The advent of gravitational wave (GW) astronomy unlocked the ability to perform unprecedented tests of general relativity (GR) in the strong-field regime. Significant work has been committed to performing black hole spectroscopy on the post-merger ringdown of ground-based signals. Analyses on recent GW events exhibit strong evidence in favor of discerning at least two quasinormal modes (QNMs) in stellar-mass binary black hole ringdowns, which will allow for high-precision tests of fundamental GR principles such as the no-hair theorem and Hawking area law.

We apply methods used in ground-based ringdown analyses to the ringdowns of massive black hole binary (MBHB) GWs expected to be detectable by LISA. We utilize a time-domain configuration that adapts methods from previous ringdown analyses, specifically using a gating and in-painting approach to isolate the ringdown portion of the signal from the inspiral-merger for likelihood calculations. We simulate MBHB signals using a select number of modes, and perform parameter estimation to measure quantities associated with the expected QNMs. By generating and analyzing the signal in this way, we can obtain constraints on the parameters for each known QNM in a LISA signal with a large signal-to-noise ratio. This analysis constitutes the first application of time-domain black hole spectroscopy methods to LISA, and provides insight into the unprecedented level of precision we can achieve in tests of GR using LISA ringdowns compared to ground-based detections.

Strong gravitational lensing provides a powerful tool for probing cosmology and fundamental physics with gravitational-wave observations. With the development of space-based gravitational-wave detectors such as LISA, Taiji Program, and TianQin Project, mergers of massive black hole binaries are expected to be detected with high precision at high redshifts, and a fraction of these signals may be affected by gravitational lensing. In this work, we first investigate the wave-optics effects in gravitationally lensed signals from massive black hole binaries observed by space-based detectors. A Bayesian inference framework is employed to analyze the lens parameters and signal characteristics, in order to assess the detectability of wave-optics lensing signatures in space-based gravitational-wave observations. Building on this analysis, we further explore the application of strongly lensed gravitational waves as standard sirens for cosmology and propose an opacity-free test of the cosmic distance duality relation. By simulating lensed gravitational-wave events detectable with future space-based detector networks, we evaluate the capability of such observations to constrain cosmological parameters and to test fundamental cosmological relations. Our results indicate that space-based gravitational-wave detector networks will provide a new observational window for studying gravitational-wave lensing and for performing precision tests of cosmology.

Extreme mass-ratio inspirals (EMRIs) observable by LISA offer a precise test of modified gravity in the strong-field regime. We investigate a minimal extension of general relativity in which quadratic $f(R)$ gravity is coupled to an ultralight scalar field through two dimensionless parameters, $\beta$ and $\xi$. We derive the modified field equations and obtain analytic source-side field solutions for the curvature and scalar response. By comparing invariant quantities in the modified theory and in general relativity, we identify the fractional difference in the effective stress-energy as a robust measure of deviation from GR. Existing binary-pulsar and solar-system constraints restrict the model to small departures, with fractional invariant differences no larger than $\mathcal{O}(10^{-5}\text{-}10^{-4})$. We use this constrained difference to place conservative upper bounds on waveform-level deviations, including cumulative changes in EMRI cycle count, phase, and chirp time over LISA-band observations. While a full detector-ready LISA observable would require a self-consistent treatment of the modified background geometry, metric perturbations, and detector response, the present results define the viable parameter space of the model and provide a controlled starting point for future EMRI waveform studies in scalar-coupled quadratic gravity.

Matched filtering is a standard technique for gravitational wave detection, but its application to massive black hole binary (MBHB) systems is computationally demanding due to the large size of template banks. This challenge motivates the exploration of quantum algorithms for accelerating the search process. In this work, we investigate quantum matched filtering approaches based on both Grover's algorithm and its extension, the Long algorithm, for MBHB signal detection. Under simplifying assumptions, the quantum framework can theoretically reduce the computational complexity from $O(N)$ to $O(\sqrt{N})$, where $N$ denotes the number of templates. Numerical simulations demonstrate that while Grover-based methods can achieve this quadratic speedup, their performance may degrade in practice due to algorithmic instability. To address this limitation, we further introduce a Long-algorithm-based quantum matched filtering scheme and present its first application to gravitational wave data analysis. The results show that the proposed approach retains the same theoretical speedup while significantly improving robustness and reducing performance fluctuations. This abstract is based on PRD,112(2025)083004 and arXiv:2603.17698

The nonlinear gravitational memory effect (the Christodoulou effect) is predicted to leave a permanent displacement between two objects affected by a gravitational wave burst. The Laser Interferometer Space Antenna (LISA) is predicted to be able to measure this effect. In this poster, we propose a novel means of measuring this effect utilizing the reference white dwarf binaries (WDBs) within LISA’s frequency band. After a gravitational wave burst from a black hole merger passes through the plane of the LISA-WDB system, the nonlinear gravitational memory will manifest as a permanent phase offset in the otherwise extremely stable WDB gravitational wave signal.

Space-based gravitational wave observatories such as LISA, Taiji, and TianQin will present unprecedented data analysis challenges. This talk reports on recent progress in exploring and tackling these challenges, developed within the framework of the second round of Taiji Data Challenge (TDC II).

We first introduce TDC II, a new suite of mock datasets designed to capture more realistic mission complexities, including numerically simulated detector orbits (with arm-length variations up to ~1%), generic second-generation time-delay interferometry (TDI) combinations, complex signal waveforms, and instrumental effects such as non-stationarities, data gaps, and glitches. To facilitate open development, we also present the Triangle toolkit, an open-source framework for realistic data simulation and analysis. Using this framework, we assess the impact of realistic orbit and different TDI channels on data analysis. By comparing results obtained with diverse settings, we quantify differences in signal response, noise characteristics, posterior distributions, and detection sensitivity. The methods developed are directly transferable to LISA and other missions. Our results provide outlooks for the development of future data analysis pipelines.

Data from space-based gravitational-wave detectors typically contain a confusion foreground formed by the superposition of numerous unresolved astrophysical sources. This foreground significantly biases the estimation of the instrumental noise power spectrum and limits subsequent signal validation and physical parameter inference. In this work, we propose a data-driven deep learning approach to directly reconstruct the instrumental noise spectrum in the presence of strong confusion foregrounds. By training on simulated LISA/Taiji data, our model effectively subtracts the overall confusion signal and reliably recovers the underlying instrumental noise power spectrum. Numerical experiments show that the method remains effective in the presence of strong signals such as massive black hole binaries (MBHB) and stochastic gravitational wave backgrounds (SGWB). These results demonstrate that learning-based approaches can extract meaningful noise information under complex signal superposition scenarios, offering a new pathway for noise modeling and data analysis in space-based gravitational-wave astronomy.

Quasars are among the most luminous active galactic nuclei (AGNs) in the universe, powered by massive accretion disks surrounding supermassive black holes at the center of most galaxies. Launching from these accretion disks are energetic outflows of gas and radiation. Among these, quasar extremely high velocity outflows (EHVOs) — defined as outflows exceeding 10% the speed of light — have been confirmed in over 150 cases by our research group, establishing them as a common feature of quasar activity in the early universe rather than rare anomalies. Because kinetic power scales with velocity cubed, EHVOs are suspected to be dominant feedback drivers of early galaxy formation and evolution, including supermassive black hole growth and star formation. However, the fastest EHVOs with speeds greater than 20% the speed of light remain severely underrepresented in existing surveys. Only two have been confirmed to date due to the detection limits of conventional spectral analysis methods.

At these extreme velocities, characteristic ion absorption signatures (CIV, SiIV, NV) are exceptionally blueshifted into the Lyman-alpha (Ly-α) forest region, a dense collection of hydrogen absorption lines from intervening intergalactic gas clouds across the universe which obscures the quasar's intrinsic spectral features. We are developing an automated Python-based detection pipeline to systematically identify EHVOs beyond this 0.2c limit. Our approach includes using the physical contrast in variability timescales between EHVO absorption and Ly-α forest lines: by dividing multi-epoch spectral observations from the Sloan Digital Sky Survey (SDSS), the static Ly-α forest is suppressed to baseline, isolating intrinsic variable EHVO features. We have demonstrated this method on a confirmed EHVO candidate, where CIV absorption consistent with v > 0.2c is clearly revealed in spectral ratios and confirmed by our automated absorption detection module. For cases where this process is insufficient, we are integrating SpenderQ, a redshift-invariant autoencoder, to reconstruct the intrinsic quasar continuum without Ly-α forest contamination. Failing to account for EHVOs in this highest velocity range may be causing a significant underestimation of their cosmological influence on the structure of the early universe. This context is directly relevant to interpreting the energetic environments from which gravitational wave signals detected by LISA will be revealed.

Gravitational waves (GWs) exhibit wave-optics effects when their wavelength is comparable to the scale of the gravitational lens. This may occur in lensing from galactic subhalos in GWs emitted by binary black-hole mergers, and is gaining interest as a novel probe of dark matter. Predictions for observables in these cases ultimately rely on evaluating a Fresnel integral that quantifies the effect of lensing on the amplitude of a GW at a given frequency. However, numerical evaluation of this Fresnel integral is tricky, and several algorithms and publicly available codes that implement it have been developed. Here, we show that the dependence of this integral on the lens position can be written as a two-dimensional Fourier transform. Modern FFT techniques then enable rapid evaluation at all-sky positions simultaneously for general lenses without symmetry. Vectorization of FFT routines allows for derivatives with respect to model parameters to be obtained with only incremental additional computational cost. If the lens is axisymmetric, further speedups can be achieved with recently developed techniques for non-uniform fast Hankel transforms. To demonstrate, we make available Fresnel Integral Optimization with Non-uniform trAnsforms (FIONA), an efficient and accurate code that is significantly faster than current methods for dense source grids, reaching 2 orders of magnitude speedups for ~ 10^6 GW-emitting points. As part of FIONA, we developed code that provides vectorized non-uniform fast Hankel transforms that may have other uses (e.g., calculation of cosmological two-point correlation functions) beyond those considered here.

We consider the potential to use LISA's data to detect Near Earth Objects (NEOs). This potential was assessed by using the masses and distances from the known NEO population compared to LISA's orbit. The mass was modeled based on brightness and asteroid type, which was used to estimate acceleration calculation that will be found in LISA data. The overall discovery potential is assessed by model simulations of the expected entire NEO population. Through these methods LISA was able to detect NEOs, but its ability is a strong function of the mass and proximity to LISA, which ultimately depend on the underlying population distribution

LISA will operate in a confusion-limited regime dominated by unresolved Galactic binaries, where detection becomes a problem of separating resolvable structure from a structured foreground rather than suppressing instrumental noise. We investigate one-class detection in this setting using continuous wavelet transform representations and autoencoder-based latent embeddings trained only on background segments. Holding data generation and preprocessing fixed, we compare local geometric deviation in latent space with explicit latent density modeling. Across independent seeds, likelihood-based scoring consistently outperforms geometry-based manifold distance in separating background-only segments from those containing resolvable sources. The improvement indicates that global distributional structure in latent space carries discriminative information not captured by nearest-neighbor deviation alone, with implications for representation-level detection strategies in LISA pipelines.

The Laser Interferometer Space Antenna (LISA) requires advanced post-processing algorithms to remove noise and recover gravitational wave signals from raw data. The performance of certain noise estimators should be verified using simulations of LISA's arm link response to gravitational waves. Including multiple gravitational wave (GW) source types in these simulations demands tedious application of available software and analysis tools. This project aims to unify research efforts by providing a platform which can quickly incorporate GW sources specified by the user. lisamultisource is an open-source python package developed to synthesize available modern waveform models and arm response simulations into one accessible, easy-to-use environment. This software leverages lisagwresponse to generate time-domain arm response functions for incoming GWs. Sources found in the LISA Data Challenges software will be accessible within lisamultisource, and its outputs are compatible with other tools such as LISANode and LISASim. lisamultisource provided simulations of six GW sources in several configurations to investigate the performance of tilt to length coupling estimators. These simulations can be injected into noise models and be used to develop data analysis techniques, such as time delay interferometry. Current efforts are underway to make this software public, with a modular and visual interface. Broadly, this software could be applied to instrumental research requiring arm response simulations that contain multiple gravitational-wave signals. Unlike existing tools, lisamultisource integrates modern waveform models into a single platform, reducing the complexity of generating multiple GW response simulations.

BBH mergers are observed by ground-based detectors such as LVK and soon by space-based detectors such as LISA. AGN dynamics may contribute to a significant fraction of LISA observables, yet BBH merger dynamics within AGN remain ambiguous despite appearing in GW observations. Monte carlo For AGN Channel Testing and Simulation (McFACTS) simulates BBH populations within AGN disks detectable by LVK and LISA. In McFACTS, BBH mergers evolve through gas drag and dynamical interactions within the disk without calculating the relativistic effects of final merger and remnant properties. I improved on McFACTS by routing inspiral BBH parameters into the surfinBH numerical relativity surrogate model to more precisely predict the merger remnant kick velocities, masses, and spins than current analytical models. The addition of this surrogate model shows that BBH mergers in McFACTS require numerical relativity to accurately describe BH merger population contributions to disk structure and lifetime.

Supermassive black hole binaries (SMBHB) will be prime multimessenger sources in the era of low-frequency gravitational waves (GW). With advanced X-ray observatory NewAthena planned for operation in the mid-2030s, it is crucial to understand how SMBHBs may be both identified and characterized through high signal-to-noise, high spectral resolution X-ray observations. To that end, we use relativistic reflection code relxill to compute the time-dependent spectra originating from mini-disks attached to two SMBHs at a separation of 100 gravitational radii. We show that SMBHBs are expected to produce distinctive reflection spectra and we investigate how simple, model-independent measurements of prominent features of the X-ray reflection spectrum of a SMBHB (such as the broad Fe K?? line and Compton hump around 20-30 keV) may be used to place constraints on binary parameters like inclination, mass ratio and spin. These results highlight the potential for X-ray spectroscopy to complement GW measurements by the Laser Interferometer Laser Antenna (LISA) and Pulsar Timing Arrays (PTAs).

Cosmological simulations are a powerful tool for studying massive black hole (MBH) binary populations. Due to their limited spatial resolution, MBH binaries (MBHBs) identified in simulations must be evolved in post-processing from kiloparsec to sub-parsec separations using binary hardening models. We present improvements to these models and to MBH subgrid dynamics, including circumbinary disk hardening, triple MBH interactions, and gravitational wave (GW) recoil kicks. Using the BRAHMA cosmological simulation suite, which incorporates multiple MBH formation mechanisms, we probe MBHBs relevant to LISA. We predict MBHB merger rates across different black hole seeding scenarios, along with the orbital eccentricities of MBHBs as they enter the LISA frequency band and the detectability of higher harmonic GW modes. Finally, by linking MBHBs to their host galaxy properties, we also provide guidance for electromagnetic (EM) follow-up of LISA GW sources.

Black hole mergers produce asymmetric gravitational wave emission, giving the remnant a recoil kick that can reach thousands of km/s, sometimes ejecting it from its host galaxy. While this effect is well established theoretically, cosmological simulations have not yet implemented realistic, spin-dependent recoil kicks drawn directly from simulated merger properties. We present ongoing work to do exactly that using ChaNGa Nbody + SPH cosmological simulations, where kicks are applied on the fly and mass ratios come directly from the simulation. We systematically explore five spin configurations: three angle orientations (aligned,random, anti-aligned) combined with two magnitude assumptions (maximum and random spin). For each configuration we report on the occupation fraction of black holes in galaxies, measuring how many remain near their galactic centers versus get ejected. This method will also enable us to track whether recoiled black holes return to their galactic centers over time.

Binary black holes (BBH) embedded within the gaseous environment of an active galactic nuclei (AGN) are more quickly hardened by interactions with the surrounding gas than dynamical encounters with other objects embedded in the disk. Mergers resulting from these BBHs are known as 'the AGN channel' and are estimated to be a significant fraction of the detections made by Ligo/Virgo/Kagra (LVK). Such BBH are also expected to appear as long lived signals detectable by the Laser Interferometer Space Antenna (LISA). To further understand how BBH evolve in AGN, we employ the Monte carlo For AGN Channel Testing and Simulation (McFACTS) software and implement a treatment for stalling the effects of gas drag on the BBH. Removing the gas drag as a driving force of the BBH evolution allows for dynamical encounters and general relativity to be the primary factors pushing the BBH towards merger. Building on previous work by the McFACTS collaboration, we investigate the implications of stalling the evolution of BBH at various semi-major axes. Assuming different physical reasons for why the stalling could take place at these set distances, we examine how these changes can affect the dynamical timescale of mergers expected from AGN in our universe and the resulting effects of the LVK and LISA populations.

Dual AGN activity is an electromagnetic precursor of SMBH mergers that produce gravitational waves detectable by LISA. To make accurate predictions of the dual AGN population and therefore the SMBH merger rates, one needs to know how often they occur and for how long they remain active. Furthermore, galaxy mergers are interesting events as gravitational torques during mergers drive gas towards galactic nuclei, which can enhance AGN activities. AGN feedback can also delay the merger times. Thus, we investigate and characterize the single and dual AGN activities using isolated and merging galaxy simulations called Stars and MUltiphase Gas in GaLaxiEs (SMUGGLE). Multiphase interstellar medium in SMUGGLE simulations yields highly variable accretion rates with short duty cycles. We found the mean active phase timescales to be in the order of ~0.1 Myr to ~1 Myr and quantified dual AGN lifetime fractions in various merger scenarios. Differences in black hole masses, galaxy morphologies, and wind speeds suggest a significant impact on single and dual AGN activities. This work presents a method for modeling realistic (dual) AGN populations in semi-analytic models and simulations, and thus for characterizing BH populations using constraints from EM and GW observations.

The detection of gravitational waves (GWs) from massive black hole binary(MBHB) coalescence motivates the development of a sub-grid model. We present RAMCOAL, which simulates the orbital evolution of MBHBs, accounting for stellar and gaseous dynamical friction (DF), stellar scattering, circumbinary disk interactions, and GW emission at scales below the simulation resolution. RAMCOAL also models MBH triplets and quadruplets which is critical in the early Universe. This model is now a flexible stand-alone library to be integrated into the any cosmological simulation code. Unlike post-processing approaches, RAMCOAL tracks the real-time evolution of MBHBs within simulations using local quantities to model dynamics and accretion. This enables more accurate predictions of both GW signals and the properties of merging black holes. We have validated RAMCOAL across isolated and merging galaxy setups at resolutions of 10, 50, and 100 pc, with and without black hole accretion and feedback. In addition, we test the model in seven galaxy merger scenarios at 100 pc resolution. These tests demonstrate that RAMCOAL is largely resolution-independent and successfully captures the effects of DF from stars, dark matter, and gas, loss-cone scattering, viscous drag from circumbinary disks, and GW emission – all within a realistic galactic environment, even at low resolutions. With RAMCOAL, we can better estimate MBHB coalescence rates and the GW background, while providing insights into the electromagnetic counterparts of GW sources. This approach bridges the gap between electromagnetic observations and GW detection, offering a more comprehensive understanding of MBHB evolution in cosmological simulations.

In this talk, I will first present predictions for massive black hole (MBH) mergers detectable by LISA down to z=0 based on the ASTRID cosmological simulation. ASTRID explicitly models MBH dynamical friction, enabling realistic orbital evolution. It incorporates relatively low-mass MBH seeds down to 4×10^4 solar mass within a large volume of 370 Mpc per side, providing a more complete picture of LISA MBH mergers. We account for orbital eccentricity in our predictions, which extends the detectable MBH mass range. We also investigate the host environments and electromagnetic signals of these LISA sources. I will also present the latest results from ASTRID-BRAHMA, a simulation that combines the physically motivated gas-based MBH seeding model of the BRAHMA suite with the cosmological volume of ASTRID. It reproduces some high-redshift MBHs observed by JWST and yields a large population of Little Red Dots (LRDs). At the same time, the larger MBH seed population significantly enhances the LISA detection rate at the high redshift, highlighting the LISA’s power to constrain MBH seeding models.

The efficiency of stellar hardening as a mechanism for massive black hole (MBH) binary orbital decay in gas-poor galaxy centers has been contested for years. Detailed numerical simulations are required to resolve these collisional processes, but they are comptuationally expensive and rarely capture the full range of galaxy types that may host LISA sources. Analytic recipes for hardening timescales are useful but often rely on inward extrapolation of galaxy properties, which in cosmological simulations can be orders of magnitude larger than the sphere of influence of the binary. Instead, we apply a merger timescale that is a function of easily accessible galaxy properties to the MBHs of the Romulus25 simulation and compare with existing MBH merger and detection rates for LISA. We emphasize the importance of nuclear star clusters (NSCs) in enhancing stellar hardening and discuss the implications for galactic stellar cores and host galaxy identification in the multi-messenger era.

Theoretical models of the evolution of supermassive black hole (SMBH) pairs in post-merger remnant galaxies are necessary to motivate observational searches for dual active galactic nuclei (AGN) and gravitational wave (GW) sources. Since LISA will detect low-frequency gravitational waves from SMBH binaries with masses in the 10^4–10^7 M⊙ range, understanding which systems can overcome the prolonged dynamical friction phase of SMBH evolution is critical for predicting its source population. We formulate a 3D dynamical model of SMBH pairs in the innermost kiloparsec of a post-merger galaxy to investigate the impact of orbital inclination with respect to the galactic disk on pairing times. We find that increasing orbital inclination prolongs SMBH pairing times on average. Pairing times for orbits with inclinations \gtrsim 45 degrees often exceed 14 Gyr, while systems with inclinations \lesssim 20 degrees evolve rapidly enough to remain viable progenitors of dual-AGN and GW sources. The model suggests that orbital geometry is a critical determinant of which SMBH pairs contribute to electromagnetic and GW survey populations.

Rubin will discover hundreds of thousands of AGN in the Legacy Survey of Space and Time. Among these, a significant population may harbor dual supermassive black holes at close separations—direct LISA gravitational wave progenitors. However, identifying dual AGN from Rubin data alone remains challenging. The LSST AGN Follow-up Group is developing coordinated multi-wavelength follow-up strategies—combining optical, near-infrared, X-ray, and radio observations—to systematically identify and characterize dual AGN candidates. We present identification methods, observational priorities, and how multi-wavelength follow-up directly constrains dual AGN merger rates, black hole masses, and orbital timescales essential for predicting the gravitational wave populations detectable by LISA and pulsar timing arrays.

A black hole merger remnant can acquire a recoil velocity, or kick, due to anisotropic gravitational wave emission during coalescence. Numerical relativity simulations predict recoil velocities up to 4000 km/s. A typical Milky Way-like galaxy has an escape velocity of around 700 km/s; thus, some kicks can be strong enough to eject black holes out of their host galaxy entirely. Despite the significant implications on black hole demographics and host galaxy properties, many semi-analytic models often exclude this gravitational recoil effect. In this study, we implement a gravitational recoil prescription, assuming a uniform black hole spin distribution, into the Semi-Analytic Galaxy Evolution models (SAGE and Dark SAGE). We present preliminary results on how recoil affects black hole growth, stellar mass assembly and metal-enrichment within a galaxy population.

I introduce EMRICode, a Python-based numerical framework for computing evolving Extreme Mass Ratio Inspiral (EMRI) trajectories and associated gravitational waveforms in Kerr spacetime. The code integrates test-particle geodesics with adaptive timestep control and includes leading-order radiation-reaction effects. It is designed to balance computational efficiency with physical fidelity, enabling large-scale inspiral surveys, parameter studies, and rapid waveform prototyping in preparation for the Laser Interferometer Space Antenna (LISA) mission. I describe the underlying equations of motion, numerical implementation, and multipole waveform generation scheme, and I assess the code's performance through extensive tests of orbital constant conservation and inspiral evolution.

Binary black holes (BBH) merging in accretion disks of active galactic nuclei are potential gravitational wave sources observable with LVK and, in future, LISA. McFACTS, an open-source population synthesis code, models black hole (BH) interaction and subsequent BBH evolution within AGN accretion disks, which is detectable with LVK. However, additional functions critical to some users can be run entirely after the initial McFACTS run. I demonstrate our new package of post-processing tools for McFACTS, including BBH merger trees and tracking for individual BBH progenitor properties across simulated time. These investigations provide crucial insight into the histories of LISA-observable BBH GW sources, following their evolution from their final observed state all the way back to their first progenitor BH, and how this evolution will appear in the frequency band of LISA and other GW observatories.

The LTPDA toolbox (Hewitson et al., 2009), developed as the data analysis environment for the LISA Technology Package, provides spectral analysis, noise characterization, filtering, and signal modelling. Every analysis object records its complete processing history, making analyses reproducible and auditable. An accompanying repository server allows teams to store and share fully-annotated objects, preserving entire annotated analysis pipelines.

The software remains widely used, but with the last release targeting MATLAB R2012b (2017), compatibility had silently broken over nearly a decade. We present a full modernization: the toolbox runs on R2025a, addressing deprecated language features and outdated interfaces, and introducing conveniences like automated SSH tunnelling for secure repository communication. The repository server has been rebuilt as a containerized Python/JavaScript stack with integrated MySQL, delivering a plug-and-play system suited to modern academic IT infrastructure and ready for the LISA mission era.

Accurate modeling of gravitational-wave signals in the LISA band requires quantifying how higher-order modes and spin-precession influence detectability across a broad parameter space. While higher-order modes are expected to be prominent in asymmetric systems, their contribution to signal-to-noise ratio (SNR) and detectability has not been systematically characterized for diverse LISA sources. In this work, we investigate these effects using waveform models such as IMRPhenomXPHM and IMRPhenomPv3 for massive black hole binaries (MBBHs), intermediate-mass black hole binaries (IMBHs), and extreme mass-ratio inspirals (EMRIs). We perform detailed mode-by-mode SNR analyses across varying mass ratios and spin configurations, constructing parameter space maps to identify regions where subdominant modes significantly enhance the total SNR. We find that higher-order modes can elevate signals above detection thresholds in specific regions, particularly for asymmetric and high-spin systems, while spin-precession introduces additional modulations that influence detectability. Ongoing work extends this analysis to inclination angle and broader parameter regimes, with implications for improving waveform modeling and LISA data analysis strategies.

Previous work has determined that there will be systematic biases in parameter estimation on massive black hole binaries observed by LISA if we use waveform templates that do not model radiation in enough higher-order multipoles. These biases become severe for binaries with total redshifted mass above about a million solar masses, which includes a large fraction of the events expected to be observed with LISA. The severity of the biases also depends sensitively on factors such as the mass ratio, inclination, and individual spins. I will discuss our work in quantifying these effects and working towards an understanding of how many modes will ultimately be required to perform unbiased parameter estimation on massive LISA sources.

The final stage of a gravitational wave signal from a binary black hole merger, known as ringdown, offers a unique opportunity to probe extreme curvature and strong-field gravity, making it a key tool for testing black hole (BH) dynamics and fundamental physics. However, physical models of BH resonances driving the ringdown regime are only known in the late-time stationary regime. Understanding their robustness when extrapolating to earlier times, and recovering a larger signal portion, is currently a key open problem in the field.

We resolve this question by performing a detailed study of ringdown models accuracy against numerical relativity simulations. We determine the validity regime of ringdown models for a given experimental accuracy, comparing the performances of state-of-the-art waveforms, and underline important limitations related to third-generation detectors like LISA.

Our methodology provides a robust framework for evaluating ringdown models validity and will be instrumental in future Quasi-Normal Mode observational searches.

Based on: https://arxiv.org/abs/2511.02915

Fast and efficient Bayesian analysis will be critical for LISA. One key approach is marginalizing the likelihood over extrinsic parameters, such as the overall phase of the signal. Current phase marginalized likelihood models are restricted to signals containing only a single mode, requiring sampling over the phase in the presence of higher modes, decreasing the efficiency of parameter estimation. Signals from sources such as Extreme Mass Ratio Inspirals, one of LISA’s targets, are expected to contain an appreciable amount of power in the higher modes, potentially leading to expensive parameter estimation. To address this, we provide a robust, accurate, analytic phase marginalized likelihood model that is applicable even in the presence of arbitrary numbers of modes. This is done by using the saddle point approximation of the integral of likelihood over the phase, with added corrections. We also discuss the possibility of using the closed form solution to potentially marginalize over other extrinsic parameters such as distance, polarization, or inclination.

Massive black hole binaries (MBHBs) and other sources within the frequency band of spaceborne gravitational wave observatories like the Laser Interferometer Space Antenna (LISA), Taiji and Tianqin pose unique challenges, as gaps and glitches during the years-long observation lead to both loss of information and spectral leakage. We propose a novel data imputation strategy based on Kalman filter and smoother to mitigate gap-induced biases in parameter estimation. Applied to a scenario where traditional windowing and smoothing technique introduce significant biases, our method mitigates the biases and demonstrates lower computational cost compared to existing data augmentation techniques such as noise inpainting. Preliminary results also show potential in treating gaps in whitened time domain. This framework presents a new gap treatment approach that balances robustness and efficiency for space-based gravitational wave data analysis. For details see:https://arxiv.org/abs/2507.02458.

LISA data analysis poses significant challenges. The data stream will always contain multiple, overlapping signals from a wide variety of sources. Moreover, the number of sources in the data will not be known a priori and must be inferred during the analysis. We refer to pipelines capable of performing full inference as Global fit. For these reasons, a set of past and planned mock data challenges of increasing complexity has been designed to help us build and improve global fit prototype pipelines. In this talk, I will describe the novelties and challenges posed by the latest simulated dataset, ranging from non-stationary noise to the presence of extreme mass-ratio inspirals, and the inclusion of multi-modal waveforms for massive black hole binaries. I will then present the advancements we are applying to one of the existing Global fit pipelines, Erebor, to face these challenges. In particular,

Global fitting of compact binary signals is commonly treated as a high-dimensional inference problem, addressed via sequential subtraction or joint Bayesian sampling. In this work, we propose an alternative formulation that views global fitting as a long-horizon decision process, where the objective is to iteratively reduce the explainability of the data with respect to a given signal class.

We cast this problem as a Markov Decision Process, in which actions correspond to subtracting physically interpretable subsets of the signal space, and rewards are defined through reductions in a likelihood-based global explainability metric. To support this, we construct a learned latent representation combining unsupervised modeling with weak physical supervision, enabling a tractable and interpretable action space.

We present preliminary results on Taiji Data Challenge datasets, demonstrating systematic reduction of residual explainability while maintaining consistency with noise and unresolved foreground assumptions. Ongoing work applies the same framework to LISA Data Challenge datasets for cross-validation and comparison.

This work introduces a new statistical framing of global fitting as a controlled information-depletion process with explicit and testable criteria for residual consistency.

Stellar-origin binary black holes (SOBH) produce long-duration gravitational wave signals in the LISA band, requiring consistent modeling of both the waveform and the detector response. SOBH sources range from minimal orbital evolution over LISA's observation window to signals that will chirp out of the frequency band over time on the order of the mission duration. We develop algorithms for the detection and characterization of these sources, with the ultimate goal of analyzing SOBHs in the LISA global fit pipeline. We validate our pipeline against the Mojito dataset produced by the European Distributed Data Processing Center (DDPC), demonstrating consistent signal recovery and the expected behaviour of our sampling operations. We report on our current status, initial results, and next steps.

The detection of gravitational waves from extreme-mass-ratio inspirals (EMRIs) in space-based antennas like Taiji and Laser Interferometer Space Antenna promises deep insights into strong-field gravity and black hole physics. However, the complex, highly degenerate, and nonconvex likelihood landscapes characteristic of EMRI parameter spaces pose severe challenges for conventional Markov chain Monte Carlo (MCMC) methods. Under realistic instrumental noise and broad priors, these methods demand impractical computational costs but are prone to becoming trapped in local maxima, leading to biased and unreliable parameter estimates. To address these challenges, we introduce flow-matching MCMC (FM-MCMC), a novel Bayesian framework that integrates continuous normalizing flows (CNFs) with parallel tempering MCMC (PTMCMC). By generating high-likelihood regions via CNFs and refining them through PTMCMC, FM-MCMC enables robust exploration of the nontrivial parameter spaces, achieves orders-of-magnitude improvement in computational efficiency, and, more importantly, ensures statistically unbiased inference. By enabling real-time, unbiased parameter inference, FM-MCMC could unlock the full scientific potential of EMRI observations, and would serve as a scalable pipeline for precision gravitational-wave astronomy.

Extreme mass-ratio inspirals (EMRIs) on generic Kerr orbits are expected to intersect the accretion disk of their central black hole periodically, producing quasi-periodic X-ray eruptions (QPEs). The timing of these disk crossings constitutes a Poincaré surface of section of the three-dimensional orbital trajectory — a sparse, high-signal-to-noise but geometrically incomplete projection of the dynamics. The associated gravitational wave signal, by contrast, encodes information about the full trajectory as a continuous time series. We compare parameter estimation from these two observational channels for a fiducial EMRI system. Using Bayesian inference on synthetic QPE crossing times and the corresponding LISA GW signal, we show that the two data streams constrain largely complementary combinations of the source parameters — including central mass, mass ratio, orbital elements, spin, and disk geometry. We discuss the biases inherent to each approach and the prospects for joint EM+GW PE as a discovery channel for LISA EMRIs.

As the Laser Interferometer Space Antenna (LISA) mission enters Phase B, the focus of the gravitational wave (GW) community has shifted from technology demonstration to the robust extraction of overlapping signals. Among the most anticipated but computationally challenging sources are Intermediate Mass Ratio Inspirals (IMRIs)—the inspiral of stellar-mass compact objects into intermediate-mass black holes (10^2 - 10^4 M_\odot). These sources are pivotal for mapping the missing link of black hole evolution and testing General Relativity in the strong-field regime.

A Conditional Variational Autoencoder (CVAE) model is employed for parameter inference on gravitational waves (GW) signals of massive black hole binaries, considering joint observations with a network of three space-based GW detectors. Our experiments show that the trained CVAE model can estimate the posterior distribution of source parameters in approximately one second, while the standard Bayesian sampling method, utilizing parallel computation across 16 CPU cores, takes an average of 20 hours for a GW signal instance. However, the sampling distributions from CVAE exhibit lighter tails, appearing broader when compared to the standard Bayesian sampling results. By using CVAE results to constrain the prior range for Bayesian sampling, the sampling time is reduced by a factor of $\sim$6 while maintaining the similar precision of the Bayesian results. This work is based on our publication in Physical Review D (111, 103053).

We study systematic noise biases in LISA and their influence on sources expected to be observed within the LISA band of gravitational wave frequencies. We focus on pre-science-data-processing (L0-L1) noise arising from optical misalignments between components in the optical bench’s interferometer and angular jitters of the LISA Moving Optical Sub-Assemblies (MOSAs) and test masses in the longitudinal output measurements. This jitter, known as tilt-to-length coupling (TTL), can appear in the data as excess noise, potentially biasing a waveform during parameter estimation. We present preliminary parameter outputs from an MCMC sampling that demonstrate the systematic biases found when these noises are left at various levels of mitigation. We plan to use open-source Python packages to estimate TTL parameters in conjunction with waveform parameters, independent of other L0-L1 effects. This can then be generalized into a TTL-estimation pipeline that can be integrated into LISA’s global fits.

Verification of the magnetic field strength at 98.304 kHz produced by the Charge Management Device (CMD) is desired to ensure prevention of interference at the Gravitational Reference System sensing frequency. The electro-magnetic compatibility requirements specify the magnetic field’s amplitude spectral density shall not exceed 0.42 pT/rt(Hz), and the amplitude of the AC magnetic field shall not exceed 1nTrms. At the University of Florida, we have performed lock-in amplifier measurements to confirm the compliance of the CMD UV Light Unit with the specified requirements. Here we present the measurement procedure and current results from these efforts.

In view of the upcoming LISA Gravitational Reference Sensor (GRS) testing campaign, we have implemented a significant upgrade to our four test mass (4-TM) torsion pendulum facility by installing a fused silica fiber. This modification is designed to suppress thermal noise levels significantly compared to those of traditional tungsten fibers. Moreover, we should have a substantial increase in the mechanical quality factor, Q. These technical enhancement allow for the establishment of more stringent upper limits on GRS acceleration noise, thereby increasing the physical significance of our on-ground experimental conclusions. In particular, we present the results of an experimental campaign dedicated to measuring the sensing stiffness. Which is a parasitic force, modeled as a spring with negative spring constant, arises from the electrostatic interaction between the test masses and the surrounding capacitive position sensor (electrode housing).

At the heart of the LISA spacecrafts are Gravitational Reference Sensors (GRS), responsible for maintaining the geodesic motion of enclosed test masses (TMs). To isolate gravitational wave signals, it is essential to minimize non-gravitational disturbances on the TMs such as electrostatic charging of the test masses. The Charge Management System (CMS) developed at the University of Florida (UF) addresses the issue by supplying 250 nm UV light to control the charge of freely floating test masses via the photoelectric effect.

We present an overview and performance characterization of the CMS, including a complete finite element analysis (FEA) model of the system. The results are compared to experimental performance data, specifically the apparent yield (AY) of the sensor which allows for estimation of the microscopic properties of the sensor surfaces. These simulations support performance verification of the CMS for LISA and offer broader applicability in similar inertial sensors for missions in geodesy, fundamental physics, and astrophysics.

LISA Pathfinder showed the presence of transient force events acting on the test masses, known as glitches. At present there is no evidence about the physical origin of these events. The glitch source that looks more consistent with the experimental landscape is related to some kind of pressure phenomenon, in particular sudden outgassing events from the various components surrounding the TM. We are performing numerical simulations in Trento with the HPC (High Performance Computing) cluster in order to have a more complete analysis for the different components of the GRS and to better understand the physics behind glitches. We have obtained new results with an improved precision for the mean momentum per molecule transferred to the TM. I discuss the impact of these findings on the LISA design and operation.

Residing in the space environment, the free-falling test masses of the LISA gravitational reference sensors are subject to bombardment by galactic cosmic rays and solar energetic particles. Over time, charge accumulates on the test masses from these charged particle interactions, leading to force noise in LISA's gravitational wave measurements. To reduce this noise, the test mass charge is controlled using UV LEDs, expelling photoelectrons via the photoelectric effect. The UV LEDs are synchronously pulsed with the 98.3 kHz electric field, where shifting the phase allows for control of the photoelectron current direction. Under typical conditions, the dominant source of test mass charging is galactic cosmic rays, for which the charge dynamics and force noise have previously been modeled. However, during solar energetic particle (SEP) events the test masses are flooded with charged particles from the Sun, leading to large, temporary increases in the test mass charge and charge noise. The charge control performance and force noise during SEP events are presented.

The Laser Interferometer Space Antenna (LISA) will open a new observational window on the Universe by detecting gravitational waves in the millihertz frequency band. Achieving the required sensitivity relies on the precise realization of a drag-free system based on the Gravitational Reference System (GRS), which monitors the motion of free-falling test masses inside each spacecraft. Building on the technological heritage of LISA Pathfinder, the Italian Space Agency (ASI) is supporting the development of key elements of the Gravitational Reference System for the LISA mission. Current activities include the industrial development of the system and its subsystems, the realization of Structural and Thermal Models, and the maturation and qualification of critical components such as the test mass release system, the vacuum system of the test mass housing, and the electronics power control unit. These efforts are carried out in close collaboration with international partners contributing additional subsystems and electronics. Future work will focus on system consolidation, qualification and integration, leading to the delivery of the qualification and flight models for the LISA spacecraft constellation.

The Test Mass Emulator (TME) is a key ground check-out element of the avionic model. It will be used in a HIL configuration to validate DFACS procedures and algorithms as well as the behavior of the FEE without a free-falling TM. The TME will replace the GRS with a fixed TM and a set of mobile sensing electrodes. Actuators with tens of nm repeatability will move the electrodes towards and away from the TM in order to emulate the behavior of the mobile TM in the EH and generate equivalent electrical signals to emulate FEE readings. TME takes as input electrode housing capacitances, simulated by the Real-Time Simulator, evaluates the target current and selects the actuator position from a calibrated lookup table that reproduces the required currents. In this work, we present theoretical and empirical assessments of actuator positioning uncertainties and TME signal accuracy that drive the definition of TME functional requirements and the spatial granularity of the GRS-TM position that can be emulated. We estimated the positioning accuracy as a function of sampling frequency and TM maximum speed. We also evaluated how electrodes' thermoelastic deformation impacts the current fluctuations. Moreover, we verified, in terms of execution time and error estimates, that the system meets the timing and frequency requirements of the test cycles. The individual contributions are then propagated to the overall positioning uncertainty using simplified Monte-Carlo models. The resulting error budget identifies the dominant contributions and their margins with respect to the target positioning requirement, leading to functional requirements definitions on actuator performance, thermal stability, and calibration strategy for the LISA TME.

The Laser Interferometer Space Antenna (LISA) mission is designed to detect low-frequency gravitational waves using free-falling test masses housed within spacecraft. These test masses are subject to slow charge build-up from external cosmic rays on the spacecraft. As the LISA constellation moves through the interplanetary magnetic field, the charged test masses will be subject to Lorentz forces. This project presents a computational study of the Lorentz force acting on the LISA test masses as a function of time. The analysis focuses on the variation of the interplanetary magnetic field intensity associated with the solar activity cycle, including solar maximum and solar minimum conditions. Using numerical methods, the resulting Lorentz force is evaluated to study how changes in the interplanetary magnetic field strength may influence the disturbance environment of the LISA mission.

NASA Goddard Space Flight Center (GSFC) is developing the laser system (LS) for the Laser Interferometer Space Antenna (LISA) mission. A complete LISA Laser system consists of a laser head (LH), frequency reference system (FRS) for frequency stabilization and power monitor detectors (PMON) for power stabilization.

We will present an overview of the LH architecture, which includes a LOM (Laser Optical Module) and a LEM (Laser Electrical Module), as well as the latest LH development activities aimed at qualifying the LOM to TRL-6 level through thermal-vacuum cycling tests. These tests with an Engineering Development Unit (EDU)-1 LOM driven by a LEM demonstrated compliance with stringent LISA laser performance requirements. The LH performance was also independently verified by CSEM under similar thermal vacuum environment.

We continue the LH development in advancing the EDU-1 LH to EDU-2 LH including significant volume reduction to meet LISA flight laser requirements.

NASA Goddard Space Flight Center (GSFC) is developing the laser system (LS) for the Laser Interferometer Space Antenna (LISA) mission. The LS under development at GSFC includes a laser head, a frequency reference system, and a power monitor detector assembly. As part of the current agreement between ESA and NASA, GSFC has already delivered several hardware units to ESA.

In this talk, we will provide an overview of the ongoing development activities and the roadmap for advancing the LISA LS toward spaceflight qualification and the delivery of multiple flight units to ESA. We will describe the optomechanical and electronic design of each subsystem, summarize recent test results, and discuss the technical challenges addressed during development. Notably, the latest laser head was independently tested by an ESA contractor, confirming the performance previously measured at GSFC in a relevant environment. We will also outline the planned test campaigns, including those that will incorporate ESA provided hardware.

Together, these efforts demonstrate steady progress toward delivering a robust, fully qualified laser system that will meet the mission requirements for LISA.

The Frequency Reference System (FRS) for the LISA laser system has been in development at BAE Systems SMS since 2022. BAE Systems SMS, in collaboration with NASA GSFC, has successfully advanced the Technology Readiness Level (TRL) of the FRS from 4 to 6. The FRS consists of the FRS-O (“Optical” – for the optical reference cavity) and FRS-E (“Electrical” – for the locking electronics). The FRS-O has space flight heritage in the GRACE FO optical cavity, used in the Laser Ranging Interferometer. The FRS-E (Electrical) was a new design leveraging Stable Laser Systems’ FPGA servo control unit. The FRS-O and FRS-E together provide feedback control to stabilize the LISA laser to the frequency noise requirement. The FRS-E passed environmental test (TVAC and vibration test) at BAE Systems SMS in Fall 2025 and has demonstrated locking performance with the FRS-O in benchtop test. The FRS-O passed TVAC test at NASA GSFC in Summer 2025, tested with FRS-E, and LEM (Laser Electronics Module) and MO (Main Oscillator), meeting performance at all operating plateaus.

Avo Photonics has been contracted by NASA Goddard Spaceflight Center (GSFC) to design, develop, and manufacture the Main Oscillator (MO) seed laser for the LISA mission. An overview of the design and the most recent test results will be presented in this poster.

The MO is a critical component of the full LISA Laser System being developed by Goddard. It serves as the seed laser for LISA’s interferometer operation. As such, minimizing intensity noise and frequency noise was an imperative consideration in Avo’s design. To achieve this, the MO utilizes a diode-pumped micro-non-planar ring oscillator laser, which has been proven to meet the LISA mission’s stringent noise requirements. Reliability is also paramount, and the design has incorporated various elements of ruggedness and redundancy to address this.

Recently, Avo Photonics completed the manufacturing of 24 engineering models of the MO. These units underwent performance testing and limited qualification testing at our facility, including thermal cycling, power characterization, and noise testing. Select units were delivered to Goddard, where they underwent a complete qualification regimen including shock and vibration, radiation, and thermal vacuum testing. The results of these tests showed that the units successfully meet the specifications required for the LISA mission.

The relative intensity noise (RIN) in the laser beam impinging each LISA test mass (TM) must be held below 100 ppm/√Hz in the mHz LISA observation band to preserve their geodesic motion. The laser system’s Main Oscillator Power Amplifier (MOPA) architecture is exploited to ensure this stability requirement is compatible with all other requirements. The Power Monitor (PMON) sub-system, developed by Avo Photonics, Inc., houses two photodetectors (both for redundancy and for out-of-loop detection) that measure the fluctuations in the laser power and feed back into the power amplifier (PA) to maintain a constant, low-noise optical power at each TM. We have built a RIN stabilization test bed to evaluate the performance of the PMON Engineering Development Units (EDU), integrated the test bed with the EDU-1 Laser Head (LH), and demonstrated less than 100 ppm/√Hz RIN stability in the observation band.

Light sent between the LISA spacecraft experiences a ~8.3 second time delay due to LISA’s 2.5 Gm arm length. Additionally, the triangular constellation evolves throughout its orbit. The Point Ahead Angle Mechanism (PAAM) will provide an adjustable point-ahead angle to compensate for the travel time of light between spacecraft and their evolving orbit.

The PAAM rotation axis must be placed with 50 μm precision on the LISA optical bench to ensure sufficiently low tilt-to-length (TTL) coupling, such that the angular jitter of the PAAM contributes less than 0.5 pm/sqrt(Hz)*U_OMS(f) piston noise. The University of Florida LISA group will measure the rotation axis placement precision of the PAAM Engineering Model 1 (EM1) using a PAAM substitute target to pre-align a laser beam to the predicted rotation axis; the PAAM EM1 is then swapped in and the TTL coupling is measured with a heterodyne interferometer. This poster presents the current status of such measurements.

Sideband locking techniques are a form of laser frequency stabilization which allow for a sideband of a laser beam, rather than the carrier, to be locked at resonance to an optical cavity. These schemes are more pragmatic for certain applications than conventional Pound-Drever-Hall (PDH) locking. For instance, such techniques could be implemented in some proposed subsystems of LISA, where additional cavity length readouts would be required. While conventional PDH locking would require additional laser heads at each spacecraft in such cases, sideband locking requires only the pick-off of a pre-stabilized beam. We therefore implemented an Electronic Sideband Locking (ESB) scheme – a compact FPGA-based technique which maintains a sideband of the laser at a cavity resonance via the tuning of a phase modulated Numerically Controlled Oscillator (NCO) signal. We demonstrate the displacement readout sensitivity of this scheme for a high-finesse cavity and compare it against the LISA required sensitivity curve, showing that it meets mission requirements for precision optical cavity readout.

Heterodyne interferometry for space applications demands robust electronic architectures capable of meeting stringent performance, environmental, and reliability requirements.

FPGAs represent an ideal solution for addressing these challenges: their (re)configurable structure enables the integration of complex designs including multi-clock domains, communication interfaces, high-speed data handling and specialized IP cores into a single device.

By leveraging Rad-Hard FPGAs, featuring hardware-managed Triple Modular Redundancy (TMR) and Majority Voting, alongside rigorous design guidelines, it is possible to achieve exceptional performance in space environment, such as the picometer-level sensitivity for detecting distance variations between reference masses in LISA satellites.

This poster presents the architectural block diagram of the Phasemeter and provides a detailed analysis of the strategies adopted to ensure long-term system reliability in demanding deep-space conditions.

Each spacecraft in LISA consists of two moving optical sub-assemblies (MOSAs). Changes in the LISA constellation incurred by celestial mechanics are accommodated by the 'moving' attribute of the MOSA. The phase relation between the two lasers on the two MOSAs is derived using a bidirectional fiber backlink that connects the two MOSAs. The Reference Interferometer on each MOSA measures such a phase relation that is required to synthesize the changes of the distances in two adjacent arms into a Michelson interferometric signal that carries information on the incoming gravitational wave. Via such a bidirectional Phase Reference Distribution System (PRDS), the two lasers serve the role of a local oscillator for each other to generate the heterodyne beat note signal with the phase information for distance measurements. In this presentation we will summarize the PRDS development at AEI spanning from an early conceptual demonstration to a trade-off study to a recent EM-level implementation.

NASA Goddard Space Flight Center (GSFC) is responsible for the development and production of fiber optic assemblies throughout the LISA mission, including the assemblies that connect the Laser System (LS) to the Optical Bench Assembly (OBA) and the back-link fibers between the OBAs . These custom-built fibers feature a combination of critical requirements that have qualified processes which have been flown individually but not combined within a single assembly: low insertion loss, high polarization extinction ratio and high optical damage threshold, all over the LISA thermal range. The Photonics Group at Goddard Space Flight Center completed an iterative fiber assembly production process, building upon 30 years of heritage of lessons learned at Goddard Space Flight Center, to create assemblies capable of meeting all requirements for the mission. The LS and OBA fiber assemblies produced by the GSFC Photonics Group have been tested up to 3W of optical power over the survival temperature range of the program and at high vacuum, which the process verified up to 8.4W under the same conditions.

The University of Florida precision interferometry group has developed an array of components and design principles to meet the numerous challenges presented when attempting to measure picometer level stability of optical and mechanical components for both ground- and space based instruments. Development of these techniques began with the design and testing of the central optical bench for the ALPS dark matter detector and were further expanded during the design of the test apparatus for the LISA mission telescope. We present a look at mechanical components, such as mountings, suspensions, and glass structures, that our group has developed to meet these requirements.

The Laser Interferometer Space Antenna (LISA) mission is a space based gravitational wave mission that will enable access to a rich array of astrophysical sources in the measurement band from 0.1 to 100 mHz. It complements ground based gravitational wave observatories, which typically search for signals at higher frequencies. Telescopes are one of the technology contributions from NASA to the European Space Agency (ESA). We will describe the key requirements for the flight telescopes and their development plan. We will also summarize the telescope technology development efforts.

We report on the successful sub milliradian parallelism testing of two optical periscopes for use in stability tests of the LISA structural and thermal model (STM) telescope. During the successful sub picometer stability testing of the STM it was found that tilt-to-length (TTL) couplings were the dominating noise source. The periscope was designed to level the input and output cavity beams as the TTL scaled with this difference in height. The parallelism was tested optically, and the mirror faces were measured with a coordinate measuring machine.

The NASA GSFC plans to test the optical pathlength stability (OPL) of the L3Harris built LISA Telescope Engineering Development Unit (EDU). To reduce TTL noise associated with the EDU off-axis design in said tests, University of Florida (UF) designed and built periscopes for integration within the EDU test setup. UF is also responsible for verifying that, within a similar setup as the EDU testing, the provided periscope’s OPL stability be below the EDU OPL stability requirement to verify utility in EDU OPL stability testing. My talk will detail the current status of the UF periscopes and periscope testing.

The ground-based gravitational-wave observatories LIGO, Virgo and KAGRA continue to improve. After the successful O4 observing run -- which detected over 200 likely binary merger events -- the observatories are offline and are focusing on instrumentation and facility upgrades. Besides sensitivity improvements, LIGO, Virgo and KAGRA are moving toward formally becoming a unified global organization, the International Gravitational Wave Observatory Network (IGWN). I will summarize the current status of the observatories, the current outlook for upcoming data collection runs, and the progress toward forming IGWN.

ODIN (Optomechanical Distributed Instrument for Inertial Sensing and Navigation) is a compact, low-mass optomechanical sensing system being developed under NASA’s InVEST program and scheduled to fly on the GRATTIS mission in 2027. Building on laser interferometry heritage from LISA Pathfinder and LISA, ODIN uses an array of optomechanical accelerometers with monolithic mechanical resonators and differential laser interferometers.

ODIN targets acceleration noise below 10⁻⁹ m/s²/√Hz and angular sensitivity below 50 μrad/√Hz with a low-SWaP architecture. These capabilities enable applications in Earth observation, planetary exploration, and navigation in GPS-denied environments. ODIN supports Earth’s mass change science objectives, aligned with GRACE-like mission, and can operate on resource-limited space platforms.

We will present the development of the ODIN flight system, its laser interferometry, laser transponder and phasemeters, as well as lab and space qualification results.

Many future space-based gravitational-wave missions are possible with distinct mission structures and technological challenges, and with correspondingly different science capabilities. As with other areas of observational astronomy, in moving beyond the current generation of missions, particularly LISA, there are many possible directions to take. Here we focus on improving a space-based GW mission concept's sky-localization ability. We present a summary of potential science opportunities, a range of possible mission concept approaches, and a set of simple tools for approximately assessing the performance of these missions in realizing the various possible science opportunities. Beyond the specific science and mission concept examples that we present, we hope that our framework may provide a foundation for the community process of prioritizing science opportunities and identifying promising mission concepts and technologies, for the next generation of space-based gravitational wave missions beyond LISA.

The origins and fate of massive compact objects such as black holes, neutron stars, and white dwarfs can only be fully answered by measuring deci-Hertz (dHz) gravitational wave emission. A plethora of innovative mission concepts have been proposed motivated by this need, promising incredible potential sensitivity and development cost. Like COBE rather than Plank, we prioritize time to first knowledge over ultimate sensitivity. We describe the highest priority science goals contained in and motivating a dHz observatory, the minimum sensitivity required to make these observations, and the fastest, least-costly mission that meets these minimum requirements. We dub this mission the Deci-hertz Interferometer Space Antenna (DISA).

A fast radio burst refers to the category of unknown bursts of radio-wave light emitted from outside of the galaxy. The origin of the fast radio bursts are generally unknown, and many researchers look into possible multi-messenger GRB or X-ray events that correspond to fast radio burst observations. However, gravitational wave multi-messenger searchers are generally unsuccessful with the current gravitational wave detectors, and most theorized progenitors would instead lie within LISA's bandwidth instead.

Multi-messenger astronomy is a promising area of astrophysics. This project examines white dwarf binaries that may be detectable through both electromagnetic and gravitational-wave observations. In preparation for LISA (Laser Interferometer Space Antenna), these systems were simulated using COSMIC (Compact Object Synthesis and Monte Carlo Investigation Code) to estimate observational expectations under a range of initial conditions. The resulting populations were mapped onto a 3D distribution of HI and H2 gas in the Milky Way to approximate dust extinction and calculate apparent magnitudes. These magnitudes were then compared with the sensitivities of telescopes such as the Vera Rubin Observatory to determine which systems would remain optically observable and could qualify as multi-messenger sources.

Binary neutron star mergers are prime multimessenger gravitational-wave events, and electromagnetic activity prior to coalescence can provide information complementary to the GW signal. I will present new three-dimensional force-free simulations of inspiraling binary neutron star magnetospheres that follow the late inspiral and show how the field structure, Poynting outflows, and precursor high-energy emission depend on the strengths and orientations of the stellar magnetic moments. We find that the outflows are strongly configuration dependent and highly anisotropic, with luminosity pulsations at twice the orbital frequency and a power-law dependence on orbital frequency whose index spans ~1-6. We also compute, for the first time in this framework, the electromagnetic forces and torques acting on the system, showing that sufficiently strong fields may induce significant crustal stresses and could affect the orbital dynamics. Using emission prescriptions motivated by isolated-pulsar studies, we construct precursor high-energy skymaps and light curves and find that, although curvature-radiation photons can in principle reach TeV-PeV energies in the final milliseconds before merger, magnetic attenuation makes the MeV band a more promising window for escaping precursor emission. These results highlight how pre-merger electromagnetic signals may complement gravitational-wave observations by helping constrain source geometry and viewing orientation, and by motivating searches for precursor emission and delayed echo-like rebrightenings in future multimessenger campaigns.

Multi-messenger detections of gravitational-wave (GW) and electromagnetic (EM) signals associated with massive black hole binaries (MBHBs) will provide an unprecedented understanding of the evolution of massive black holes and galaxies. They will depend on our ability to localize the merging binary within a particular host galaxy and on our understanding of the EM signatures of both. Motivated by this, we developed a model for spectral energy distributions (SEDs) from accreting MBHBs across three distinct accretion regimes: low, intermediate, and super-Eddington. Our model accounts for the unique geometry of MBHB systems, including circumbinary and individual minidisks. We explore how different binary configurations affect their spectral properties and investigate conditions under which accreting MBHBs can outshine their hosts. This model serves as a tool for the multi-messenger community to guide observations and interpret the environments of the LISA sources.

The Laser Interferometer Space Antenna (LISA) mission will detect mergers of massive black hole binary (MBHB) systems at the centers of relatively low-mass galaxies for the first time, but identifying their host galaxies within the large gravitational-wave localization volume presents a major challenge. Since these lower-mass host galaxies are likelier to be gas-rich, we investigate whether the stellar populations of MBHB merger host galaxies detectable by LISA contain signatures of a recent starburst, which may be expected if the MBHB resulted from a major galaxy merger. We use the Romulus25 cosmological simulation to select samples of MBHB mergers, and produce synthetic optical spectra of their host galaxies through stellar population synthesis and dust radiative transfer. We find that the MBHB merger host galaxies often display a starburst 2 to 3 Gyrs before the MBHB merger, and systematically low star-formation rates at the time of the MBHB merger, in contrast to a mass- and redshift-matched control sample of simulated galaxies without recent MBHB mergers. Our results suggest that observationally, rest-frame optical spectra of galaxies within LISA localization volumes can aid in identifying the exact MBHB host galaxy, by probing the rapidly-declining star-formation rate over the past ∼1 Gyr through various spectral features.

Supermassive black hole binary (SMBHB) systems are potential sources of electromagnetic (EM) radiation, with the EM signatures influenced by gas dynamics, orbital dynamics, and radiation processes. The gas dynamics are governed by general relativistic magnetohydrodynamics (GRMHD) in a time-dependent spacetime. Using a general relativistic ray tracing code we post-process numerical relativity simulations of these binary systems. We will present progress in calculating the electromagnetic signals of inspiraling and merging supermassive binary black hole binary systems.

LISA detections of gravitational waves from black hole binaries (BHBs) may provide sufficient merger-time predictions and sky localization to signal EM observatories to pre-point towards the system, offering an opportunity for multi-messenger astronomy. BHBs will cast a hierarchy of shadows on the image due to multiple lensing, the shapes and sizes of which should be able to constrain parameters of the system, such as the black holes' masses, spins, and separation distance. This work uses ray-tracing code BOTHROS on a BHB spacetime to measure the shapes and sizes of primary, secondary, and tertiary shadows, including Einstein rings, cast by the BHB on the image. Binary phase and inclination angle are sampled efficiently through a novel curvilinear adaptive mesh refinement technique with a braid group model that we call the Venus FlyTrap method (VFT). VFT allows us to quickly find and compute arbitrarily obscure BHB shadows in preparation for future explorations of other BHB systems.

As LISA will look for gravitational waves with frequencies in the approximate range from 0.001 to 0.1 Hz, it is useful to consider the electromagnetic counterpart of astrophysical sources with similar frequencies in their light curve. I will discuss the case of a gamma-ray burst with a strong quasi-periodic oscillation with a frequency less than 0.1 Hz, present possible physical interpretations and prospects for multimessenger studies.

The upcoming Laser Interferometer Space Antenna (LISA) mission will open a new window on millihertz gravitational-wave sources, including compact binaries and massive black hole systems. Many of these sources are expected to have electromagnetic counterparts, making coordinated ground-based observations essential. I present ongoing work at the Mount Lemmon Observing Facility (MLOF), in Mount Lemmon, Arizona using the University of Minnesota’s 60-inch (1.5 m) infrared-optimized Dall-Kirkham telescope to support preparatory, contemporaneous, and follow-up studies of future LISA detections. Recent upgrades have enabled remote, rapid-response observations and fast instrument switching, allowing the telescope to operate as a dedicated time-domain follow-up facility. In the preparatory phase, we identify and study candidate LISA sources, such as ultra-compact binaries and AM CVn systems through time-domain photometry and color selection. These observations refine orbital periods and variability properties that can later be compared directly with gravitational-wave measurements. Our current CCD system supports minute-scale cadence imaging suitable for monitoring compact binaries and accreting systems. While not optimized for very short exposures, the system enables stable, repeated light-curve measurements that track orbital modulation, eclipses, and accretion-driven variability. The potential addition of a fast-readout CMOS detector would further enhance high-cadence capabilities and improve sensitivity to rapid variability, strengthening the connection between electromagnetic light curves and LISA’s gravitational-wave phase evolution. In the contemporaneous and follow-up phases, moderate-resolution spectroscopy allows redshift measurements of host galaxies and classification of transient or accretion-driven counterparts. Together, these capabilities demonstrate how a flexible, mid-sized optical/NIR telescope can contribute meaningfully to the coordinated multi-messenger effort that will maximize LISA’s scientific return.

The General Coordinates Network (GCN, https://gcn.nasa.gov) is a public collaboration platform run by NASA for the astronomy research community to share alerts and rapid communications about high-energy, multimessenger, and transient phenomena. Over the past 33 years, GCN has helped enable many seminal advances by disseminating observations, quantitative near-term predictions, requests for follow-up observations, and observing plans. GCN distributes alerts between space- and ground-based observatories, physics experiments, and thousands of astronomers around the world. With new transient discovery and follow-up instruments, this coordination effort is more important and complex than ever. LISA alerts for inspiralling compact objects will present new challenges in the coordination of follow-up observations given the timescales. The existing GCN infrastructure is ideal for meeting the needs of the LISA mission and providing a conduit to our community of thousands of follow-up observers. I will present a summary of GCN’s capabilities and how we can aid in LISA alert format development, provide streams for simulated alerts, and will describe features in development that could serve the LISA community into the future.

Massive black hole binaries (MBHBs) formed during galaxy mergers are among the primary targets of the Laser Interferometer Space Antenna (LISA). Identifying their electromagnetic (EM) counterparts will be crucial for host-galaxy localization and multi-messenger science. I present a series of studies exploring EM signatures from MBHBs embedded in circumbinary disks. First, I investigate self-lensing flares, produced when one black hole gravitationally lenses emission from its companion, generating distinctive periodic features in the light curve. Second, I show that as a binary approaches merger, the accreting minidisks become tidally disrupted, producing a several-order-of-magnitude drop in thermal X-ray emission and leaving the system temporarily X-ray dark. Finally, I examine non-thermal emission powered by magnetic reconnection and show that while it does not obscure the thermal X-ray drop, it can produce a spectral reconfiguration in which the dominant high-energy emission shifts from thermal minidisk radiation to a harder non-thermal component near merger. Together, these results highlight promising observational pathways for identifying EM counterparts to LISA sources and probing the astrophysics of accreting massive black hole binaries.