arXiv subjects

Manan Agarwal

Publications and source records attributed to Manan Agarwal.

26 records · Page 2Linked to original sources

arXiv2023

Applications of Deep Learning to physics workflows

Modern large-scale physics experiments create datasets with sizes and streaming rates that can exceed those from industry leaders such as Google Cloud and Netflix. Fully processing these datasets requires both sufficient compute power and efficient workflows. Recent advances in Machine Learning (ML) and Artificial Intelligence (AI) can either improve or replace existing domain-specific algorithms to increase workflow efficiency. Not only can these algorithms improve the physics performance of current algorithms, but they can often be executed more quickly, especially when run on coprocessors such as GPUs or FPGAs. In the winter of 2023, MIT hosted the Accelerating Physics with ML at MIT workshop, which brought together researchers from gravitational-wave physics, multi-messenger astrophysics, and particle physics to discuss and share current efforts to integrate ML tools into their workflows. The following white paper highlights examples of algorithms and computing frameworks discussed during this workshop and summarizes the expected computing needs for the immediate future of the involved fields.

hep-ex↗

arXiv2022

A $Gaia$ EDR3 search for tidal tails in disintegrating open clusters

We carry out a search for tidal tails in a sample of open clusters with known relatively elongated morphology. We identify the member stars of these clusters from the precise astrometric and deep photometric data from $Gaia$ Early Data Release 3 using the robust membership determination algorithm, ML-MOC. We identify 46 open clusters having a stellar corona beyond the tidal radius, 20 of which exhibit extended tails aligned with the cluster orbit direction in galactocentric coordinates. Notably we find NGC 6940 (at a distance of $\sim1$ kpc) is the furthest open cluster exhibiting tidal tails that are $\sim50$ pc from its center, while also identifying $\sim40$ pc long tidal tails for the nearby Pleiades. Using the minimum spanning tree length for the most massive stars relative to all cluster members, we obtain the mass segregation ratio ($\rmλ_{MSR}$) profiles as a function of the number of massive stars in each cluster. From these profiles, we can classify the open clusters into four classes based on the degree of mass segregation experienced by the clusters. We find that clusters in the most mass segregated classes are the oldest on average and have the flattest mass function slope. Of the 46 open clusters studied in this work, 41 exhibit some degree of mass segregation. Furthermore, we estimate the initial masses (M$\rm_{i}$) of these open clusters finding that some of them, having M$\rm_{i}\gtrsim 10^{4} M_{\odot}$, could be the dissolving remnants of Young Massive Clusters.

astro-ph.GA↗

arXiv2022

UOCS-VIII. UV Study of the open cluster NGC 2506 using ASTROSAT

We study an intermediate-age open cluster NGC 2506 using the \textit{ASTROSAT}/UVIT data and other archival data. We identified 2175 cluster members using a machine learning-based algorithm, ML--MOC, on Gaia EDR3 data. Among the cluster members detected in UVIT filters, F148W, F154W, and F169M, we detect 9 blue straggler stars (BSS), 3 yellow straggler stars (YSS) and 3 red clump (RC) stars. We construct multi-wavelength spectral energy distributions (SEDs) of these objects to characterize them and to estimate their parameters. We discovered hot companions to 3 BSS, 2 YSS and 3 RC candidates and estimated their properties. The hot companions with estimated temperatures, T$\mathrm{_{eff}}$ $\sim$ 13250--31000 K, are WDs of extremely low-mass ($\sim$ 0.20 M$_\odot$), low-mass ($\sim$ 0.20--0.40 M$_\odot$), normal mass ($\sim$ 0.40--0.60 M$_\odot$), and high-mass ($\sim$ 0.8 M$_\odot$). We suggest that systems with extremely low mass and low mass WDs as companions are formed via Case-A/Case-B mass transfer mechanism. A BSS is the likely progenitor of the high mass WD, as a star with more than the turn-off mass of the cluster is needed to form a high mass WD. Thus, systems with high mass WD are likely to be formed through merger in triple systems. We conclude that mass transfer as well as merger pathways of BSS formation are present in this cluster.

astro-ph.GA↗

arXiv2022

UOCS VII. Blue Straggler Populations of Open Cluster NGC 7789 with UVIT/AstroSat

NGC 7789 is a $\sim$1.6 Gyr old, populous open cluster located at $\sim$2000 pc. We characterize the blue straggler stars (BSS) of this cluster using the Ultraviolet (UV) data from the UVIT/AstroSat. We present spectral energy distributions (SED) of 15 BSS candidates constructed using multi-wavelength data ranging from UV to IR wavelengths. In 8 BSS candidates, a single temperature SED is found to be satisfactory. We discover hot companions in 5 BSS candidates. The hot companions with Teff $\sim$11750-15500 K, R $\sim$0.069-0.242 R$_{\odot}$, and L $\sim$0.25-1.55 L$_{\odot}$, are most likely extremely low mass (ELM) white dwarfs (WDs) with masses smaller than $\sim$0.18 M$_{\odot}$, and thereby confirmed post mass transfer systems. We discuss the implication of this finding in the context of BSS formation mechanisms. Two additional BSS show excess in one or more UV filters, and may have a hot companion, however, we are unable to characterize them. We suggest that at least 5 of the 15 BSS candidates (33%) studied in this cluster have formed via the mass transfer mechanism.

astro-ph.SR↗

arXiv2021

Determination of dynamical ages of open clusters through the A$^+$ parameter -- I

The sedimentation level of blue straggler stars (BSS) has been shown to be a great tool to investigate the dynamical states of globular clusters (GCs). The area enclosed between the cumulative radial distributions of BSS and a reference population up to the half-mass radius of the clusters, $A^+_{\mathrm{rh}}$, is known to be a measure of the sedimentation of BSS in GCs. In this work, we calculate $A^+_{\mathrm{rh}}$ for 11 open clusters (OCs) using a combined list of main-sequence turn-off stars, sub-giant branch stars, and red-giant branch stars as reference population. The BSS, the reference populations, and the cluster members are identified using the proper motions and parallaxes from the Gaia DR2 data. In a subset of clusters, the BSS are confirmed cluster members on the basis of radial velocity information available in the literature. Using the Pearson and Spearman rank correlation coefficients, we find weak correlations between the estimated values of $A^+_{\mathrm{rh}}$ and other markers of dynamical ages of the clusters, i.e., the number of central relaxations a cluster has experienced since its formation, and the structural parameters of the clusters. Based on statistical tests, we find that these correlations are similar to the corresponding correlations among the less evolved GCs, albeit within large errors.

astro-ph.GA↗

arXiv2021

Tidal tails in the disintegrating open cluster NGC 752

We utilize the robust membership determination algorithm, ML-MOC, on the precise astrometric and deep photometric data from Gaia Early Data Release 3 within a region of radius 5$^{\circ}$ around the center of the intermediate-age galactic open cluster NGC 752 to identify its member stars. We report the discovery of the tidal tails of NGC 752, extending out to $\sim$35 pc on either side of its denser central region and following the cluster orbit. From comparison with PARSEC stellar isochrones, we obtain the mass function of the cluster with a slope, $χ=-1.26\pm0.07$. The high negative value of $χ$ is indicative of a disintegrating cluster undergoing mass-segregation. $χ$ is more negative in the intra-tidal regions as compared to the outskirts of NGC 752. We estimate a present day mass of the cluster, M$\rm_{C}=297\pm10$ M$_{\odot}$. Through mass-loss due to stellar evolution and tidal interactions, we further estimate that NGC 752 has lost nearly 95.2-98.5 % of its initial mass, $\rm M_{i}~=~0.64~-2~\times~10^{4}~M_{\odot}$.

astro-ph.GA↗

arXiv2021

ML-MOC: Machine Learning (kNN and GMM) based Membership Determination for Open Clusters

The existing open cluster membership determination algorithms are either prior dependent on some known parameters of clusters or are not automatable to large samples of clusters. In this paper, we present, ML-MOC, a new machine learning based approach to identify likely members of open clusters using the Gaia DR2 data, and no a priori information about cluster parameters. We use the k-Nearest Neighbours (kNN) algorithm and the Gaussian Mixture Model (GMM) on the high-precision proper motions and parallax measurements from Gaia DR2 data to determine the membership probabilities of individual sources down to G ~20 mag. To validate the developed method, we apply it on fifteen open clusters: M67, NGC 2099, NGC 2141, NGC 2243, NGC 2539, NGC 6253, NGC 6405, NGC 6791, NGC 7044, NGC 7142, NGC 752, Blanco 1, Berkeley 18, IC 4651, and Hyades. These clusters differ in terms of their ages, distances, metallicities, extinctions and cover a wide parameter space in proper motions and parallaxes with respect to the field population. The extracted members produce clean colour-magnitude diagrams and our astrometric parameters of the clusters are in good agreement with the values derived by the previous works. The estimated degree of contamination in the extracted members range between 2% and 12%. The results show that ML-MOC is a reliable approach to segregate the open cluster members from the field stars.

astro-ph.IM↗

arXiv2020

Blue Straggler Populations of Seven Open Clusters with Gaia DR2

Blue straggler stars (BSS) are well studied in globular clusters but their systematic study with secure membership determination is lacking in open clusters. We use Gaia DR2 data to determine accurate stellar membership for four intermediate-age open clusters, Melotte 66, NGC 2158, NGC 2506 and NGC 6819, and three old open clusters, Berkeley 39, NGC 188 and NGC 6791, to subsequently study their BSS populations. The BSS radial distributions of five clusters, Melotte 66, NGC 188, NGC 2158, NGC 2506, and NGC 6791, show bimodal distributions, placing them with Family II globular clusters which are of intermediate dynamical ages. The location of minima, $r_\mathrm{min}$, in the bimodal BSS radial distributions, varies from 1.5$r_c$ to 4.0$r_c$, where $r_c$ is the core radius of the clusters. We find a positive correlation between $r_\mathrm{min}$ and $N_{\mathrm{relax}}$, the ratio of cluster age to the current central relaxation time of the cluster. We further report that this correlation is consistent in its slope, within the errors, to the slope of the globular cluster correlation between the same quantities, but with a slightly higher intercept. This is the first example in open clusters that shows BSS radial distributions as efficient probes of dynamical age. The BSS radial distributions of the remaining two clusters, Berkeley 39 and NGC 6819, are flat. The estimated $N_{\mathrm{relax}}$ values of these two clusters, however, indicate that they are dynamically evolved. Berkeley 39 especially has its entire BSS population completely segregated to the inner regions of the cluster.

astro-ph.GA↗