arXiv subjects

Raoul Canameras

Publications and source records attributed to Raoul Canameras.

5 recordsLinked to original sources

arXiv2026

HOLISMOKES XXI: Detecting strongly lensed type Ia supernovae from time series of multi-band LSST-like imaging data -- Part II

Strong gravitationally lensed supernovae (LSNe) are rare but extremely valuable probes of cosmology and astrophysics. Prompt identification within the alert streams of time-domain surveys such as the Rubin Legacy Survey of Space and Time (LSST) is essential for timely follow-up observations. In our previous study, Bag et al. (2026), we introduced a deep-learning framework for detecting LSNe Ia directly from multi-band, multi-epoch image cutouts. The model employs a convolutional LSTM architecture to capture spatiotemporal correlations in time-series imaging data, enabling classification updates as new observations arrive. In this work, we extend that framework by incorporating greater realism into the simulations. In particular, we present a method to construct realistic image time series from single-epoch observations by introducing epoch-to-epoch point spread function variations with corresponding variance-map corrections. The dataset is based on HSC PDR3 observations and includes simulated lensed host-galaxy arcs, SN light-curve variations, and Poisson noise. We also introduce an additional negative class consisting of SN Ia occurring in the foreground lens galaxy, representing a challenging source of false positives. Despite these additional complexities, the model retains strong performance. The receiver operating characteristic improves rapidly during the first few observations, reaching a true-positive rate of $\sim60\%$ at a false-positive rate of $\mathcal{O}(10^{-4})$ by the seventh observation and $\sim80\%$ by the tenth. We also investigate potential confusion with sibling SNe occurring in LRGs and identify the configurations that best mimic lensed systems. These results demonstrate that the image-time-series approach remains robust under more realistic observing conditions, and is well suited for real-time LSN searches in LSST and other time-domain surveys.

astro-ph.IM↗

arXiv2026

HOLISMOKES XVII: Detecting strongly lensed SNe Ia from time series of multi-band LSST-like imaging data

Strong gravitationally lensed supernovae (LSNe), though rare, are exceptionally valuable probes for cosmology and astrophysics. Upcoming time-domain surveys like the Vera Rubin Observatory's Legacy Survey of Space and Time (LSST) offer a major opportunity to discover them in large numbers. Early identification is crucial for timely follow-up observations. We develop a deep learning pipeline to detect LSNe using multi-band, multi-epoch image cutouts. Our model is based on a 2D convolutional long short-term memory (ConvLSTM2D) architecture, designed to capture both spatial and temporal correlations in time-series imaging data. Predictions are made after each observation in the time series, with accuracy improving as more data arrive. We train the model on realistic simulations derived from Hyper Suprime-Cam (HSC) data, which closely matches LSST in depth and filters. This work focuses exclusively on Type Ia supernovae (SNe Ia). LSNe Ia are injected onto HSC luminous red galaxies (LRGs) at various phases of evolution to create positive examples. Negative examples include variable sources from HSC Transient Survey (including unclassified transients), and simulated unlensed SNe Ia in LRG and spiral galaxies. Our multi-band model shows rapid classification improvements during the initial few observations and quickly reaches high detection efficiency: at a fixed false-positive rate (FPR) of $0.01\%$, the true-positive rate (TPR) reaches $\gtrsim 60\%$ by the 7th observation and exceeds $\gtrsim 70\%$ by the 9th. Among the negative examples, SNe in LRGs remain the primary source of FPR, as they can resemble their lensed counterparts under certain conditions. The model detects quads more effectively than doubles and performs better on systems with larger image separations. Although trained and tested on HSC-like data, our approach applies to any cadenced imaging survey, particularly LSST.

astro-ph.IM↗

arXiv2025

Using Convolutional Neural Networks to Search for Strongly Lensed Quasars in KiDS DR5

Gravitationally strongly lensed quasars (SL-QSO) offer invaluable insights into cosmological and astrophysical phenomena. With the data from ongoing and next-generation surveys, thousands of SL-QSO systems can be discovered expectedly, leading to unprecedented opportunities. However, the challenge lies in identifying SL-QSO from enormous datasets with high recall and purity in an automated and efficient manner. Hence, we developed a program based on a Convolutional Neural Network (CNN) for finding SL-QSO from large-scale surveys and applied it to the Kilo-degree Survey Data Release 5 (KiDS DR5). Our approach involves three key stages: firstly, we pre-selected ten million bright objects (with $r$-band $\tt{MAG\_AUTO} < 22$), excluding stars from the dataset; secondly, we established realistic training and test sets to train and fine-tune the CNN, resulting in the identification of 4195 machine candidates, and the false positive rate (FPR) of $\sim$1/2000 and recall of 0.8125 evaluated by using the real test set containing 16 confirmed lensed quasars; thirdly, human inspections were performed for further selections, and then 272 SL-QSO candidates were eventually found in total, including 16 high-score, 118 median-score, and 138 lower-score candidates, separately. Removing the systems already confirmed or identified in other papers, we end up with 229 SL-QSO candidates, including 7 high-score, 95 median-score, and 127 lower-score candidates, and the corresponding catalog is publicly available online. We have also included an excellent quad candidate in the appendix, discovered serendipitously during the fine-tuning process of the CNN.

astro-ph.GA↗

arXiv2024

Detecting unresolved lensed SNe Ia in LSST using blended light curves

Strongly gravitationally lensed supernovae (LSNe) are promising probes for providing absolute distance measurements using gravitational-lens time delays. Spatially unresolved LSNe offer an opportunity to enhance the sample size for precision cosmology. We predict that there will be approximately three times as many unresolved as resolved LSNe Ia in the Legacy Survey of Space and Time (LSST) by the {\it Rubin} Observatory. In this article, we explore the feasibility of detecting unresolved LSNe Ia from a pool of preclassified SNe Ia light curves using the shape of the blended light curves with deep-learning techniques. We find that $\sim 30\%$ unresolved LSNe Ia can be detected with a simple 1D convolutional neural network (CNN) using well-sampled $rizy$-band light curves (with a false-positive rate of $\sim 3\%$). Even when the light curve is well observed in only a single band among $r$, $i$, and $z$, detection is still possible with false-positive rates ranging from $\sim 4$ to $7\%$ depending on the band. Furthermore, we demonstrate that these unresolved cases can be detected at an early stage using light curves up to $\sim20$ days from the first observation with well-controlled false-positive rates, providing ample opportunity to trigger follow-up observations. Additionally, we demonstrate the feasibility of time-delay estimations using solely LSST-like data of unresolved light curves, particularly for doubles, when excluding systems with low time delays and magnification ratios. However, the abundance of such systems among those unresolved in LSST poses a significant challenge. This approach holds potential utility for upcoming wide-field surveys, and overall results could significantly improve with enhanced cadence and depth in the future surveys.

astro-ph.IM↗

arXiv2024

Systematic comparison of neural networks used in discovering strong gravitational lenses

Efficient algorithms are being developed to search for strong gravitational lens systems owing to increasing large imaging surveys. Neural networks have been successfully used to discover galaxy-scale lens systems in imaging surveys such as the Kilo Degree Survey, Hyper-Suprime Cam (HSC) Survey and Dark Energy Survey over the last few years. Thus, it has become imperative to understand how some of these networks compare, their strengths and the role of the training datasets as most of the networks make use of supervised learning algorithms. In this work, we present the first-of-its-kind systematic comparison and benchmarking of networks from four teams that have analysed the HSC Survey data. Each team has designed their training samples and developed neural networks independently but coordinated apriori in reserving specific datasets strictly for test purposes. The test sample consists of mock lenses, real (candidate) lenses and real non-lenses gathered from various sources to benchmark and characterise the performance of each of the network. While each team's network performed much better on their own constructed test samples compared to those from others, all networks performed comparable on the test sample with real (candidate) lenses and non-lenses. We also investigate the impact of swapping the training samples amongst the teams while retaining the same network architecture. We find that this resulted in improved performance for some networks. These results have direct implications on measures to be taken for lens searches with upcoming imaging surveys such as the Rubin-Legacy Survey of Space and Time, Roman and Euclid.

astro-ph.GA↗