Search

Traditional weather forecasting relies on domain expertise and computationally intensive numerical simulation systems. Recently, with the development of a data-driven approach, weather forecasting based on deep learning has been receiving attention. Deep learning-based weather forecasting has made stunning progress, from various backbone studies using CNN, RNN, and Transformer to training strategies using weather observations datasets with auxiliary inputs. All of this progress has contributed to the field of weather forecasting; however, many elements and complex structures of deep learning models prevent us from reaching physical interpretations. This paper proposes a SImple baseline with a spatiotemporal context Aggregation Network (SIANet) that achieved state-of-the-art in 4 parts of 5 benchmarks of W4C22. This simple but efficient structure uses only satellite images and CNNs in an end-to-end fashion without using a multi-model ensemble or fine-tuning. This simplicity of SIANet can be used as a solid baseline that can be easily applied in weather forecasting using deep learning., Comment: 1st place solution for stage1 and Core Transfer in the Weather4Cast competition on NeurIPS 22

Deep learning-based weather prediction models have advanced significantly in recent years. However, data-driven models based on deep learning are difficult to apply to real-world applications because they are vulnerable to spatial-temporal shifts. A weather prediction task is especially susceptible to spatial-temporal shifts when the model is overfitted to locality and seasonality. In this paper, we propose a training strategy to make the weather prediction model robust to spatial-temporal shifts. We first analyze the effect of hyperparameters and augmentations of the existing training strategy on the spatial-temporal shift robustness of the model. Next, we propose an optimal combination of hyperparameters and augmentation based on the analysis results and a test-time augmentation. We performed all experiments on the W4C22 Transfer dataset and achieved the 1st performance., Comment: Core Transfer Track 1st place solution in Weather4Cast competition at NeuIPS22

We study the galaxy merger fraction and its dependence on star formation mode in the5.4 square degrees of the North Ecliptic Pole-Wide field. We select 6352 galaxies withAKARI 9{\mu}m detections, and identify mergers among them using the Gini coefficientand M20derived from the Subaru/HSC optical images. We obtain the total infraredluminosity and star formation rate of galaxies using the spectral energy distributiontemplates based on one band, AKARI 9{\mu}m. We classify galaxies into three differentstar formation modes (i.e. starbursts, main sequence, and quiescent galaxies) andcalculate the merger fractions for each. We find that the merger fractions of galaxiesincrease with redshift atz<0.6. The merger fractions of starbursts are higher thanthose of main sequence and quiescent galaxies in all redshift bins. We also examinethe merger fractions of far-infrared detected galaxies which have at least one detectionfromHerschel/SPIRE. We find thatHerscheldetected galaxies have higher mergerfraction compared to non-Herscheldetected galaxies, and bothHerscheldetected andnon-Herscheldetected galaxies show clearly different merger fractions depending onthe star formation modes., Comment: 13 pages, 11 figures, Accepted for publication by MNRAS

Context. The North Ecliptic Pole (NEP) field provides a unique set of panchromatic data, well suited for active galactic nuclei (AGN) studies. Selection of AGN candidates is often based on mid-infrared (MIR) measurements. Such method, despite its effectiveness, strongly reduces a catalog volume due to the MIR detection condition. Modern machine learning techniques can solve this problem by finding similar selection criteria using only optical and near-infrared (NIR) data. Aims. Aims of this work were to create a reliable AGN candidates catalog from the NEP field using a combination of optical SUBARU/HSC and NIR AKARI/IRC data and, consequently, to develop an efficient alternative for the MIR-based AKARI/IRC selection technique. Methods. A set of supervised machine learning algorithms was tested in order to perform an efficient AGN selection. Best of the models were formed into a majority voting scheme, which used the most popular classification result to produce the final AGN catalog. Additional analysis of catalog properties was performed in form of the spectral energy distribution (SED) fitting via the CIGALE software. Results. The obtained catalog of 465 AGN candidates (out of 33 119 objects) is characterized by 73% purity and 64% completeness. This new classification shows consistency with the MIR-based selection. Moreover, 76% of the obtained catalog can be found only with the new method due to the lack of MIR detection for most of the new AGN candidates. Training data, codes and final catalog are available via the github repository. Final AGN candidates catalog will be also available via the CDS service after publication., Comment: Accepted for publication in Astronomy and Astrophysics

To understand the cosmic accretion history of supermassive black holes, separating the radiation from active galactic nuclei (AGNs) and star-forming galaxies (SFGs) is critical. However, a reliable solution on photometrically recognising AGNs still remains unsolved. In this work, we present a novel AGN recognition method based on Deep Neural Network (Neural Net; NN). The main goals of this work are (i) to test if the AGN recognition problem in the North Ecliptic Pole Wide (NEPW) field could be solved by NN; (ii) to shows that NN exhibits an improvement in the performance compared with the traditional, standard spectral energy distribution (SED) fitting method in our testing samples; and (iii) to publicly release a reliable AGN/SFG catalogue to the astronomical community using the best available NEPW data, and propose a better method that helps future researchers plan an advanced NEPW database. Finally, according to our experimental result, the NN recognition accuracy is around 80.29% - 85.15%, with AGN completeness around 85.42% - 88.53% and SFG completeness around 81.17% - 85.09%., Comment: 12 pages, 12 figures

The $AKARI$ space infrared telescope has performed near- to mid-infrared (MIR) observations on the North Ecliptic Pole Wide (NEPW) field (5.4 deg$^2$) for about one year. $AKARI$ took advantage of its continuous nine photometric bands, compared with NASA's $Spitzer$ and WISE space telescopes, which had only four filters with a wide gap in the MIR. The $AKARI$ NEPW field lacked deep and homogeneous optical data, limiting the use of nearly half of the IR sources for extra-galactic studies owing to the absence of photometric redshifts (photo-zs). To remedy this, we have recently obtained deep optical imaging over the NEPW field with 5 bands ($g$, $r$, $i$, $z$, and $Y$) of the Hyper Suprime-Camera (HSC) on the Subaru 8m telescope. We optically identify AKARI-IR sources along with supplementary $Spitzer$ and WISE data as well as pre-existing optical data. In this work, we derive new photo-zs using a $\chi^2$ template-fitting method code ($Le$ $Phare$) and reliable photometry from 26 selected filters including HSC, $AKARI$, CFHT, Maidanak, KPNO, $Spitzer$ and WISE data. We take 2026 spectroscopic redshifts (spec-z) from all available spectroscopic surveys over the NEPW to calibrate and assess the accuracy of the photo-zs. At z < 1.5, we achieve a weighted photo-z dispersion of $\sigma_{\Delta{z/(1+z)}}$ = 0.053 with $\eta$ = 11.3% catastrophic errors., Comment: 20 pages, 13 figures, accepted for publication in MNRAS. For summary video, please see http://youtu.be/hjNJRCoBIgg

The north ecliptic pole (NEP) field is a natural deep field location for many satellite observations. It has been targeted manytimes since it was surveyed by the AKARI space telescope with its unique wavelength coverage from the near- to mid-infrared(mid-IR). Many follow-up observations have been carried out and made this field one of the most frequently observed areas witha variety of facilities, accumulating abundant panchromatic data from X-ray to radio wavelength range. Recently, a deep opticalsurvey with the Hyper Suprime-Cam (HSC) at the Subaru telescope covered the NEP-Wide (NEPW) field, which enabled us toidentify faint sources in the near- and mid-IR bands, and to improve the photometric redshift (photo-z) estimation. In this work,we present newly identified AKARI sources by the HSC survey, along with multi-band photometry for 91,861 AKARI sourcesobserved over the NEPW field. We release a new band-merged catalogue combining various photometric data from GALEXUV to the submillimetre (sub-mm) bands (e.g., Herschel/SPIRE, JCMT/SCUBA-2). About 20,000 AKARI sources are newlymatched to the HSC data, most of which seem to be faint galaxies in the near- to mid-infrared AKARI bands. This cataloguemotivates a variety of current research, and will be increasingly useful as recently launched (eROSITA/ART-XC) and futurespace missions (such as JWST, Euclid, and SPHEREx) plan to take deep observations in the NEP field., Comment: 17 pages, 12 Figures, accepted in MNRAS, (in press)

In order to understand the interaction between the central black hole and the whole galaxy or their co-evolution history along with cosmic time, a complete census of active galactic nuclei (AGN) is crucial. However, AGNs are often missed in optical, UV and soft X-ray observations since they could be obscured by gas and dust. A mid-infrared (mid-IR) survey supported by multiwavelength data is one of the best ways to find obscured AGN activities because it suffers less from extinction. Previous large IR photometric surveys, e.g., $WISE$ and $Spitzer$, have gaps between the mid-IR filters. Therefore, star forming galaxy (SFG)-AGN diagnostics in the mid-IR were limited. The $AKARI$ satellite has a unique continuous 9-band filter coverage in the near to mid-IR wavelengths. In this work, we take advantage of the state-of-the-art spectral energy distribution (SED) modelling software, CIGALE, to find AGNs in mid-IR. We found 126 AGNs in the NEP-Wide field with this method. We also investigate the energy released from the AGN as a fraction of the total IR luminosity of a galaxy. We found that the AGN contribution is larger at higher redshifts for a given IR luminosity. With the upcoming deep IR surveys, e.g., $JWST$, we expect to find more AGNs with our method., Comment: 14 pages, 18 figures. Accepted for publication in MNRAS. For associated video, please see https://www.youtube.com/watch?v=B01jL0Bol9Q&feature=emb_logo

Galaxy interactions are often accompanied by an enhanced star formation rate (SFR). Since molecular gas is essential for star formation, it is vital to establish whether, and by how much, galaxy interactions affect the molecular gas properties. We investigate the effect of interactions on global molecular gas properties by studying a sample of 58 galaxies in pairs and 154 control galaxies. Molecular gas properties are determined from observations with the JCMT, PMO, CSO telescopes, and supplemented with data from the xCOLD GASS and JINGLE surveys at $^{12}$CO(1-0) and $^{12}$CO(2-1). The SFR, gas mass ($M_\mathrm{H_{2}}$), and gas fraction ($f_{gas}$) are all enhanced in galaxies in pairs by $\sim$ 2.5 times compared to the controls matched in redshift, mass, and effective radius, while the enhancement of star formation efficiency (SFE $\equiv$ SFR/$M_{H_{2}}$) is less than a factor of 2. We also find that the enhancements in SFR, $M_{H_{2}}$ and $f_{gas}$ increase with decreasing pair separation and are larger in systems with smaller stellar mass ratio. Conversely, the SFE is only enhanced in close pairs (separation $<$ 20 kpc) and equal-mass systems; therefore most galaxies in pairs lie in the same parameter space on the SFR-$M_{H_{2}}$ plane as controls. This is the first time that the dependence of molecular gas properties on merger configurations is probed statistically with a relatively large sample and with a carefully-selected control sample for individual galaxies. We conclude that galaxy interactions do modify the molecular gas properties, although the strength of the effect is merger configuration dependent., Comment: 26 pages, 11 figures, accepted for publication in ApJ

We study the effects of bulge elongation on the star formation activity in the centers of spiral galaxies using the data from the Sloan Digital Sky Survey Data Release 7. We construct a volume-limited sample of face-on spiral galaxies with $M_r < -$19.5 mag at 0.02 $\leq z <$ 0.055 by excluding barred galaxies, where the aperture of the SDSS spectroscopic fibre covers the bulges of the galaxies. We adopt the ellipticity of bulges measured by Simard et al. (2011) who performed two-dimensional bulge+disc decompositions using the SDSS images of galaxies, and identify nuclear starbursts using the fibre specific star formation rates derived from the SDSS spectra. We find a statistically significant correlation between bulge elongation and nuclear starbursts in the sense that the fraction of nuclear starbursts increases with bulge elongation. This correlation is more prominent for fainter and redder galaxies, which exhibit higher ratios of elongated bulges. We find no significant environmental dependence of the correlation between bulge elongation and nuclear starbursts. These results suggest that non-axisymmetric bulges can efficiently feed the gas into the centre of galaxies to trigger nuclear starburst activity., Comment: 9 pages, 7 figures, accepted for publication in MNRAS

We study the star formation activity of nearby galaxies with bars using a sample of late-type galaxies at 0.02$\leq z \leq$ 0.05489 and $M_r <-19.5$ from the Sloan Digital Sky Survey. We compare the physical properties of strongly and weakly barred galaxies with those of non-barred galaxies that have stellar mass and redshift distributions similar to barred galaxies. We find that the star formation activity of strongly barred galaxies probed by starburstiness, $\it{g-r}$, NUV$-r$, and mid-infrared [3.4]$-$[12] colors is, on average, lower than that of non-barred galaxies. However, weakly barred galaxies do not show such a difference between barred and non-barred galaxies. The amounts of atomic and molecular gas in strongly barred galaxies are smaller than those of non-barred galaxies, and the gas metallicity is higher in strongly barred galaxies than in non-barred galaxies. The gas properties of weakly barred galaxies again show no difference from those of non-barred galaxies. We stack the optical spectra of barred and non-barred galaxies in several mass bins and fit to the stacked spectra with a spectral fitting code, STARLIGHT. We find no significant difference in stellar populations between barred and non-barred galaxies for both strongly and weakly barred galaxies. Our results are consistent with the idea that the star formation activity of barred galaxies is enhanced in the past along with significant gas consumption, and is currently lower than or similar to that of non-barred galaxies. The past star formation enhancement depends on the strength of bars., Comment: 13 pages, 11 figures, accepted for publication in ApJ

We describe the overall characteristics and the performance of an optical CCD camera system, Camera for QUasars in EArly uNiverse (CQUEAN), which is being used at the 2.1 m Otto Struve Telescope of the McDonald Observatory since 2010 August. CQUEAN was developed for follow-up imaging observations of red sources such as high redshift quasar candidates (z >= 5), Gamma Ray Bursts, brown dwarfs, and young stellar objects. For efficient observations of the red objects, CQUEAN has a science camera with a deep depletion CCD chip which boasts a higher quantum efficiency at 0.7 - 1.1 um than conventional CCD chips. The camera was developed in a short time scale (~ one year), and has been working reliably. By employing an auto-guiding system and a focal reducer to enhance the field of view on the classical Cassegrain focus, we achieve a stable guiding in 20 minute exposures, an imaging quality with FWHM >= 0.6" over the whole field (4.8' * 4.8'), and a limiting magnitude of z = 23.4 AB mag at 5-sigma with one hour total integration time., Comment: Accepted for publication in PASP. 26 pages including 5 tables and 24 figures

To perform imaging observation of optically red objects such as high redshift quasars and brown dwarfs, the Center for the Exploration of the Origin of the Universe (CEOU) recently developed an optical CCD camera, Camera for QUasars in EArly uNiverse(CQUEAN), which is sensitive at 0.7-1.1 um. To enable observations with long exposures, we developed an auto-guiding system for CQUEAN. This system consists of an off-axis mirror, a baffle, a CCD camera, a motor and a differential decelerator. To increase the number of available guiding stars, we designed a rotating mechanism for the off-axis guiding camera. The guiding field can be scanned along the 10 acrmin ring offset from the optical axis of the telescope. Combined with the auto-guiding software of the McDonald Observatory, we confirmed that a stable image can be obtained with an exposure time as long as 1200 seconds., Comment: Accepted for publication in Journal of Korean Astronomical Society (JKAS)