Astrometric microlensing probes of the isolated neutron star population with Roman

TL;DR

This work investigates how the Roman Space Telescope's Galactic Bulge Time Domain Survey can detect and characterize isolated neutron stars via astrometric microlensing. Using PopSyCLE with four Maxwellian natal-kick distributions, it shows that NSs imprint a distinctive spur in the $(\log_{10} t_{\rm E}, \log_{10} \theta_{\rm E})$ space, enabling relatively clean NS candidate selection when both photometry and astrometry are available. The study predicts ~11,000 detectable microlensing events, including ~100 NS lenses, and demonstrates a Bayesian classifier that leverages high-signal observables to distinguish NS from other remnants, with performance dependent on kick magnitude. The results, including publicly released simulated datasets, provide a framework for NS mass and kick inferences from Roman data and highlight the importance of NS dynamics modelling and follow-up for robust population constraints.

Abstract

Notoriously hard to detect and study, isolated neutron stars (NS) could provide valuable answers to fundamental questions about stellar evolution and explosion physics. With the upcoming Roman Space Telescope, scheduled for launch in 2026, a new and powerful channel for their detection - astrometric microlensing - will become available. We set out to create a realistic sample of simulated gravitational microlensing events as observed by Roman with the Galactic Bulge Time Domain Survey. We focus in particular on the population of NS lenses, which has until now been largely understudied. We use state-of-the-art Galactic models tailored for application to microlensing by compact objects. We simulate four different NS populations with Maxwellian natal kick distributions: $\bar{v} = (150, \ 250, \ 350, \ 450)$ km/s. We apply projected Roman precision, cadence, and detectability criteria. We find the parameter space $\log_{10} t_{\rm E}$ - $\log_{10} θ_{\rm E}$, which will be accessible to Roman observations, to be maximally efficient for classification of stellar remnants. We find a feature in this space that is characteristic to NS; using this feature, optimal samples of NS candidates can be constructed from Roman-like datasets. We describe the dependence of observable parameter distributions on the assumed mean kick velocities. As the effects of natal kicks are very complex and mutually counteracting, we suggest more detailed studies focused on the dynamics of NS are needed in anticipation of Roman and future surveys. We estimate Roman will observe approximately $11\,000$ microlensing events - including $\sim100$ with NS lenses - whose both photometric and astrometric signal are detectable; the event yield decreases by $38\%$ if gap-filling low-cadence observations are not included. We make all simulated microlensing event datasets publicly available in preparation for Roman data.

Figures (8)

• Figure 1: Layout of the simulation fields, overplotted on the Roman GBTDS footprint as recommended by the Roman Observations Time Allocation Committee. Colours represent the total number of simulated microlensing events per simulation field over the average for all fields in the run with the PopSyCLELam2020 default 350 km/s mean NS kick velocity (before applying the detectability cuts).
• Figure 2: Yield of events with NS lenses detectable with Roman in both photometry and astrometry over the entire GBTDS duration as a function of mean kick velocity $v_{\rm kick}$, before (green, dotted) and after (purple, solid) applying the effect of natal kicks on volume density of NS in the Galaxy. Errorbars represent Poisson noise; the $\sigma^2 = \sum_i w_i^2$ formula Barlow1987 is applied to weighted Poisson counts. While larger kicks increase the probability of lensing by a single NS, this effect is counterbalanced in weighting by a lower amount of NS available as lenses in the inner Galactic regions.
• Figure 3: Distributions of NS (purple) and non-NS (grey) events in $\log_{10} t_{\rm E}$--$\log_{10}\theta_{\rm E}$ space. All simulated events regardless of passing detectability cuts are plotted; shade denotes iso-density levels estimated with a Gaussian KDE. Black circles highlight the NS events that passed the detectability cuts. The NS distribution exhibits a characteristic 'spur' feature that becomes stronger and more shifted leftwards with increasing $v_{\rm kick}$, yielding more detectable NS outlying from the main distribution.
• Figure 4: All six subplots present the same $\log_{10} t_{\rm E}$--$\log_{10}\theta_{\rm E}$ space distribution of detectable NS lenses from the $v_{\rm kick}$ = 450 km/s simulation run (coloured) compared to that of lenses from other (Star, WD, BH) classes (light grey). The NS lenses are coloured by event parameters: lens velocity $v_L$, lens mass $M_L$, source distance $D_L$, lens distance $\mu_L$, and a measure of background area subject to light deflection by the lens per unit time $\theta_{\rm E} \cdot\mu_{L}$, respectively (from top left row-wise). NS located in the 'spur' region distinguishable from other lens classes are nearby, have high proper motions, and are relatively very likely to cause lensing events.
• Figure 5: Like Fig. \ref{['fig:ns_spur_coloured']}, but coloured by parameters relevant to correcting for changed NS density: weight and Galactocentric $z$ coordinate. In the top subplot, black outlines denote events we select as 'in spur' to estimate the total yield of $\approx$6 expected 'spur' NS lenses in GBTDS, assuming $v_{\rm kick} = 450$ km/s.