arXiv subjects

Andrés Alejandro Plazas Malagón

Publications and source records attributed to Andrés Alejandro Plazas Malagón.

5 recordsLinked to original sources

arXiv2026

Multi-Method Machine Learning Classification of Astronomical Sources with Rubin Data Previews: a Spectroscopy-Free Approach with DP1 and DP2

We present a multi-method machine-learning study for the binary photometric classification of point-like and extended sources in the Rubin Observatory Data Preview 1 (DP1) and Data Preview 2 (DP2). The workflow runs on the Rubin Science Platform using DP1 and DP2 ugrizy photometry, the DP1 ancillary photometric-redshift catalog, and the Monster astrometric and photometric reference catalog, without relying on external spectroscopy. Six methods (SVM, Random Forest, XGBoost, k-Nearest Neighbors, an Artificial Neural Network, and a soft Voting ensemble) are benchmarked in the deepest DP1 field, the Extended Chandra Deep Field South. These methods reach 10-fold cross-validation macro-F1 scores of 0.95-0.97 on the balanced training subset and test macro-F1 scores of 0.87-0.92 on the imbalanced held-out set, with RF and the SV ensemble tied at the top (0.918); the differences among the leading methods are within the statistical uncertainty. Permutation importance identifies FWHM as the dominant feature, and the six methods agree on 79.5% of the full catalog. The same pipeline is then applied without retuning the settings to the two other DP1 fields with full ugrizy coverage and photometric-redshift support, Euclid Deep Field South and the Low Galactic Latitude Field. Together, the three fields span a factor of approximately 7 in E(B-V) and approximately 10 in stellar density, yet headline macro-F1 remains above 0.87 in every field. We then apply the same pipeline to four DP2 fields (ECDFS, COSMOS, EDFS, and ELAIS-S1). Using colors and morphology, the best method reaches a test macro-F1 of 0.83 to 0.88 across these fields, and the six methods agree on 77 to 86% of each catalog. Adding the per-band photometric uncertainties as features raises the best macro-F1 to about 0.95 in ECDFS DP1 and 0.97 in ECDFS DP2 (and to approximately 0.97-0.98 in the other fields).

astro-ph.IM↗

arXiv2026

Weak-lensing mass calibration of \emph{Planck} Sunyaev--Zel'dovich clusters with HSC-SSP Year~3

We present a weak gravitational lensing mass calibration of 19 \textit{Planck} Sunyaev--Zel'dovich (SZ) selected galaxy clusters using shape measurements from the Hyper Suprime-Cam Subaru Strategic Program (HSC-SSP) Year 3 shape catalog. We measure the stacked weak-lensing signal $ΔΣ(R)$ using per-cluster lensing weights that match the measurement pipeline's stacking scheme, and construct an analytical covariance matrix that includes shape noise and projected large-scale structure contributions. Our primary constraint on the SZ mass bias comes from a forward-modeling approach that integrates over the halo mass function while accounting for the \textit{Planck} SZ selection function, Eddington bias from log-normal scatter in the SZ mass proxy, and cluster miscentering. Fitting four free parameters, the log mass bias $\ln(1-b)$, the miscentered fraction $f_{\rm mis}$, the offset scale $σ_{\rm off}$, and the SZ scatter $σ_{\ln M}$, over the radial range $0.5$--$5.0\,h^{-1}\,\mathrm{Mpc}$, we obtain $1-b = 0.73^{+0.10}_{-0.11}$ with $χ^2/\mathrm{dof} = 5.2/5$ at an effective redshift $z_{\rm eff}\simeq 0.24$. This measurement is consistent with recent weak-lensing calibrations of SZ-selected clusters and supports the picture that significant mass bias corrections are required to reconcile cluster abundance measurements with primary cosmic microwave background constraints on cosmological parameters.

astro-ph.CO↗

arXiv2026

Identifying and Measuring Satellite Streaks in DECam Images

The rapid growth of satellite constellations, particularly Starlink, is increasingly affecting ground-based astronomy. In this project, we developed a workflow to detect, identify, and measure the brightness of trails from artificial satellites and other orbiting objects in archival images from the Dark Energy Camera (DECam), available through the NOIRLab Data Archive. We filtered images with visible streaks, retrieved detector-level images, applied the Hough Transform (via satmetrics) to detect and align trails, and performed surface brightness photometry. We also used SatChecker to obtain likely identifications for each trail. Our sample of nine measured streaks includes Starlink satellites, a navigation satellite, a decommissioned science satellite, and a rocket body. Our results show that satellites and other orbiting objects are consistently detectable in DECam images, but their brightness varies significantly, reflecting design and operational differences across object types and models. While the methodology proved effective, detecting faint streaks was challenging, and short-lived glints remain an even harder problem for future work. This proof-of-concept establishes a foundation for larger statistical studies of satellite impacts on astronomical surveys. The code is available at https://github.com/iausathub/reca-streaks

astro-ph.IM↗

arXiv2026

Deep Learning for Point Spread Function Modeling in Cosmology

We present the development of a data-driven, AI-based model of the Point Spread Function (PSF) that achieves higher accuracy than the current state-of-the-art approach, "PSF in the Full Field-of-View'' (PIFF). PIFF is widely used in leading weak-lensing surveys, including the Dark Energy Survey (DES), the Hyper Suprime-Cam (HSC) Survey, and the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST). The PSF characterizes how a point source, such as a star, is imaged after its light traverses the atmosphere and telescope optics, effectively representing the "blurred fingerprint'' of the entire imaging system. Accurate PSF modeling is essential for weak gravitational lensing analyses, as biases in its estimation propagate directly into cosmic shear measurements -- one of the primary cosmological probes of the expansion history of the Universe and the growth of large-scale structure for dark energy studies. To address the limitations of PIFF, which constructs PSF models independently for each CCD and therefore loses spatial coherence across the focal plane, we introduce a deep-learning-based framework for PSF reconstruction. In this approach, an autoencoder is trained on stellar images obtained with the Hyper Suprime-Cam (HSC) of the Subaru Telescope and combined with a Gaussian process to interpolate the PSF across the telescope's full field of view. This hybrid model captures systematic variations across the focal plane and achieves a reconstruction error of $3.4 \times 10^{-6}$ compared to PIFF's $3.7 \times 10^{-6}$, laying the foundation for integration into the LSST Science Pipelines.

astro-ph.IM↗

arXiv2025

Recommendations to overcome language barriers in the Vera C. Rubin Observatory Research Ecosystem

The report presents a comprehensive set of five recommendations to reduce language barriers within the Vera C. Rubin Observatory Research Ecosystem, promoting greater inclusion of researchers who are speakers of English as an additional language. Recognizing that English linguistic hegemony in science limits participation and productivity, the document proposes multilingual presentation formats, academic writing training, a Virtual Writing Center, language support programs, and writing retreats. Each recommendation is grounded in both pedagogical theory and empirical evidence, with an emphasis on collaborative, socially embedded approaches to scientific writing. The proposed academic writing training integrates constructivist and socio-cultural perspectives, emphasizing genre awareness, rhetorical competence, and reflective practices. The Virtual Writing Center would serve as a permanent infrastructure offering personalized tutoring and peer review support, while the language support programs address ongoing needs through workshops, consultations, and access to language tools. Writing retreats provide immersive environments for focused work and mentorship. The recommendations also encourage ethical use of AI tools for translation and writing assistance, fostering digital literacy alongside linguistic proficiency. Collectively, these initiatives aim to transform language from a barrier into a resource, recognizing multilingualism as an asset in global research collaboration. Rather than offering a one-size-fits-all solution, the document advocates for adaptable, community-driven strategies that can evolve within the diverse institutional and disciplinary contexts of the Rubin Research Ecosystem. By implementing these practices, the Ecosystem could lead efforts to democratize scientific communication and foster a more equitable, multilingual research culture.

astro-ph.IM↗