SearcharxivSearch

arXiv subjects

Roberto Molinaro

Publications and source records attributed to Roberto Molinaro.

At least 19 recordsLinked to original sources

Classical Cepheids in the JWST and Roman Era: A New Theoretical Framework and its Metallicity Dependence

Classical Cepheids are fundamental distance indicators that anchor the cosmic distance ladder through their Period-Luminosity, Period-Luminosity-Color, and Period-Wesenheit relations, while also providing valuable constraints on young stellar populations. Recent advances in space-based near-infrared observations highlight the need for theoretical calibrations of Classical Cepheid observables in modern photometric systems optimized for precision distance-scale applications. This work provides a homogeneous theoretical framework for Classical Cepheids in the JWST/NIRCam and Roman/WFI systems and explores their use as standard candles in the near-infrared. We exploit an extensive grid of nonlinear convective pulsation models spanning a broad range of masses, effective temperatures, periods, metallicities, convective efficiencies, and mass-luminosity relations. We derive mean magnitudes, colors, pulsation amplitudes, amplitude ratios, and new theoretical Period-Luminosity-Color, Period-Wesenheit, and metal-dependent Period-Wesenheit relations. The theoretical Wesenheit relations are compared with recent JWST observations of Classical Cepheids in six SN Ia host galaxies from the SH0ES collaboration. We find excellent agreement between predictions and observations, with the largest systematic effects arising from the adopted mass-luminosity relation. Theoretical distance moduli are fully consistent with empirical JWST determinations when a mildly overluminous mass-luminosity relation is adopted instead of the canonical one. These results provide a self-consistent pulsation-based calibration for JWST and Roman observations and a theoretical framework for future precision studies of the local distance scale.

astro-ph.SR

Do AI weather models miss extremes?

First-generation AI weather models are often reported to underperform at extremes, mostly in reanalysis-based evaluations of deterministic regression systems. We verify eleven physical and AI forecast systems against European synoptic, solar, and rain-gauge stations over ten months for 10 m wind, 2 m temperature, hourly shortwave accumulation, and hourly precipitation, scoring mean absolute error (MAE) against ECMWF IFS in ERA5 1991-2020 climatological regimes. Among these systems, AI models do not show a uniform relative-skill deficit in the tails. Jua EPT-2.1 Europa leads all-conditions wind (+8.4%), while Jua EPT-2 HRRR leads temperature overall (+12.1%) and in the heat regime (+19.6 +/- 2.2%). EPT-2.1 Europa and DWD ICON Global lead at gale-force wind. Jua EPT-2.1 Helios leads solar overall (+10.2 +/- 1.7%), in overcast conditions (+16.4 +/- 3.4%), and in the clear-sky tail (+24.8 +/- 5.4%). For precipitation, three Jua models gain 14-15% at moderate intensity and 9-11% at P75-P95; EPT-2 Reasoning remains ahead above P95 (+1.7 +/- 0.5%). Failures are model-specific: ECMWF AIFS loses 4.9 +/- 2.0% in the heat tail, while NOAA GFS loses 22.8 +/- 2.0% there. Every model, including numerical weather prediction systems, shows a shared conditional bias toward the centre of the observed distribution, with an inter-model spread several times smaller than the shared signal. Missing relative skill at extremes is therefore not a property of AI weather models as a class, but of particular AI and physical models.

physics.ao-ph

Determination of [Fe/H] in Fundamental mode Classical Cepheids from Gaia DR3 light curve Fourier parameters

Estimating the iron abundance ([Fe/H]) of Classical Cepheids (CCs) is crucial both for validating the cosmic distance scale through the Period-Luminosity relation and for tracing the chemical evolution of young stellar populations. However, standard spectroscopic determinations are highly time-consuming, limiting their application to large or distant samples. This work delivers a practical, multi-band alternative by investigating whether light-curve Fourier parameters derived from Gaia mission photometry can serve as reliable proxies for [Fe/H]. Our reference sample comprises 398 fundamental-mode CCs from Gaia DR3 with high-resolution spectroscopic metallicities available in the literature, spanning a wide range of pulsation periods. Light curves in the G, G_BP, and G_RP bands were modeled with a truncated Fourier series to extract characteristic amplitudes, amplitude ratios, and phase differences. We implemented a replicated Random Forest algorithm as a feature selection tool to isolate the most predictive parameters based on their total increase in node purity. Using the selected features, we calibrated an empirical, multi-variable relation via linear regression, incorporating higher-order polynomial terms to capture complex morphological dependencies. The primary unweighted OLS model yields a residual rms of 0.213 dex and a Pearson correlation of 0.61. As an external validity check, we applied this calibration to large target samples of CCs in the Milky Way, LMC, and SMC. The resulting photometric metallicity distributions correctly reproduce the expected chemical enrichment history and peak values of the three galaxies, proving the reliability of this relation for population studies in the Gaia DR4 era.

astro-ph.SR

Universal Diffusion-Based Probabilistic Downscaling

We introduce a universal diffusion-based downscaling framework that lifts deterministic low-resolution weather forecasts into probabilistic high-resolution predictions without any model-specific fine-tuning. A single conditional diffusion model is trained on paired coarse-resolution inputs (~25 km resolution) and high-resolution regional reanalysis targets (~5 km resolution), and is applied in a fully zero-shot manner to deterministic forecasts from heterogeneous upstream weather models. Focusing on near-surface variables, we evaluate probabilistic forecasts against independent in situ station observations over lead times up to 90 h. Across a diverse set of AI-based and numerical weather prediction (NWP) systems, the ensemble mean of the downscaled forecasts consistently improves upon each model's own raw deterministic forecast, and substantially larger gains are observed in probabilistic skill as measured by CRPS. These results demonstrate that diffusion-based downscaling provides a scalable, model-agnostic probabilistic interface for enhancing spatial resolution and uncertainty representation in operational weather forecasting pipelines.

cs.LG

New theoretical predictions on Type II Cepheids: towards a self consistent Pop. II distance scale

Type II Cepheids are pulsating stars that can be used as standard candles for old stellar populations due to their characteristic Period-Luminosity and Period-Luminosity-Colour relations. They are traditionally divided in 3 sub-classes, namely BL Her, W Vir and RV Tauri. In this paper we focus on the first two sub-classes, to provide a new theoretical scenario and develop tools and relations to be adopted in distance scale and old stellar populations studies. We have built new nonlinear convective pulsation models of Type II Cepheids, computed along selected stellar evolution tracks and spanning a wide range of pulsation period and stellar parameters. Three chemical compositions have been taken into account. For each assumed Z and Y, models have been computed following stellar evolution predictions for off Zero Age Horizontal Branch evolution of stellar masses lower than typical RR Lyrae stars, crossing the classical instability strip as BL Her or W Vir pulsating stars. A new theoretical prediction for the instability strip boundaries of these classes of variable stars has been obtained together with their dependence on metal abundance. The predicted light and radial velocity curves have been computed along the evolution inside the strip, showing how the amplitude and the morphology are affected by the position relative to the edges and by the luminosity and mass values. The transformation of bolometric light curves into various photometric systems allowed us to provide new theoretical Period-Luminosity and Period-Wesenheit relations for BL Her and W Vir. These relations are consistent with previously published RR Lyrae model results but with a smaller metallicity dependence. Moreover, the application of the inferred theoretical relations to Magellanic and Galactic Type II Cepheid data provides results in good agreement with some independent distance estimates in the literature.

astro-ph.SR

EPT-2 Technical Report

We present EPT-2, the latest iteration in our Earth Physics Transformer (EPT) family of foundation AI models for Earth system forecasting. EPT-2 delivers substantial improvements over its predecessor, EPT-1.5, and sets a new state of the art in predicting energy-relevant variables-including 10m and 100m wind speed, 2m temperature, and surface solar radiation-across the full 0-240h forecast horizon. It consistently outperforms leading AI weather models such as Microsoft Aurora, as well as the operational numerical forecast system IFS HRES from the European Centre for Medium-Range Weather Forecasts (ECMWF). In parallel, we introduce a perturbation-based ensemble model of EPT-2 for probabilistic forecasting, called EPT-2e. Remarkably, EPT-2e significantly surpasses the ECMWF ENS mean-long considered the gold standard for medium- to longrange forecasting-while operating at a fraction of the computational cost. EPT models, as well as third-party forecasts, are accessible via the app.jua.ai platform.

cs.LG

The VMC Survey -- LIV. Anomalous Cepheids in the Magellanic Clouds Period-Luminosity relations in the near-infrared bands

Anomalous Cepheids (ACs) are less studied metal-poor pulsating stars ([Fe/H]<-1.5) compared to Classical Cepheids (CCs) and RR Lyrae stars. They follow distinct Period-Luminosity (PL) and Period-Wesenheit (PW) relations and pulsate in either the fundamental (F) or first overtone (1O) mode. Our goal is to assess the precision and accuracy of AC-based distances and evaluate their potential for establishing an independent distance scale. We derive new PL and PW relations for F-mode, 1O-mode, and, for the first time, combined F+1O ACs in the Magellanic Clouds. We study their wavelength dependence and apply the relations to estimate distances to Local Group stellar systems hosting ACs, while also confirming AC classifications. Our analysis is based on near-infrared time-series photometry in the Y, J, and Ks bands for about 200 ACs in the Magellanic Clouds from the VISTA survey of the Magellanic Clouds system (VMC, 2009-2018). VMC data are complemented with optical photometry from Gaia DR3 and OGLE-IV, which also provide periods and pulsation modes. Custom light-curve templates were used to derive precise intensity-averaged magnitudes for 118 ACs in the Large Magellanic Cloud (LMC) and 75 in the Small Magellanic Cloud. These data were used to derive multi-band PL and PW relations, calibrated using the geometric LMC distance from eclipsing binaries. We find that PL relation slopes increase and dispersions decrease with wavelength. Using Gaia parallaxes, we determine the LMC distance modulus and the LMC-SMC relative distance. We also confirm the AC nature of several new candidates in Galactic Globular Clusters and derive a distance modulus for the Draco dSph galaxy of 19.425+/-0.048 mag. A 0.1 mag discrepancy with RR Lyrae-based distances may reflect metallicity effects. Future spectroscopic surveys and Gaia DR4 will help refine the AC distance scale and quantify metallicity impacts.

astro-ph.SR

Spatial Age Distribution of Classical Cepheids in Spiral Galaxies: The Cases of M31 and M33

Classical Cepheids can be used as age indicators due to well-established period-age and period-age-color relations. \citet{Desomma2021} refined these relations by including a metallicity term and different Mass-Luminosity assumptions. In this study, we apply the period-age-metallicity relation for the first time to samples of Classical Cepheids in M31 and M33. For both galaxies, we consider Cepheid coordinates and spatial distributions, along with the metallicity gradients by \citet{Zaritsky1994} and \citet{Magrini2007}, to provide a metallicity estimate for each pulsator. By applying the period-age-metallicity relation, we derive individual ages for each Cepheid. Combining these ages and spatial distributions, we construct detailed age maps for both galaxies. Our analysis confirms a radial age gradient, with younger Cepheids preferentially found toward the galactic centers. In M31, we confirm an outer ring at $\sim 11$ kpc, consistent with previous studies, and identify for the first time an inner ring at $\sim 7$ kpc, possibly associated with star formation episodes. Comparing age gradients at different angles, we find a consistent general trend of ages increasing monotonically with radius. At the same time, we observe smaller-scale differences, particularly in the $90^\circ$-$180^\circ$ quadrant, suggesting asymmetric star formation and possible dynamical influences. In contrast, M33 displays a steeper global age gradient, indicating a higher concentration of young stars toward its center. This study highlights the utility of Cepheids as stellar population tracers, providing insights into the star formation and dynamical evolution of spiral galaxies. Future works will extend this methodology to additional galaxies.

astro-ph.GA

Stellar Pulsation and Evolution: a Combined Theoretical Renewal and Updated Models (SPECTRUM) -- I: Updating radiative opacities for pulsation models of Classical Cepheid and RR-Lyrae

Pulsating stars are universally recognized as precise distance indicators and tracers of stellar populations. Their variability, combined with well-defined relationships between pulsation properties and intrinsic evolutionary parameters such as luminosity, mass, and age, makes them essential for understanding galactic evolution and retrieving star formation histories. Therefore, accurate modeling of pulsating stars is crucial for using them as standard candles and stellar population tracers. This is the first paper in the "Stellar Pulsation and Evolution: a Combined Theoretical Renewal and Updated Models" (SPECTRUM) project, which aims to present an update of Stellingwerf's hydrodynamical pulsation code, by adopting the latest radiative opacity tables commonly used in stellar evolution community. We assess the impact of this update on pulsation properties, such as periods, instability strip topology, and light curve shapes, as well as on Period Wesenheit and Period-Luminosity relations for Classical Cepheids and RR Lyrae stars, comparing the results with those derived using older opacity data. Our results indicate that the opacity update introduces only minor changes: instability strip boundary locations shift by no more than $100K$ in effective temperature, and pulsation periods vary within $1\sigma$ compared to previous evaluations. Light curves exhibit slight differences in shape and amplitude. Consequently, the theoretical calibration of the Cepheid or RRL-based extragalactic distance scale remains largely unaffected by the opacity changes. However, achieving consistency in opacity tables between stellar evolution and pulsation codes is a significant step toward a homogeneous and self-consistent stellar evolution and pulsation framework.

astro-ph.SR

EPT-1.5 Technical Report

We announce the release of EPT-1.5, the latest iteration in our Earth Physics Transformer (EPT) family of foundation AI earth system models. EPT-1.5 demonstrates substantial improvements over its predecessor, EPT-1. Built specifically for the European energy industry, EPT-1.5 shows remarkable performance in predicting energy-relevant variables, particularly 10m & 100m wind speed and solar radiation. Especially in wind prediction, it outperforms existing AI weather models like GraphCast, FuXi, and Pangu-Weather, as well as the leading numerical weather model, IFS HRES by the European Centre for Medium-Range Weather Forecasts (ECMWF), setting a new state of the art.

cs.AI

Generative AI for fast and accurate statistical computation of fluids

We present a generative AI algorithm for addressing the pressing task of fast, accurate, and robust statistical computation of three-dimensional turbulent fluid flows. Our algorithm, termed as GenCFD, is based on an end-to-end conditional score-based diffusion model. Through extensive numerical experimentation with a set of challenging fluid flows, we demonstrate that GenCFD provides an accurate approximation of relevant statistical quantities of interest while also efficiently generating high-quality realistic samples of turbulent fluid flows and ensuring excellent spectral resolution. In contrast, ensembles of deterministic ML algorithms, trained to minimize mean square errors, regress to the mean flow. We present rigorous theoretical results uncovering the surprising mechanisms through which diffusion models accurately generate fluid flows. These mechanisms are illustrated with solvable toy models that exhibit the mathematically relevant features of turbulent fluid flows while being amenable to explicit analytical formulae. Our codes are publicly available at https://github.com/camlab-ethz/GenCFD.

cs.LG

Poseidon: Efficient Foundation Models for PDEs

We introduce Poseidon, a foundation model for learning the solution operators of PDEs. It is based on a multiscale operator transformer, with time-conditioned layer norms that enable continuous-in-time evaluations. A novel training strategy leveraging the semi-group property of time-dependent PDEs to allow for significant scaling-up of the training data is also proposed. Poseidon is pretrained on a diverse, large scale dataset for the governing equations of fluid dynamics. It is then evaluated on a suite of 15 challenging downstream tasks that include a wide variety of PDE types and operators. We show that Poseidon exhibits excellent performance across the board by outperforming baselines significantly, both in terms of sample efficiency and accuracy. Poseidon also generalizes very well to new physics that is not seen during pretraining. Moreover, Poseidon scales with respect to model and data size, both for pretraining and for downstream tasks. Taken together, our results showcase the surprising ability of Poseidon to learn effective representations from a very small set of PDEs during pretraining in order to generalize well to unseen and unrelated PDEs downstream, demonstrating its potential as an effective, general purpose PDE foundation model. Finally, the Poseidon model as well as underlying pretraining and downstream datasets are open sourced, with code being available at https://github.com/camlab-ethz/poseidon and pretrained models and datasets at https://huggingface.co/camlab-ethz.

cs.LG

The Hertzsprung progression of Classical Cepheids in the Gaia era

A new fine grid of nonlinear convective pulsation models for the so-called "bump Cepheids" is presented to investigate the Hertzprung progression (HP) phenomenon shown by their light and radial pulsation velocity curves. The period corresponding to the center of the HP is investigated as a function of various model assumptions, such as the efficiency of super-adiabatic convection, the mass-luminosity relation, and the metal and helium abundances. The assumed mass-luminosity relation is found to significantly affect the phenomenon but variations in the chemical composition as well as in the stellar mass (at fixed mass-luminosity relation) also play a key role in determining the value of the HP center period. Finally, the predictive capability of the presented theoretical scenario is tested against observed light curves of bump Cepheids in the ESA Gaia database, also considering the variation of the pulsation amplitudes and of the Fourier parameters $R_{21}$ and $\Phi_{21}$ with the pulsation period. A qualitative agreement between theory and observations is found for what concerns the evolution of the light curve morphology as the period moves across the HP center, as well for the pattern in period-amplitude, period-$R21$ and period-$\Phi_{21}$ planes. A larger sample of observed Cepheids with accurate light curves and metallicities is required in order to derive more quantitative conclusions.

astro-ph.SR

Classical Cepheid Pulsation properties in the Rubin-LSST filters

Homogeneous multi-wavelength observations of classical Cepheids from the forthcoming Rubin-LSST have the potential to significantly contribute to our understanding of the evolutionary and pulsation properties of these pulsating stars. Updated pulsation models for Classical Cepheid stars have been computed under various assumptions about chemical compositions, including relatively low metallicity ($Z$ = $0.004$ with $Y$ =$0.25$ and $Z$=$0.008$ with $Y$ =$0.25$), solar metallicity ($Z$=$0.02$ with $Y$=$0.28$), and supersolar metallicity environments ($Z$ = $0.03$ with $Y$ = $0.28$). From the predicted periods, intensity-weighted mean magnitudes, and colors, we have derived the first theoretical pulsation relations in the Rubin-LSST filters (ugrizy), including period-luminosity-color, period-Wesenheit, and period-age-color relations. We find that the coefficients of these relations are almost insensitive to the efficiency of superadiabatic convection but are significantly affected by the assumption of the mass-luminosity relation and the adopted chemical composition. Metal-dependent versions of these relations are also derived, representing valuable tools for individual distance determinations and correction for metallicity effects on the cosmic distance scale.

astro-ph.SR

The VMC Survey -- L. Type II Cepheids in the Magellanic Clouds

Type II Cepheids (T2C) are less frequently used counterparts of classical Cepheids which provide the primary calibration of the distance ladder for measuring $H_0$ in the local Universe. In the era of the Hubble Tension, T2C variables with the RR Lyrae stars (RRL) and the tip of the red giant branch (TRGB) can potentially provide classical Cepheid independent calibration of the cosmic distance ladder. Our goal is to provide an absolute calibration of the Period-Luminosity, Period-Luminosity-Color and Period-Wesenheit relations(PL,PLC and PW) of T2Cs in the Large Magellanic Cloud (LMC). We exploited time-series photometry in the near-infrared (NIR) bands for a sample of more than 320 T2Cs in the Magellanic Clouds (MC). These observations were acquired during 2009-2018 in the context of the VMC ESO public survey (The VISTA near-infrared YJKs survey of the Magellanic System). The NIR photometry was supplemented with well-sampled optical light curves and accurate pulsation periods from the OGLE IV survey and the Gaia mission. We used the best-quality NIR light curves to generate custom templates for modelling sparsely sampled light curves in YJKs bands; in turn, we derived accurate and precise intensity-averaged mean magnitudes and pulsation amplitudes of 339 T2Cs in the MC. We used optical and NIR mean magnitudes to derive PL/PLC/PW relations for T2Cs in multiple bands, which were calibrated with the geometric distance to the LMC and with the Gaia parallaxes. We used our new empirical calibrations of PL/PW relations to obtain distances to 22 T2C-host Galactic globular clusters, which were found to be systematically smaller by 0.1 mag and 0.03-0.06 mag compared with the literature. A better agreement is found between our distances and those based on RRLs in globular clusters, providing strong support for using these population II stars with the TRGB for future distance scale studies.

astro-ph.SR

High-resolution Spectroscopic Metallicities of Milky Way Cepheid Standards and their impact on the Leavitt Law and the Hubble constant

Milky Way Cepheid variables with accurate {\it Hubble Space Telescope} photometry have been established as standards for primary calibration of the cosmic distance ladder to achieve a percent-level determination of the Hubble constant ($H_0$). These 75 Cepheid standards are the fundamental sample for investigation of possible residual systematics in the local $H_0$ determination due to metallicity effects on their period-luminosity relations. We obtained new high-resolution ($R\sim81,000$), high signal-to-noise ($S/N\sim50-150$) multi-epoch spectra of 42 out of 75 Cepheid standards using ESPaDOnS instrument at the 3.6-m Canada-France-Hawaii Telescope. Our spectroscopic metallicity measurements are in good agreement with the literature values with systematic differences up to $0.1$ dex due to different metallicity scales. We homogenized and updated the spectroscopic metallicities of all 75 Milky Way Cepheid standards and derived their multiwavelength ($GVIJHK_s$) period-luminosity-metallicity and period-Wesenheit-metallicity relations using the latest {\it Gaia} parallaxes. The metallicity coefficients of these empirically calibrated relations exhibit large uncertainties due to low statistics and a narrow metallicity range ($\Delta\textrm{[Fe/H]}=0.6$~dex). These metallicity coefficients are up to three times better constrained if we include Cepheids in the Large Magellanic Cloud and range between $-0.21\pm0.07$ and $-0.43\pm0.06$ mag/dex. The updated spectroscopic metallicities of these Milky Way Cepheid standards were used in the Cepheid-Supernovae distance ladder formalism to determine $H_0=72.9~\pm 1.0$\textrm{~km~s$^{-1}$~Mpc$^{-1}$}, suggesting little variation ($\sim 0.1$ ~km~s$^{-1}$~Mpc$^{-1}$) in the local $H_0$ measurements due to different Cepheid metallicity scales.

astro-ph.SR

Representation Equivalent Neural Operators: a Framework for Alias-free Operator Learning

Recently, operator learning, or learning mappings between infinite-dimensional function spaces, has garnered significant attention, notably in relation to learning partial differential equations from data. Conceptually clear when outlined on paper, neural operators necessitate discretization in the transition to computer implementations. This step can compromise their integrity, often causing them to deviate from the underlying operators. This research offers a fresh take on neural operators with a framework Representation equivalent Neural Operators (ReNO) designed to address these issues. At its core is the concept of operator aliasing, which measures inconsistency between neural operators and their discrete representations. We explore this for widely-used operator learning techniques. Our findings detail how aliasing introduces errors when handling different discretizations and grids and loss of crucial continuous structures. More generally, this framework not only sheds light on existing challenges but, given its constructive and broad nature, also potentially offers tools for developing new neural operators.

cs.LG

Convolutional Neural Operators for robust and accurate learning of PDEs

Although very successfully used in conventional machine learning, convolution based neural network architectures -- believed to be inconsistent in function space -- have been largely ignored in the context of learning solution operators of PDEs. Here, we present novel adaptations for convolutional neural networks to demonstrate that they are indeed able to process functions as inputs and outputs. The resulting architecture, termed as convolutional neural operators (CNOs), is designed specifically to preserve its underlying continuous nature, even when implemented in a discretized form on a computer. We prove a universality theorem to show that CNOs can approximate operators arising in PDEs to desired accuracy. CNOs are tested on a novel suite of benchmarks, encompassing a diverse set of PDEs with possibly multi-scale solutions and are observed to significantly outperform baselines, paving the way for an alternative framework for robust and accurate operator learning. Our code is publicly available at https://github.com/bogdanraonic3/ConvolutionalNeuralOperator

cs.LG