Program

In this talk I will give an overview of the South African Astronomical Observatory and the work we do, from astrophysical research, telescope and data provision and the development of new instruments and technologies. Our human capital development and science engagement programmes are important aspects of our aim to explore the Universe for the benefit of society. I will touch on the Intelligent Observatory project to network telescopes and use AI technology to enhance operations, and the potential to expand this network to other African observatories.

We report the status of the NIR microlensing exoplanet search project, the Prime-focus Infrared Microlensing Experiment (PRIME). We have built a new 1.8m wide field infrared telescope at the Sutherland in South Africa.One of the largest NIR camera has been manufactured by using four H4RG-10 detectors loaned from the Roman project. Thanks to 1.45 deg.^2 FOV, we can conduct the first high cadence microlensing survey in H-band towards the central region of the galactic bulge, where high dust extinction prevents optical observations. Because the stellar density is higher at the lower galactic latitude, we expect higher event rate. We can compare the planet abundances in high and low stellar density for the first time, which is important for the study of the planetary formation scenarios. If the PRIME telescope and Roman observe the same fields simultaneously, different light curves will be observed due to the different line of sights, so-called the space-based microlensing parallax. This enables us to measure the mass and the distance of the lens system and enhance the Roman’s yields. We started observation in 2023 and continued in 2024. The telescope was used for the ToO observations for various transients.

I will discuss the design principles adopted for the PRIME building and review the construction phase and installation of the dome, both of which faced challenges.

We designed, built and commissioned the PRIME infrared camera in October 2022. It is being in continuous operation since then for more than two years. We present the camera design, performance, the photometric pipeline developed for it and our experience the camera use for the followups of the Gamma Ray Bursts.

I would like to introduce the SAND spectrometer for the PRIME telescope, and its synergy with other telescoeps, especially, IRSF and Subaru telescopes. SAND (South Africa Near-infrared Doppler instrument) is a time-stable high-dispersion spectrograph, covering the z- and Y-bands simultaneously (849 – 1085 nm) with the maximum spectral resolution of ∼60,000. Its main science is to monitor the radial velocity of M-dwarfs with the precision of a few m/s level, which enables us to search for exoplanets in their habitable zone as well as statistical investigation of young planets and stars. SAND is based on our experince of the first IR-only Dopper instrument for the Subaru telescope, IRD. Thus, SAND is a fiber-fed spectrograph, and we can easily change telescope used to collect the starlight by switching the fiber connection. It will be operated mainly with two telescopes (PRIME and IRSF),
.

The PRIME (PRime-focus Infrared Microlensing Experiment) telescope, installed at the South African Astronomical Observatory (SAAO), leverages its wide field of view (approximately 1.45 deg²) to conduct near-infrared microlensing surveys in the Galactic center region. It achieves near-infrared observations and a wide field of view by employing four HAWAII-4RG-10 (H4RG-10) detectors provided by NASA/GSFC.

Accurate photometric observations require correcting for the nonlinearity of the detector system. The conventional method, such as that used in the JWST pipeline (JWST-STScI-005167, SM-12), corrects for nonlinearity by using a stable light source to measure how a linear signal is transformed into a nonlinear signal. This method relies on the assumption that dark current is negligibly small relative to the light source or entirely linear.

In the case of PRIME, however, the detectors are operated at a high temperature (119.5 K as of December 2024), leading to an increase in dark current and the emergence of pixels exhibiting atypical (nonlinear) dark current behavior. For such pixels, the assumptions underlying the conventional method break down, making it impossible to apply accurate corrections. To address this issue, a new method was developed to correct nonlinearity even in the presence of such challenges, and its practicality has been validated.

This presentation will explain the new correction method itself, and its accuracy compared to the conventional method. It will also introduce how this new method has been integrated into the current image reduction process for PRIME.

With relatively few telescopes operating in the near infrared (NIR), and even fewer available for Target of Opportunity observations of transient sources, PRIME is a powerful facility for NIR transient detection. The PRime-focus Infrared Microlensing Experiments (PRIME) camera is part of the joint NASA-JAXA project supporting the Nancy Grace Roman Space Telescope engineering and science studies. It is installed on the 1.8 m PRIME telescope with a ~1.5 square degree FOV (0.5 arcseconds/pixel) dedicated to the project. The instrument is equipped with multiple broad band and narrow band filters between 0.9 µm to 1.8 µm. The facility is located at the South African Astronomical Observatory and has been in continuous operation since October 2022. After 2 years of operation, we present the results of the high energy transient team’s efforts in following up detections made by Swift, Fermi, Ligo Virgo KAGRA, Einstein Probe and other X-ray facilities along with the performance of PRIME in observing these targets.

Mira variable stars, evolved from stars aged 100 Myr and 10 Gyr, are useful tracers of stellar populations. There have large efforts to search for Miras and other variables towards the Galactic bulge, including those made in South Africa, but the collection of known Miras is still highly incomplete. The PRIME bulge survey is crucial to fill the gap of the surveys because its observations in the multi bands including Y, J, and H allows to detect Miras over a large range of interstellar reddening. I will present expectations for the PRIME survey and discuss related future studies like spectroscopic follow-up and the JASMINE space astrometry mission in the future.

In 2024 PRIME conducted its first full season survey of the Galactic Bulge. A transient detection system was implemented based on the difference imaging technique. This revealed ~200 microlensing and other transient events during 2024. I will present these results and discuss the prospects for real-time microlensing alerts for 2025.

The inner regions of planetary systems have been well-characterized by transits and radial velocities, revealing diverse architectures that are often much different than the Solar System. The outer regions are comparatively less understood, but microlensing is gradually allowing a clearer picture to emerge. This talk will review topics relating to the demographics and dynamics of exoplanetary systems, considering both inner and outer system architectures. I will discuss progress towards understanding the prevalence of Solar System-like architectures.

I present a somewhat preliminary statistical analysis of mass measurements and upper limits from high angular resolution follow-up images and microlensing parallax measurements for the 30 exoplanets in the Suzuki et al. (2016) statistical sample. These results indicate that larger mass host stars are more likely to host planets of a given mass ratio, q, in the range -4.3 < log(q) < -1.5. They also suggest that a significant fraction of wide orbit exoplanets may be hosted by white dwarfs

In this talk, I will introduce the motivation of the KMTNet AnomalyFinder pipeline and review its performance on the 7-year KMTNet data. Then, I will introduce the KMTNet Mass-ratio Function derived from the AnomalyFinder planetary sample, which is the largest statistical sample of microlensing planets to date, and I will compare it with other statistical samples.

Planet mass-ratio function is one of the important result from statistical analysis of the large microlensing survey data as it can be compared to the planet population synthesis models to constrain some parameters on the planet formation. It is also important that planet-to-host mass-ratio can be precisely measured in each planetary event. We derive the planet mass-ratio function by using the combined sample consisting of Suzuki et al. 2016 (MOA, PLANET, micro-FUN) and recent KMT survey data, which includes 84 planets in total. We will show the resultant mass-ratio function and compare it with the previous results and theoretical predictions.

OGLE-2017-BLG-0114 is a microlensing event with an anomaly caused by a planet on either a wide (s ≈ 3) or close (s ≈ 1/3) orbit. This event is truly exceptional. It is the only known planetary event where the light-curve shape, even after removing the anomaly, cannot be explained without accounting for the orbital motion of the source, known as the xallarap effect. This makes the event an ideal test case for addressing this effect in microlensing modeling. In my presentation, I will outline the current state of the analysis of this unusual even and discuss model fitting strategy.

Discovering new exoplanets via gravitational microlensing is a primary objective of NASA’s Roman Space Telescope, scheduled for launch no later than May 2027. To ensure that the Roman Mission Science Operations System pipeline meets the Level-1 science requirements for precise mass measurements of exoplanets detected via gravitational microlensing, it is essential to assess how higher-order effects, such as orbital motion, influence the derivation of key parameters in microlensing models. Neglecting orbital motion in modeling has been identified as a potential source of error in parallax estimates—parameters critical for determining lens mass in microlensing events. To quantify the impact of orbital motion on mass measurements, the Microlensing Modeling Group of the Roman Galactic Exoplanet Survey Project Infrastructure Team (RGES-PIT) is modeling 10,000 simulated microlensing light curves generated by the RGES-PIT’s Simulation Team, both with and without orbital motion. This presentation will share the results of our investigation, emphasizing how the inclusion or omission of orbital motion during the modeling affects the recovery of parallax parameters and, consequently, mass measurements.

MOA-2010-BLG-328 is a binary-lens binary-source planetary event with a unique light-curve model. The original analysis by Furusawa et al. 2013 reports a degenerate light curve, where several solutions are possible depending on which higher order effects are considered. Ultimately, they considered that the “parallax + orbital motion” and the “xallarap” models provided the best fits.

Eight years later, we reobserved this event with Keck and Hubble in an attempt to directly measure the lens’ flux and relative proper motion to the source. With this high-angular resolution data acting as constraints, we can remodel the event and determine which model, and therefore, which higher-order effects provide the best fits. We found it necessary to model parallax, orbital motion, binary source effects and xallarap together. Even with a lens detection, the solution remains unclear due to the degeneracy between the microlensing parallax and orbital motion, which cannot be broken with NIR data alone for M dwarfs in the disk.

Therefore, we conclude that the system is either a ∼ 0.2 Solar mass star at 2 − 3 kpc, or a ∼ 0.5 Solar mass star at 4 – 5 kpc. This study highlights the importance of obtaining data in multiple pass-bands, in addition to considering all the higher order effects that could be present on the light curve.

We report on the analysis of the microlensing event MOA-2020-BLG-108. This event was discovered in June 2020 at Galactic coordinates (l, b) = (3.228,−1.488) by the MOA collaboration. We fitted the light curve with the standard single-source single-lens (1S1L) and the single-source binary-lens (1S2L) models. The results show that the 1S2L model is Δchi^2=4440 and ΔBIC=4380 better than the 1S1L model. Furthermore, five solutions of the 1S2L model were found to reproduce the MOA-2020-BLG-108 light curve. Higher-order effects such as parallax and lens orbital motion were added to the analysis, but other than the finite source effect, these effects were not significantly detected. In addition, we conducted a Bayesian analysis to estimate the parameters of the lens system. The result was that the five solutions were consistent in that the host was a late-type star of M_L,H ~ 0.6 M_Sun and the companion was a giant planet beyond the snow line.

Current generation high-cadence microlensing survey projects such as KMTNet are rapidly increasing the number of exoplanet candidates discovered with microlensing, with fainter magnitude limits than previous searches. However, the faintness of the source stars involved in these events makes it hard for high-cadence surveys to measure both microlensing parallax, π_E, and source colors. This difficulty in turn prevents estimation of the angular Einstein radius, θ_E, and lens mass. To help resolve this issue, we use the Dark Energy Camera (DECam) at Cerro-Tololo Interamerican Observatory to perform a low-cadence survey of a portion of the KMTNet survey footprint, obtaining either g & i or r & z photometry, depending on the extinction of the field. We then match our sources with OGLE/KMTNet event candidates. Here we present the early results of our work to analyze these source colors, measure parallaxes, and ultimately arrive at lens mass measurements.

Microlensing can reveal populations of dim compact objects that are otherwise very hard to find. All-sky surveys depending on their design have the potential to look for these compact objects throughout the sky and help us understand the rate at which these events are expected to happen. the Transiting Exoplanet Survey Satellite (TESS) primarily focuses on finding transiting exoplanets, and we aim at using its comprehensive all-sky survey and high cadence to look for microlensing candidates. We are using traditional detection algorithms used by the community along with innovative machine learning algorithms trained and tested on simulated TESS microlensing light curves and TESS-SPOC Full Frame Image (FFI) light curves. This project is important from various perspectives; the simulations provide an understanding of what we can expect from TESS in microlensing, and our algorithmic approach tested on TESS light curves will provide an opportunity to evaluate various detection methods available to the community and for future all-sky surveys.

None

None

The Roman Galactic Bulge Time Domain Survey will observe the Galactic Bulge with exquisite precision, allowing the detection of thousand of cold planets via microlensing. With more than 100 millions sources monitored at <15 min cadence, the volume of data generated is challenging. I will present a summary of the pipelines implemented by the Microlensing Science Operations System (MSOS, developed by the Science Support Center at IPAC) to achieve the Roman mission requirements. I will first detail the role of MSOS inside the global Roman environment. I will then develop on the different algorithms deployed to achieve precise photometry in crowded fields, the modeling of the lightcurves, the assessment of the physical parameters of the lenses as well as the overall pipelines efficiencies. Finally, I will present the data products that MSOS will deliver to the community.

The constraints on the mass and distance of the lens inferred from the microlensing light curve can be significantly improved with a well-informed prior on the lens flux contribution to the total flux. These constraints can be further refined by including measurements of the lens’s proper motion relative to the source. In this study, we aim to estimate the lens flux contribution and relative proper motion using a 2D image of the source-lens system captured at a much later epoch, well after the microlensing event, by the Roman Space Telescope. This late observation provide a maximally separated point spread functions (PSFs), corresponding to the lens and the source, due to their relative motion and hence maximizes the detectability of the lens in the image. We derive analytical estimates of the uncertainties in the lens flux and relative proper motion measurements based on 2D image observables, such as the centroid, spread, and skewness of the combined PSFs, assuming a Gaussian approximation. Additionally, we exploit the color-dependent centroid shift observable through multi-band photometry of Roman and demonstrate how it can further constrain lens properties. We discuss how our error estimates can be applied to assess the likelihood of accurately measuring the mass and distance of the host lens in planetary microlensing events, thereby resolving the degeneracy in determining the planet’s mass and its separation from the host star. Furthermore, our framework can set upper limits on the flux of dark lenses, contributing to the discovery and characterization of a population of dark compact objects in the Galaxy.

As of today, over two thousand exoplanets have been discovered within multiple-planetary systems, though only a little more than ten have been detected through microlensing. Using the recently released VBMicrolensing code, an in-depth study was conducted on triple-lens systems to assess the Roman Space Telescope’s observational capabilities in these scenarios.
VBMicrolensing is a code designed to calculate magnification in microlensing events, not just for single or binary systems, but also for more complex configurations such as triple-lens systems or those with more than three lenses. I will present the various aspects of the code in details as well as its effectiveness.
I will also discuss preliminary results on the capabilities of Roman to detect multi-planetary systems, by simulating a large sample of triple lens and various configurations. These early results show that Roman should be capable to detect triple lens planets in 20% of the cases where the source passes close to the central caustic of the systems.

Microlensing surveys such as KMTNet, OGLE, and MOA have provided extensive datasets of microlensing events, many of which overlap with the observational fields of future Euclid and Roman surveys. In this study, we perform a detailed statistical analysis to derive the distributions of proper motions of lenses relative to the sources for each event based on the exitsting information and suitable Galactic models. Projected to the times of future surveys, such predictions allow us to estimate the number of microlensing events for which the detection of the lens is in principle possible, with a particular attention on published planetary events. Our findings offer valuable insight into the capabilities of upcoming space-based observatories and their role in advancing our understanding of microlensing phenomena and their astrophysical significance.

Core-Accretion and gravitational instability models are two leading scenarios for planet formation. They have distinguishable implications on the mass-period distribution of long-period, cold planets. I will discuss how micro-lensing, direct imaging, astrometric, and time-domain surveys may provide constraints on the likelihood of these planet-formation channels. A
closely related issue is the possibility of detachment of cold planets from their host stars. I will discuss various physical mechanisms which may have contributed to the launch of freely floating planets and some potentially
observable signatures.

The Rubin Observatory has begun on-sky testing with the commissioning camera, and will soon be installing LSSTCam and preparing to begin the LSST survey. I will describe the image processing methods Rubin uses to support time domain science and the resulting products that will be available for users, along with our ongoing testing to ensure that the pipelines perform well in the dense fields that are important targets for microlensing studies.

Rubin Observatory is a groundbreaking new wide-field survey facility currently being commissioned in Chile. It is due to begin science operations for the Legacy Survey of Space and Time (LSST) in October. Rubin pioneered a novel community-driven approach to developing an observing strategy for LSST that will serve as many science goals as possible, and the Phase 3 recommendations for this strategy was released in Oct 2024. This represents the survey strategy that will be used for at least the early years of the survey, though it will continue to be reviewed throughout operations.
I will describe the survey strategy and examine its implications for the detection of microlensing events, and potential synergies with contemporaneous surveys such as Roman.

Vera C. Rubin Observatory is an 8.4~m telescope in Chile and will begin its scientific operations by the end of 2025. It has a 9.6 square degree field of view and will continuously observe a wide area of the southern sky, including the Galactic plane. Once it reaches its full capabilities, it will produce 40 million alerts per night, and during the 10 years of the Legacy Survey of Space and Time (LSST) is expected to discover thousands of microlensing events. However, the average cadence is likely to be too low to characterize planetary and stellar binary events, and follow-up is going to be necessary to recover the underlying populations. Additionally, LSST’s cadence, depth and volume will put our current methods of detecting microlensing events to the test and will likely call for new strategies in handling and quickly assessing large volumes of data.

In this talk, I would like to summarize lessons we learned from running ground-based follow-up campaigns over the last years for alerts from lower-cadence wide-area surveys as a pathfinder to LSST follow-up.
We have followed both stellar/planetary events, as well as long-duration, potential stellar remnant, events. The former is more likely to be detected by LSST.
I will present which methods have worked, and which must be adapted for LSST. I will discuss the classification of LSST alerts and our selection for future follow-up.

I will first review the current status of the Earth Two mission, in particular its microlensing survey, and then discuss potential synergies with other ground and space facilities.

Earth 2.0 (ET) Mission consists of six transit telescopes and one microlensing telescope to be launched to Earth-Sun L2 orbit in 2028. The microlensing telescope aims to search free-floating planets and measure their masses through satellite parallax from simultaneous observation from the ground. The diffraction limit designs and undersampled images challenge the data processing pipeline. We develop a preliminary pipeline and an image-level simulator to understand the noise sources and estimate the yields. I will present the pipeline design, the simulation setup, and the effects of some investigated noise sources for the ET microlensing project.

Euclid will observe in March 2025 a ~6.5 square degrees are in the Galactic Bulge with its Visible Instrument (VIS) providing a reference dataset for future microlensing studies in the area. We summarize the survey requirements, the technical constraints and limitations and present the implementation plan. The footprint of the survey, planned data products, and the currently consolidated elements of the timeline of implementation and foreseeable data releases will be also presented.

The ESA Euclid mission will spend ~45 hours at the end of March 2025 observing the Roman fields. These precursor observations will take place almost 2 years prior to the Roman’s first observational season. While this is very exciting, it is also occurring at a significantly less time span than the one used in the Bachelet+ 2022 simulations. In addition, Euclid’s very high magnitude sensitivity (down to VIS=25+ mag) means that the observed fields will be extremely crowded, much more crowded than most fields the microlensing community has ever had to work on to date. Combining Euclid and Roman is a very powerful way of mass constraints by measuring relative source lens proper motion. However, for most cases, it will be a fraction of the PSF, in very crowded fields. So we will need to evaluate which photometric approaches would be the best to use.

Here I will present a new tool to study the upcoming Euclid observations and fit the potential microlensing targets included in the frames. I will explain all the numerous methods that can be used for sky background fitting, aperture photometry, star identification and modelling. Then, I will demonstrate how I am using the method presented in Vandorou+ 2020 and Blackman+ 2021 to fit the source and lens parameters (separation, flux ratio and direction). Finally, I will illustrate some representative examples of the simulated events I have worked on and I will discuss the accuracy distribution we should expect for source-lens separations that are inside one-pixel.

The Nancy Grace Roman Space Telescope (Roman) is NASA’s next major astrophysics mission, scheduled for launch in late 2026. Roman will feature a wavelength range, aperture, and angular resolution comparable to the Hubble Space Telescope but with approximately 100 times the field of view and 1,000 times the sky-mapping speed. This capability allows it to survey large sky areas rapidly or repeatedly observe smaller areas with high frequency. A key community survey during Roman’s primary mission will be the Roman Galactic Bulge Time Domain Survey (RGBTDS). This survey will monitor about 2 square degrees of the Galactic center with a cadence of approximately 15 minutes using a wide 1–2 micron filter, spanning six seasons of 60–72 days each, for a total survey duration of 380–440 days. The RGBTDS aims to detect around 1,500 cold bound planets and hundreds of free-floating planets using microlensing, as well as about 100,000 hot and warm transiting planets. Roman’s transit and microlensing data will enable the first comprehensive statistical survey of exoplanets within a single stellar population, covering planets with radii or masses greater than twice that of Earth across all orbital distances. I will review Roman’s anticipated exoplanet science output and discuss the efforts of the RGES Project Infrastructure Team (RGES-PIT), which ensures the success of primary RGES scientific microlensing objectives. On behalf of RGES-PIT, I will outline the team’s activities over the next four years and highlight opportunities for involvement from the microlensing, exoplanet, and broader astronomical communities.

As we prepare for the flood of data from the upcoming Vera C. Rubin Observatory and Nancy Grace Roman Telescope, it is important to understand how fitting the microlensing lightcurves impacts our physical interpretation of them. In particular, we want to understand the results of fitting millions of lightcurves without being able to treat them individually. We simulated sets of lightcurves using PopSyCLE which are representative of previous ground based surveys and upcoming Roman telescope. We explore parameter bias when fitting events including noise and gaps; fitting PSPL models to binary events; and the effect of using models with and without parallax, including how parallax priors affect the fits. We assess the impact both on the individual event level and on the interpretation of the population.

We know of nearly six thousand exoplanets today, and almost all of them orbit single stars. The discovery of these planets has taught us a lot about planet formation and the demographics of planets around single stars. It is estimated that more than half of all the stars in our galaxy exist in binary systems [1][2]. Yet, we know of only a handful of circumbinary planets. Most of these planets were discovered by Kepler, and they orbit eclipsing binary star systems very close to the stability limit [3]. Microlensing is a promising avenue for discovering circumbinary planets over a broader range of parameter space. While ground-based microlensing surveys have only discovered ~ 8-10 circumbinary planets so far [4], Roman’s higher cadence and continuous light curve coverage might allow us to find many more of these systems, thus allowing us to answer questions about formation mechanisms and occurrence rates of planets in binary star systems. In this talk, I will outline our efforts to characterize the detectability of circumbinary systems with different architectures in the Roman microlensing survey. This work will not only give us an estimate of yields of various kinds of circumbinary planets from Roman but will also help inform future searches for circumbinary events in Roman data.







References:



[1] Offner S. S. R., et al., 2023, doi:10.48550/arXiv.2203.10066

[2] Raghavan, et al., 2010, ApJS, 190, 1

[3] Doyle, L. R., Carter, J. A., Fabrycky, D. C., et al. 2011

[4] Kuang et al., 2023, doi:10.1093/mnras/stad461

The Roman Galactic Bulge Time Domain Survey will be the first survey able to access large numbers of both hot and cool exoplanets, and will do so across Galactic distance scales. With around 1500 microlensing detections of cool exoplanets, and up to 200,000 transit detections of hot exoplanets, Roman will allow a reliable and extensive exploration of exoplanet demography and its relation to Galactic environment. I’ll discuss how the combination of transit and microlensing samples from Roman will be particularly powerful for statistical constraints on exoplanet demography and planet formation theory. This tantalising prospect motivates a new approach to exoplanet demography that is able to maximally combine the information from both microlensing and transit datasets within the Galactic context. I will present a framework being developed by Manchester PhD student, Akshay Priyadarshi, and myself that is based on an extension of synthetic stellar population synthesis modelling that is now routinely used for modelling microlensing populations. This approach enables the very different selection biases of microlensing and transits to be treated consistently, using the same underlying Galactic model. It allows improvements in exoplanet demographic understanding to come not just through enlarged exoplanet datasets but also through improvements in our knowledge of Galactic structure. I will report on initial trials of this approach based on the Kepler DR25 transit sample and I’ll discuss how we will aim to integrate this with the Roman microlensing sample, and potentially with other survey datasets.

Free-floating planets (FFPs), especially the low-mass ones, are expected to be common in our Galaxy. As a powerful microlensing survey, KMTNet has not detected free-floating planet events with low theta_E (<1 uas). The possible reason is that the photometry pipeline is not sensitive to the single short-timescale signals. For searching these low-mass free-floating planets, we are setting up a new full-frame photometry pipeline in KMTNet historical images. The pipeline includes the register from frames to a reference frame, stamp cutting and preprocessing, image subtraction, and photometry based on the pySIS package. We tested the pipeline on the bright star on one square degree images in 2018, and I will introduce the recent results, including the false positive rate, new microlensing event candidates, and the future perspective.

The Roman Space Telescope’s Galactic Bulge Time Domain Survey will provide a new window into the demographics of free-floating planets, planets unbound to any star. The mass and velocity distributions of this population hold key insights into their formation mechanisms and the growth of planetary systems in a short-lived period of early dynamical instability that is otherwise difficult to probe observationally. In this talk, I will show the results of the first ab initio simulations of free-floating planet formation and connect the results to particular features in the free-floating planet mass function that Roman will be able to observe.

Gravitational microlensing provides a unique method to detect distant and intrinsically faint objects that are challenging to detect otherwise, such as free-floating planets (FFPs). The distribution of FFPs remains elusive, with widely varying estimates reported from recent ground-based studies (Mroz et al. 2017 , Sumi et al. 2023). The K2 “second-light” mission of the Kepler Space Telescope featured a dedicated microlensing study towards the Galactic bulge during Campaign 9 (K2C9). This data was blindly searched by McDonald et al. 2021 who reported on 5 new event candidates. In our study, we fully automate the McDonald et al. (2021) pipeline selection and report a new event candidate selection that we are using to recover Kepler’s FFP detection efficiency to estimate the galactic abundance of FFPs. This study provides good preparation for a future Roman survey for FFPs that will provide critical insights for precision tests of planet formation theories.

Free-floating planets (FFPs) may be the most abundant type of terrestrial-mass exoplanets, yet they are much more challenging to detect than their bound counterparts. The short-duration magnification of background stars by FFP-induced gravitational microlensing, our best tool for studying this elusive population, requires high-cadence, wide-field surveys to detect. The Transiting Exoplanet Survey Satellite (TESS), though designed to detect close-bound exoplanets via transits, has a cadence as short as 200 seconds and has monitored hundreds of millions of stars, providing a unique dataset for short-duration transient events.

A preliminary search of TESS light curves found one short-duration event of interest because its morphology matches expectations for a terrestrial-mass FFP. But the low expected microlensing rate, short observational baseline, and source magnetic activity challenge this interpretation. Therefore, we explore two false positive scenarios: 1) a stellar flare and 2) centrifugal breakout of entrapped matter in a strong magnetosphere. Both are unlikely but cannot be quantified. Therefore, we are unable to determine if the FFP or one of the false positive interpretations are quantitatively more likely. The event may constitute a first example of the new classes of pernicious false positive that future space-based microlensing efforts will encounter. By characterizing such events, our search across all TESS sectors will support upcoming studies of rogue worlds with dedicated space-based microlensing surveys.

We describe CLEoPATRA:Contemporaneous LEnsing Parallax and Automatic TRansient Assay – a space mission concept to obtain microlensing parallax measurements for free floating planets. The original concept is to observe the Roman Space Telescope FOR simultaneously during its extended mission to measure the slight shift in time and amplitude of peak brightening from a spaceborne observatory located in Earth orbit approximately 0.01 AU from Roman at the second Lagrangian point. We discuss calculations we conducted in support of this concept. We discuss the possibility of a Low Earth Orbit choice for the mission concept to facilitate the measurement of acceleration parallax of point mass microlensing signals to obtain masses of very low-mass free floating planets independently of any other observatory.

Interferometric observations of gravitational microlensing events offer an opportunity for precise, efficient, and direct mass and distance measurements of lensing objects, especially those of isolated neutron stars and black holes. However, such observations were previously possible for only a handful of extremely bright events. The recent development of a dual-field interferometer, GRAVITY Wide, has made it possible to reach out to significantly fainter objects, and increase the pool of microlensing events amenable to interferometric observations by two orders of magnitude.

In this talk, I will describe our efforts to use this new instrument to study microlensing events. In particular, I will present the first successful resolution of microlensed images with this new observing mode, and discuss the future prospects of using dual-field interferometric observations for the follow-up of microlensing events.

Although millions of isolated black holes are expected to reside in the Galaxy, only one detection has been confirmed so far. Studying a population of these objects would provide extremely valuable input for stellar evolution, supernova physics, and, by extension, gravitational wave studies. The only known channel through which they can be detected is gravitational microlensing.

The difficulty in finding microlensing black holes lies in having to choose a small subset of events based on characteristics of their lightcurves to allocate expensive and scarce follow-up resources that can confirm the identity of the lens. Current methods either rely on simple cuts in parameter space without using the full distribution information or are only effective on a small subset of events.

We have developed a new classifier that combines posterior constraints on microlensing parameters and a Galactic simulation to estimate the lens class probability. This method is fast, flexible, and relies only on information easily obtained from the lightcurve. It is also particularly efficient at differentiating black holes from other lens classes. Therefore, it is well suited to be applied as a filter for large databases or real-time data streams, picking out the optimal black hole candidates for allocating additional resources.

We have applied our classifier to ~10,000 microlensing events from the OGLE survey and found 23 new high-probability black hole candidates, which we analysed in detail. The classifier also suggests that follow-up of the only known isolated black hole was a high-risk strategy that paid off. While in this application we focused on point source – point lens photometric microlensing, the classification framework can be used with any set of microlensing parameters, as well as for other science cases. We make this framework available via the open-source Python package `popclass`.

Gravitational microlensing provides a powerful method for detecting and characterizing dark/hidden objects in the Milky Way, but non-bulge events often suffer from incomplete source information, hindering precise mass measurements. In this talk, I present an analysis of microlensing events identified by Gaia Science Alerts (GSA), focusing on non-bulge events. Leveraging photometric observations with galactic models and Gaia DR3 data, I refined event parameters to better constrain the lens masses. Among these events, four displayed binary lens features, three exhibited finite source effects, and eight had blend fluxes consistent with zero, suggesting the presence of dark lenses—potentially red dwarfs, white dwarfs, neutron stars, or black holes. These findings highlight the value of Gaia’s alerts for probing hidden compact objects across the Milky Way.

The Milky Way likely contains 100 million black holes; however, only a single isolated black hole and a few dozen binary black holes have ever been found. A more complete census of the Milky Way’s black hole population is critical to understanding how massive stars die, how black holes are formed, whether black holes receive kicks or retain their companions, and how black holes get close enough to merge via gravitational waves. The most promising method for mapping the demographics of black holes is gravitational microlensing. The explosion of wide-area, long-duration photometric surveys (ZTF, Rubin, LS4, Euclid, Roman, etc.) coupled with current and future astrometric capabilities (adaptive optics, HST, JWST, Roman, etc.) will likely increase the number of detected isolated and binary black holes by 100-fold or more. We will present the infrastructure needed to bring about this revolution in our understanding of black holes. This includes photometric pipelines to identify strong black hole lens candidates, organizational platforms for tracking and following-up microlensing events, astrometric follow-up instruments and analysis methods, modeling packages to jointly fit photometry and astrometry and extract black hole properties, and population simulations to infer demographic properties. We will present results of our current experiments to detect black holes with much smaller samples (1-2 candidates per year) and discuss lessons learned and predictions for the coming surveys.

In this talk, I will present a live demonstration of an open-source, self-directed learning resource, being developed for graduate students new to the field of gravitational microlensing. This is primarily a collection of Jupyter notebooks, which offer an interactive, choose-your-own-adventure approach to learning about key microlensing concepts. The project includes Python-based code examples, exercises with model solutions, and in-depth explorations of topics such as difference imaging, light curve fitting, galactic models, and survey simulations. By working through these notebooks, students can develop practical coding skills while strengthening their intuition for microlensing phenomena. As an ongoing open-source project hosted on GitHub, we encourage community collaboration and contributions.

With the about 3000/yr microlensing events discovered by current ground surveys and more by Roman and ET missions, fast and stable codes are required to fit the binary and multiple lens events. The advanced contour integration code VBBinaryLensing is widely recognized in microlensing field for its efficiency and flexibility, and the recently published VBMicroLensing extends these benefits to multiple lensing. In this talk, I will introduce the acceleration of VBBinaryLensing and VBMicroLensing through compiling and algorithmic optimization. By inlining the complex arithmetic operations, we achieve ~3 times speedup of VBMicroLensing v4.0 up to magnification ~100, with the speedup decreases at higher magnification, which contributes to the release of v4.1. Furthermore, by replacing sorted linked list with skiplist data structure to store image contours, which reduces the algorithmic complexity of ordering images, another ~1.5 times speedup is achieved at magnification ~200 for both codes, with further increases in speedup at higher magnification.

The initial mass function (IMF) of the Galactic bulge can be obtained from the luminosity function measured by the Hubble Space Telescope if the statistics of binary stellar systems is well constrained. Such statistics can be estimated with gravitational microlensing, which have large occurrence rate in the bulge and is sensitive to masses of the lensing objects, including low-mass stars. We propose to search for and analyse binary-lens/point-source and single-lens/binary-source events in databases of large microlensing surveys allowing to explore the lower-mass end of the bulge IMF even in unresolved binary systems. We are implementing a fully-automated approach in the search and characterisation of new and previously identified events. With the proper consideration of the detection efficiency, such a large statistics of binaries will provide important constraints for the binary fraction and mass-ratio distribution. In this talk, I will present the methods and current status of this effort, including its performance for a selection of binary events with well-separated bumps. Some tools developed in this project may be valuable for future analysis of data from the Roman Space Telescope.

Understanding the distribution and properties of stellar remnants, brown dwarfs and exoplanets is essential for models of stellar evolution, planet formation and black hole formation. At the same time, these objects are still difficult to detect, not only by traditional methods, but also in microlensing events alerted by survey teams. Monitoring events outside the central bulge region remains a challenge, preventing more comprehensive studies of these populations and their properties. LCO’s geographically distributed network of telescopes is being used as part of the OMEGA key projects to provide high-cadence photometry and time-series spectroscopy of transient targets discovered by wide-area surveys such as Gaia and ZTF. The OMEGA project’s custom target and monitoring managers automatically prioritize monitoring requests in response to survey alerts. We provide updates on analysis of current and past events, project status, and future plans

The effect of wave optics on gravitational lensing produces a wavelength dependent magnification ($\mu(\lambda)$) which could, in principle, be measured from the spectrum of lensed sources. We refer to this phenomenon as “wavelensing”. The characteristic oscillation pattern of $\mu(\lambda)$ appears when $\lambda$ is comparable to the Schwarzschild radius of the lens, but is destroyed by the finite size of incoherent sources. In this work we discuss a few basic conditions that might enable the observation of wavelensing, addressing the effects of the full wave optics for uniform finite sources. We identify potential astrophysical sources, lenses, and wavelength ranges where the wavelensing signature might be observable using current spectrographs.

The modeling of multiple lens system is well defined problem but hard to solve. Past methods could not analyze more than three lens system. The analyses suffered by sever round-off error caused by caustic singularities. In near future, high-precision high-cadence observations from space are expected. To analyze such data, efficient high precision algorithm for multiple is particularly important. We are on the way of developing new algorithm. In the new algorithm, most of the problems in the past algorithms. In principle, there is no limitation on the number of lenses. Almost all of the singularities can be avoided. I will report the progress of the new algorithm.

We present measurements of the microlensing optical depth and event rate toward the Galactic bulge using the dataset from the 2006-2014 MOA-II survey, which covers $22$ bulge fields spanning ~42 deg^2 between -5 deg < l < 10 deg and -7 deg < b < -1 deg.
In the central region with |l|<5 deg, we estimate an optical depth of tau = [1.75\pm0.04]\times 10^{-6}\exp[{(0.34\pm0.02)(3-|b|)}] and an event rate of Gamma = 16.08\pm0.28]\times 10^{-6}\exp[{(0.44\pm0.02)(3-|b|)}] star^{-1} year^{-1} using a sample consisting of 3525 microlensing events, with Einstein radius crossing times of t_E < 760 days and source star magnitude of I_s<21.4 mag.
We confirm our results are consistent with the latest measurements from OGLE-IV 8 year dataset \citep{mro19}.
We find our result is inconsistent with a prediction based on Galactic models, especially in the central region with |b|<3 deg.
These results can be used to improve the Galactic bulge model, and more central regions can be further elucidated by future microlensing experiments, such as the PRime-focus Infrared Microlensing Experiment (PRIME) and Nancy Grace Roman Space Telescope.

The exceptional photometric accuracy and the continuous sampling obtained by the Roman Galactic Exoplanet Survey will allow the detection of very small anomalies and perturbations caused by additional bodies. This will pose severe challenges for microlensing codes, which must be prepared to account for these elusive features and report physically consistent models. New strategies for modeling complex microlensing events are therefore necessary in order to exploit the full potential of Roman. Here we report about the latest evolution of RTModel, a modeling platform based on the principle of template matching for the fitting of binary lensing events. New model categories have been added, templates are now customizable by the user, a new strategy for the search for offset degeneracy partners is included, external constraints coming from lens-flux analysis can be forced into the fit. With these innovations and further technical optimizations, the efficiency in the recovery of the correct planetary models of simulated events has been pushed to 95%. New routes to apply the philosophy of template matching to triple lenses and binary-source-binary-lens events will also be explored and incorporated in future versions.

A few inconsistencies in the foundational literature on how the periodic shift of the effective line of sight due to the Earth’s revolution around the Sun affects gravitational microlensing events have caused some issues for the proper interpretation of acquired data and the accurate reporting and reproducibility of results. I will point out how to avoid potential pitfalls.

The lensing wave effect produced by microlenses embedded in a lens galaxy is an unavoidable phenomenon in strongly lensed gravitational waves (SLGWs). These microlensing effects present both obstacles and opportunities, challenging the detection and enabling novel applications of SLGWs. However, analyzing these effects requires computing a complete diffraction integral over each microlens, a task that becomes immensely time-consuming given the sheer number of microlenses (from 10^3 to 10^6). In this talk, I will introduce a new adaptive mesh algorithm that significantly accelerates the computation of diffraction integrals, alongside a new method for identifying gravitational wave host galaxies based on microlensing-induced diffraction patterns.