Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userPredicting the global far-infrared SED of galaxies via machine learning techniques
162
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
Dust plays an important role in shaping a galaxys spectral energy distribution (SED). It absorbs ultraviolet (UV) to near-infrared (NIR) radiation and re-emits this energy in the far-infrared (FIR). The FIR is essential to understand dust in galaxies. However, deep FIR observations require a space mission, none of which are still active today. We aim to infer the FIR emission across six Herschel bands, along with dust luminosity, mass, and effective temperature, based on the available UV to mid-infrared (MIR) observations. We also want to estimate the uncertainties of these predictions, compare our method to energy balance SED fitting, and determine possible limitations of the model. We propose a machine learning framework to predict the FIR fluxes from 14 UV-MIR broadband fluxes. We used a low redshift sample by combining DustPedia and H-ATLAS, and extracted Bayesian flux posteriors through SED fitting. We trained shallow neural networks to predict the far-infrared fluxes, uncertainties, and dust properties. We evaluated them on a test set using a root mean square error (RMSE) in log-space. Our results (RMSE = 0.19 dex) significantly outperform UV-MIR energy balance SED fitting (RMSE = 0.38 dex), and are inherently unbiased. We can identify when the predictions are off, for example when the input has large uncertainties on WISE 22, or when the input does not resemble the training set. The galaxies for which we have UV-FIR observations can be used as a blueprint for galaxies that lack FIR data. This results in a virtual FIR telescope, which can be applied to large optical-MIR galaxy samples. This helps bridge the gap until the next FIR mission.
rate research
Read More
We develop a machine learning-based framework to predict the HI content of galaxies using more straightforwardly observable quantities such as optical photometry and environmental parameters. We train the algorithm on z=0-2 outputs from the Mufasa cosmological hydrodynamic simulation, which includes star formation, feedback, and a heuristic model to quench massive galaxies that yields a reasonable match to a range of survey data including HI. We employ a variety of machine learning methods (regressors), and quantify their performance using the root mean square error ({sc rmse}) and the Pearson correlation coefficient (r). Considering SDSS photometry, 3$^{rd}$ nearest neighbor environment and line of sight peculiar velocities as features, we obtain r $> 0.8$ accuracy of the HI-richness prediction, corresponding to {sc rmse}$<0.3$. Adding near-IR photometry to the features yields some improvement to the prediction. Compared to all the regressors, random forest shows the best performance, with r $>0.9$ at $z=0$, followed by a Deep Neural Network with r $>0.85$. All regressors exhibit a declining performance with increasing redshift, which limits the utility of this approach to $zla 1$, and they tend to somewhat over-predict the HI content of low-HI galaxies which might be due to Eddington bias in the training sample. We test our approach on the RESOLVE survey data. Training on a subset of RESOLVE data, we find that our machine learning method can reasonably well predict the HI-richness of the remaining RESOLVE data, with {sc rmse}$sim0.28$. When we train on mock data from Mufasa and test on RESOLVE, this increases to {sc rmse}$sim0.45$. Our method will be useful for making galaxy-by-galaxy survey predictions and incompleteness corrections for upcoming HI 21cm surveys such as the LADUMA and MIGHTEE surveys on MeerKAT, over regions where photometry is already available.
Massive Population II galaxies undergoing the first phase of vigorous star formation after the initial Population III stage should have high energy densities and silicate-rich interstellar dust. We have modeled the resulting far-infrared spectral energy distributions (SEDs), demonstrating that they are shifted substantially to bluer (`warmer) wavelengths relative to the best fitting ones at z ~ 3, and with strong outputs in the 10 - 40 micron range. When combined with a low level of emission by carbon dust, their SEDs match that of Haro 11, a local moderately-low-metallicity galaxy undergoing a very young and vigorous starburst that is likely to approximate the relevant conditions in young Population II galaxies. We expect to see similar SEDs at high redshifts (z >= 5) given the youth of galaxies at this epoch. In fact, we find a progression with redshift in observed galaxy SEDs, from those resembling local ones at 2 < z < 4 to a closer resemblance with Haro 11 at 5 < z < 7. In addition to the insight on conditions in high redshift galaxies, this result implies that estimates of the total infrared luminosities at z ~ 6 based on measurements near lambda ~ 1 mm can vary by factors of 2 - 4, depending on the SED template used. Currently popular modified blackbodies or local templates can result in significant underestimates compared with the preferred template based on the SED of Haro 11.
Understanding the ionizing spectrum of low-metallicity galaxies is of great importance for modeling and interpreting emission line observations of early/distant galaxies. Although a wide suite of stellar evolution, atmosphere, population synthesis, and photoionization models, taking many physical processes into account now exist, all models face a common problem: the inability to explain the presence of nebular HeII emission, which is observed in many low metallicity galaxies, both in UV and optical spectra. Several possible explanations have been proposed in the literature, including Wolf-Rayet (WR) stars, binaries, very massive stars, X-ray sources, or shocks. However, none has so far been able to explain the major observations. We briefly discuss the HeII problem, available empirical data, and observed trends combining X-ray, optical and other studies. We present a simple and consistent physical model showing that X-ray binaries could explain the long-standing nebular HeII problem. Our model, described in Schaerer et al. (2019), successfully explains the observed trends and strength of nebular HeII emission in large samples of low metallicity galaxies and in individual galaxies, which have been studied in detail and with multi-wavelength observations. Our results have in particular important implications for the interpretation of galaxy spectra in the early Universe, which will be obtained with upcoming and future facilities.
Modelling and interpreting the SEDs of galaxies has become one of the key tools at the disposal of extragalactic astronomers. Ideally, we could hope that, through a detailed study of its SED, we can infer the correct physical properties and the evolutionary history of a galaxy. In the past decade, panchromatic SED fitting, i.e. modelling the SED over the entire UV-submm wavelength regime, has seen an enormous advance. Several advanced new codes have been developed, nearly all based on Bayesian inference modelling. In this review, we briefly touch upon the different ingredients necessary for panchromatic SED modelling, and discuss the methodology and some important aspects of Bayesian SED modelling. The current uncertainties and limitations of panchromatic SED modelling are discussed, and we explore some avenues how the models and techniques can potentially be improved in the near future.
To explore the connection between the global physical properties of galaxies and their far-infrared (FIR) spectral energy distributions (SEDs), we study the variation in the FIR SEDs of a set of hydrodynamically simulated galaxies that are generated by performing dust radiative transfer in post-processing. Our sample includes both isolated and merging systems at various stages of the merging process and covers infrared (IR) luminosities and dust masses that are representative of both low- and high-redshift galaxies. We study the FIR SEDs using principle component analysis (PCA) and find that 97% of the variance in the sample can be explained by two principle components (PCs). The first PC characterizes the wavelength of the peak of the FIR SED, and the second encodes the breadth of the SED. We find that the coefficients of both PCs can be predicted well using a double power law in terms of the IR luminosity and dust mass, which suggests that these two physical properties are the primary determinants of galaxies FIR SED shapes. Incorporating galaxy sizes does not significantly improve our ability to predict the FIR SEDs. Our results suggest that the observed redshift evolution in the effective dust temperature at fixed IR luminosity is not driven by geometry: the SEDs of $z sim 2-3$ ultraluminous IR galaxies (ULIRGs) are cooler than those of local ULIRGs not because the high-redshift galaxies are more extended but rather because they have higher dust masses at fixed IR luminosity. Finally, based on our simulations, we introduce a two-parameter set of SED templates that depend on both IR luminosity and dust mass.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
