoniadis, J., Araújo, H., Arduini, G., Ardoini, A. S., Arnold, T., Augst, N., Badurina, L., & al., . et . (2024). Terrestrial very-long-baseline atom interferometry: Workshop summary. AVS Quantum Science, 6(2).
- [32]Jeff Crowder and Neil J Cornish. Beyond lisa: Exploring future gravitational wave missions. Physical Review D—Particles, Fields, Gravitation, and Cosmology, 72(8):083005, 2005.DOI
- [33]Tomofumi Shimoda, Satoru Takano, Ching Pin Ooi, Naoki Aritomi, Yuta Michimura, Masaki Ando, and Ayaka Shoda. Torsionbar antenna: A ground-based mid-frequency and low-frequency gravitational wave detector. International Journal of Modern Physics D, 29(04):1940003, 2020.DOI
- [34]Badurina, L., Bentine, E., Blas, D., Bongs, K., Bortoletto, D., Bowcock, T., Bridges, K., Bowden, W., Buchmueller, O., Burrage, C., & al., . et . (2020). Aion: an atom interferometer observatory and network. Journal of Cosmology and Astroparticle Physics, 2020(05), 011.
- [35]Zhan, M.-S., Wang, J., Ni, W.-T., Gao, D.-F., Wang, G., He, L.-X., Li, R.-B., Zhou, L., Chen, X., Zhong, J.-Q., & al., . et . (2020). Zaiga: Zhaoshan long-baseline atom interferometer gravitation antenna. International Journal of Modern Physics D, 29, 1940005.
- [36]Baker, J., Barke, S. F., Bender, P. L., Berti, E., Caldwell, R., Conklin, J. W., Cornish, N., Ferrara, E. C., Holley-Bockelmann, K., Kamai, B., & al., . et . (2019). Space based gravitational wave astronomy beyond lisa. arXiv:1907.11305
- [37]Branchesi, M., Falanga, M., Harms, J., Jani, K., Katsanevas, S., Lognonne, P., Badaracco, F., Cacciapuoti, L., Cappellaro, E., Dell’Agnello, S., & al., . et . (2023). Lunar gravitational-wave detection. Space Science Reviews, 219(8), 67.
- [38]Norman Lafave and Thomas L Wilson. Lunar ligo: A new concept in gravitational wave astronomy. In Lunar and Planetary Inst., Twenty-Fourth Lunar and Planetary Science Conference. Part 2: GM, 1993.
- [39]Andrea Cozzumbo, Benedetta Mestichelli, Marco Mirabile, Lavinia Paiella, Jacopo Tissino, and Jan Harms. Opportunities and limits of lunar gravitational-wave detection. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 382(2271), 2024.DOI
- [40]Hanlin Song, Han Yan, Yacheng Kang, Xian Chen, Junjie Zhao, and Lijing Shao. Probing intermediate-mass black hole binaries with the lunar gravitational-wave antenna. npj Space Exploration, 1(1):2, 2025.DOI
- [41]Han Yan, Xian Chen, Jinhai Zhang, Fan Zhang, Mengyao Wang, and Lijing Shao. Toward a consistent calculation of the lunar response to gravitational waves. Physical Review D, 109(6):064092, 2024.DOI
- [42]Karan Jani and Abraham Loeb. Gravitational-wave lunar observatory for cosmology. Journal of Cosmology and Astroparticle Physics, 2021(06):044, 2021.DOI
- [43]Jani, K., Abernathy, M., Berti, E., Boschi, V., Chakrabarti, S., Cocoros, A., Conklin, J. W., Creighton, T., Dell’Agnello, S., Diels, J.-C., & al., . et . (2025). Laser interferometer lunar antenna (lila): Advancing the us priorities in gravitational-wave and lunar science. arXiv:2508.11631
- [44]Harms, J., Ambrosino, F., Angelini, L., Braito, V., Branchesi, M., Brozato, E., Cappellaro, E., Coccia, E., Coughlin, M., Della Ceca, R., & al., . et . (2021). Lunar gravitational-wave antenna. The Astrophysical Journal, 910(1), 2.
- [45]Ajith, P., Seoane, P. A., Sedda, M. A., Arcodia, R., Badaracco, F., Banerjee, B., Belgacem, E., Benetti, G., Benetti, S., Borick, A., & al., . et . (2025). The lunar gravitational-wave antenna: mission studies and science case. Journal of Cosmology and Astroparticle Physics, 2025(01), 108.
- [46]Katsanevas, S. (2020). Ideas for exploring the moon with a large european lander.
- [47]Francis S Johnson, James M Carroll, and Dallas E Evans. Vacuum measurements on the lunar surface. Journal of Vacuum Science and Technology, 9(1):450–456, 1972.DOI
- [48]Jan Harms. Seismic background limitation of lunar gravitational-wave detectors. Physical Review Letters, 129(7):071102, 2022.DOI
- [49]De Angelis, G., Badavi, F., Clem, J., Blattnig, S., Clowdsley, N., Ratliff, K., Rathi, J., & Wilson, J. (2007). Modeling of the lunar radiation environment. Nuclear Physics B-Proceedings Supplements, 166, 169–183.
- [50]M Horányi, JR Szalay, S Kempf, J Schmidt, E Grün, R Srama, and Z Sternovsky. A permanent, asymmetric dust cloud around the moon. Nature, 522(7556):324–326, 2015.DOI
- [51]Niu, J.-S., Wang, Y.-Y., Bian, X., Wei, L.-F., Hou, Q., & Zhang, X.-M. (2024). Lunar-based laser interferometer gravitational wave observatory. Progress in Astronomy, 42:203–223.
- [52]Eric Larose, Michel Campillo, A Khan, and Y Nakamura. Lunar subsurface investigated from correlation of seismic noise. Geophysical Research Letters, 32:L16201, 2005.DOI
- [53]Lognonné, P., Feuvre, M. L., Johnson, C. L., & Weber, R. C. (2009). Moon meteoritic seismic hum: steady state prediction. Journal of Geophysical Research: Planets, 114(E12).
- [54]University, L., & Physics, L. C. for T. (2024). Workshop on lunar-based gravitational wave detection (cigo). <a href="https://doi.org/https://lctp.lzu.edu.cn/content.jsp?wbtreid=1021&wbnewsid=2402" rel="noopener noreferrer" target="_blank">https://doi.org/https://lctp.lzu.edu.cn/content.jsp?wbtreid=1021&wbnewsid=2402</a>
- [55]Engineering, C. for O. S. and . (2024). Cigo project discussion meeting. <a href="https://doi.org/http://www.cose.edu.cn/info/1038/2160.htm" rel="noopener noreferrer" target="_blank">https://doi.org/http://www.cose.edu.cn/info/1038/2160.htm</a>
- [56]Djuna Lize Croon and David James Weir. Gravitational waves from phase transitions. Contemporary Physics, 65(2):75–96, 2024.DOI
- [57]Ruiyu Zhou and Lingong Bian. Gravitational waves from cosmic strings after a first-order phase transition. Chinese Physics C, 46(4):043104, 2022.DOI
- [58]Curt Cutler. Angular resolution of the lisa gravitational wave detector. Physical Review D, 57(12):7089, 1998.DOI
- [59]Arkadiusz B laut. Accuracy of estimation of parameters with lisa. Physical Review D—Particles, Fields, Gravitation, and Cosmology, 83(8):083006, 2011.DOI
- [60]Chao Zhang, Yungui Gong, Hang Liu, Bin Wang, and Chunyu Zhang. Sky localization of space-based gravitational wave detectors. Physical Review D, 103(10):103013, 2021.DOI
- [61]Park, R. S., Flanagan, W. M., Williams, J. G., & Boggs, D. H. (2021). The jpl planetary and lunar ephemerides de440 and de441. The Astronomical Journal, 161, 105.
- [62]Travis Robson, Neil J Cornish, and Chang Liu. The construction and use of lisa sensitivity curves. Classical and Quantum Gravity, 36(10):105011, 2019.DOI
- [63]Stanislav Babak, Martin Hewitson, and Antoine Petiteau. Lisa sensitivity and snr calculations. arXiv preprint arXiv:2108.01167, 2021.arXiv
- [64]Neil J Cornish. Detecting a stochastic gravitational wave background with the laser interferometer space antenna. Physical Review D, 65(2):022004, 2001.DOI
- [65]Bangalore Suryanarayana Sathyaprakash and Bernard F Schutz. Physics, astrophysics and cosmology with gravitational waves. Living reviews in relativity, 12(1):2, 2009.DOI
- [66]Christopher J Moore, Robert H Cole, and Christopher PL Berry. Gravitational-wave sensitivity curves. Classical and Quantum Gravity, 32(1):015014, 2014.DOI
- [67]Collaboration, L. S. (2020). pygwinc: Gravitational wave interferometer noise calculator.
- [68]Evan D Hall. Cosmic explorer: A next-generation ground-based gravitational-wave observatory. Galaxies, 10(4):90, 2022.DOI
- [69]Neil J Cornish and Shane L Larson. Space missions to detect the cosmic gravitational-wave background. Classical and Quantum Gravity, 18(17):3473, 2001.DOI
- [70]Peter L Bender and Dieter Hils. Confusion noise level due to galactic and extragalactic binaries. Classical and Quantum Gravity, 14(6):1439, 1997.DOI
- [71]Sina M Koehlenbeck, Conor M Mow-Lowry, Gerald Bergmann, Robin Kirchoff, Philip Koch, Gerrit Kühn, Johannes Lehmann, Patrick Oppermann, Janis Wöhler, and David S Wu. A study on motion reduction for suspended platforms used in gravitational wave detectors. Scientific Reports, 13(1):2388, 2023.DOI
- [72]Matichard, F., Lantz, B., Mittleman, R., Mason, K., Kissel, J., Abbott, B., Biscans, S., McIver, J., Abbott, R., Abbott, S., & al., . et . (2015). Seismic isolation of advanced ligo: Review of strategy, instrumentation and performance. Classical and Quantum Gravity, 32(18), 185003.
- [73]Jan Harms and Ho Jung Paik. Newtonian-noise cancellation in full-tensor gravitational-wave detectors. Physical Review D, 92(2):022001, 2015.DOI
- [74]Louis J Rubbo, Neil J Cornish, and Olivier Poujade. Forward modeling of space-borne gravitational wave detectors. Physical Review D, 69(8):082003, 2004.DOI
- [75]Hong-Bo Jin and Cong-Feng Qiao. Tetrahedron constellation of gravitational wave observatory. Science China Physics, Mechanics & Astronomy, 68(2):220414, 2025.DOI
- [76]Xin-Chun Hu, Xiao-Hong Li, Yan Wang, Wen-Fan Feng, Ming-Yue Zhou, Yi-Ming Hu, Shou-Cun Hu, Jian-Wei Mei, and Cheng-Gang Shao. Fundamentals of the orbit and response for tianqin. Classical and Quantum Gravity, 35(9):095008, 2018.DOI
- [77]Chunyu Zhang, Yungui Gong, and Chao Zhang. Source localizations with the network of space-based gravitational wave detectors. Physical Review D, 106(2):024004, 2022.DOI