Multi-band Optical Photometry Research Articles

An Early-time Optical and Ultraviolet Excess in the Type-Ic SN 2020oi

Abstract We present photometric and spectroscopic observations of Supernova 2020oi (SN 2020oi), a nearby (∼17 Mpc) type-Ic supernova (SN Ic) within the grand-design spiral M100. We undertake a comprehensive analysis to characterize the evolution of SN 2020oi and constrain its progenitor system. We detect flux in excess of the fireball rise model δ t ≈ 2.5 days from the date of explosion in multiband optical and UV photometry from the Las Cumbres Observatory and the Neil Gehrels Swift Observatory, respectively. The derived SN bolometric luminosity is consistent with an explosion with M ej = 0.81 ± 0.03 M ⊙, E k = 0.79 ± 0.09 × 1051 erg s−1, and M Ni56 = 0.08 ± 0.02 M ⊙. Inspection of the event’s decline reveals the highest Δm 15,bol reported for a stripped-envelope event to date. Modeling of optical spectra near event peak indicates a partially mixed ejecta comparable in composition to the ejecta observed in SN 1994I, while the earliest spectrum shows signatures of a possible interaction with material of a distinct composition surrounding the SN progenitor. Further, Hubble Space Telescope pre-explosion imaging reveals a stellar cluster coincident with the event. From the cluster photometry, we derive the mass and age of the SN progenitor using stellar evolution models implemented in the BPASS library. Our results indicate that SN 2020oi occurred in a binary system from a progenitor of mass M ZAMS ≈ 9.5 ± 1.0 M ⊙, corresponding to an age of 27 ± 7 Myr. SN 2020oi is the dimmest SN Ic event to date for which an early-time flux excess has been observed, and the first in which an early excess is unlikely to be associated with shock cooling.

Optical photometry of two transitional millisecond pulsars in the radio pulsar state

ABSTRACT We present ULTRACAM multiband optical photometry of two transitional millisecond pulsars (tMSPs), PSR J1023+0038 and PSR J1227−4853, taken while both were in their radio pulsar states. The light curves show significant asymmetry about the flux maxima in all observed bands, suggesting an asymmetric source of heating in the system. We model the light curves using the icarus binary code, using models with an additional ‘hotspot’ heating contribution and an anisotropic heat redistribution model to treat the asymmetry. Our modelling reveals companion stars with underfilled Roche lobes in both PSRs J1023+0038 and J1227−4853, with Roche lobe filling factors in the range f ∼ 0.82−0.92. While the volume-averaged filling factors are closer to unity, significant underfilling is unexpected from tMSPs as they must rapidly overfill their Roche lobes to start transferring mass, which occurs on time-scale of weeks or months. We discuss the motivation and validity of our extensions to the models and the implications of the underfilled Roche lobe, and suggest future work to further investigate the role of the filling factor in the tMSP cycle.

Galactic cirri in deep optical imaging

The ubiquitous presence of Galactic cirri in deep optical images represents a major obstacle to study the low surface brightness features of extragalactic sources. To address this issue, we have explored the optical properties of cirri using g, r, i, and z bands in the Sloan Digital Sky Survey (SDSS) Stripe82 region. Using state-of-the-art, custom made, image processing techniques, including the modeling and removal of the instrumental scattered light produced by the stars, we managed to isolate the optical diffuse emission by the cirri, allowing their photometric characterization. We find that their optical colors are driven by the dust column density: The cirri become redder as their 100 μm emission increases. Remarkably, the optical colors of the Galactic cirri differ significantly from those of extragalactic sources, with a characteristic bluer r-i color for a given g-r, allowing one to detect these by using a simple color relation. Our results show the high potential of deep multi-band optical photometry, on its own, identifying the presence of cirri at a higher spatial resolution than those provided by far-infrared observations. The combination of very deep data and multi-band photometry (as the one produced by LSST and Euclid) would make it possible to build dust maps of unprecedented quality.

Further Evidence for Significant Luminosity Evolution in Supernova Cosmology

Abstract Supernova (SN) cosmology is based on the assumption that the corrected luminosity of SNe Ia would not evolve with redshift. Recently, our age dating of stellar populations in early-type host galaxies (ETGs) from high-quality spectra has shown that this key assumption is most likely in error. It has been argued though that the age–Hubble residual (HR) correlation from ETGs is not confirmed from two independent age data sets measured from multiband optical photometry of host galaxies of all morphological types. Here we show, however, that one of the data sets is based on highly uncertain and inappropriate luminosity-weighted ages derived, in many cases, under serious template mismatch. The other data set employs more reliable mass-weighted ages, but the statistical analysis involved is affected by regression dilution bias, severely underestimating both the slope and significance of the age–HR correlation. Remarkably, when we apply regression analysis with a standard posterior sampling method to this data set comprising a large sample (N = 102) of host galaxies, very significant (>99.99%) correlation is obtained between the global population age and HR with the slope (−0.047 ± 0.011 mag Gyr−1) highly consistent with our previous spectroscopic result from ETGs. For the local age of the environment around the site of SNe, a similarly significant (>99.96%) correlation is obtained with a steeper slope (−0.057 ± 0.016 mag Gyr−1). Therefore, the SN luminosity evolution is strongly supported by the age dating based on multiband optical photometry and can be a serious systematic bias in SN cosmology.

Resolved and Integrated Stellar Masses in the SDSS-IV/MaNGA Survey. II. Applications of PCA-based Stellar Mass Estimates

Abstract A galaxy’s stellar mass is one of its most fundamental properties, but it remains challenging to measure reliably. With the advent of very large optical spectroscopic surveys, efficient methods that can make use of low signal-to-noise spectra are needed. With this in mind, we created a new software package for estimating effective stellar mass-to-light ratios that uses a principal component analysis (PCA) basis set to optimize the comparison between observed spectra and a large library of stellar population synthesis models. In Paper I, we showed that with a set of six PCA basis vectors we could faithfully represent most optical spectra from the Mapping Nearby Galaxies at APO (MaNGA) survey, and we tested the accuracy of our M/L estimates using synthetic spectra. Here, we explore sources of systematic error in our mass measurements by comparing our new measurements to data from the literature. We compare our stellar mass surface density estimates to kinematics-derived dynamical mass surface density measurements from the DiskMass Survey and find some tension between the two that could be resolved if the disk scale heights used in the kinematic analysis were overestimated by a factor of ∼1.5. We formulate an aperture-corrected stellar mass catalog for the MaNGA survey, and compare to previous stellar mass estimates based on multiband optical photometry, finding typical discrepancies of 0.1 dex. Using the spatially resolved MaNGA data, we evaluate the impact of estimating total stellar masses from spatially unresolved spectra, and we explore how the biases that result from unresolved spectra depend upon the galaxy’s dust extinction and star formation rate. Finally, we describe an SDSS Value-Added Catalog that will include both spatially resolved and total (aperture-corrected) stellar masses for MaNGA galaxies.

Machine-learning-based Brokers for Real-time Classification of the LSST Alert Stream

The unprecedented volume and rate of transient events that will be discovered by the Large Synoptic Survey Telescope (LSST) demands that the astronomical community update its followup paradigm. Alert-brokers -- automated software system to sift through, characterize, annotate and prioritize events for followup -- will be critical tools for managing alert streams in the LSST era. The Arizona-NOAO Temporal Analysis and Response to Events System (ANTARES) is one such broker. In this work, we develop a machine learning pipeline to characterize and classify variable and transient sources only using the available multiband optical photometry. We describe three illustrative stages of the pipeline, serving the three goals of early, intermediate and retrospective classification of alerts. The first takes the form of variable vs transient categorization, the second, a multi-class typing of the combined variable and transient dataset, and the third, a purity-driven subtyping of a transient class. While several similar algorithms have proven themselves in simulations, we validate their performance on real observations for the first time. We quantitatively evaluate our pipeline on sparse, unevenly sampled, heteroskedastic data from various existing observational campaigns, and demonstrate very competitive classification performance. We describe our progress towards adapting the pipeline developed in this work into a real-time broker working on live alert streams from time-domain surveys.

Simultaneous Multiwavelength Observations of V404 Cygni during its 2015 June Outburst Decay Strengthen the Case for an Extremely Energetic Jet-base

Abstract We present results of multiband optical photometry of the black hole X-ray binary system V404 Cyg obtained using Wheaton College Observatory’s 0.3 m telescope, along with strictly simultaneous INTEGRAL and Swift observations during 2015 June 25.15–26.33 UT, and 2015 June 27.10–27.34 UT. These observations were made during the 2015 June outburst of the source when it was going through an epoch of violent activity in all wavelengths ranging from radio to γ-rays. The multiwavelength variability timescale favors a compact emission region, most likely originating in a jet outflow, for both observing epochs presented in this work. The simultaneous INTEGRAL/Imager on Board the Integral Satellite (IBIS) 20–40 keV light curve obtained during the June 27 observing run correlates very strongly with the optical light curve, with no detectable delay between the optical bands as well as between the optical and hard X-rays. The average slope of the dereddened spectral energy distribution was roughly flat between the - and V-bands during the June 27 run, even though the optical and X-ray flux varied by >25× during the run, ruling out an irradiation origin for the optical and suggesting that the optically thick to optically thin jet synchrotron break during the observations was at a frequency larger than that of V-band, which is quite extreme for X-ray binaries. These observations suggest that the optical emission originated very close to the base of the jet. A strong emission line, probably originating in a quasi-spherical nebula around the source, also contributes significantly in the R C -band. Our data, in conjunction with contemporaneous data at other wavelengths presented by other groups, strongly suggest that the jet-base was extremely compact and energetic during this phase of the outburst.

The bumpy light curve of Type IIn supernova iPTF13z over 3 years

A Type IIn supernova (SN) is dominated by the interaction of SN ejecta with the circumstellar medium (CSM). Some SNe IIn (e.g., SN 2006jd) have episodes of re-brightening ("bumps") in their light curves. We present iPTF13z, a SN IIn discovered by the intermediate Palomar Transient Factory (iPTF) and characterised by several bumps in its light curve. We analyse this peculiar behaviour trying to infer the properties of the CSM and of the SN explosion, as well as the nature of its progenitor star. We obtained multi-band optical photometry for over 1000 days after discovery with the P48 and P60 telescopes at Palomar Observatory. We obtained low-resolution optical spectra in the same period. We did an archival search for progenitor outbursts. We analyse our photometry and spectra, and compare iPTF13z to other SNe IIn. A simple analytical model is used to estimate properties of the CSM. iPTF13z was a SN IIn showing a light curve with five bumps during its decline phase. The bumps had amplitudes between 0.4 and 0.9 mag and durations between 20 and 120 days. The most prominent bumps appeared in all our different optical bands. The spectra showed typical SN IIn characteristics, with emission lines of H$\alpha$ (with broad component FWHM ~$10^{3}-10^{4} ~{\rm ~km ~s^{-1}}$ and narrow component FWHM ~$10^2 \rm ~km ~s^{-1}$) and He I, but also with Fe II, Ca II, Na I D and H$\beta$ P-Cygni profiles (with velocities of ~$10^{3}$ ${\rm ~km ~s^{-1}}$). A pre-explosion outburst was identified lasting $\gtrsim 50$ days, with $M_r \approx -15$ mag around 210 days before discovery. Large, variable progenitor mass-loss rates (~> 0.01 $M_{\odot} \rm ~yr^{-1}$) and CSM densities (~> 10$^{-16}$ g cm$^{-3}$) are derived. We suggest that the light curve bumps of iPTF13z arose from SN ejecta interacting with denser regions in the CSM, possibly produced by the eruptions of a luminous blue variable star.

VdBH 222: a starburst cluster in the inner Milky Way

It has been suggested that the compact open cluster VdBH 222 is a young massive distant object. We set out to characterise VdBH 222 using a comprehensive set of multi-wavelength observations. We obtained multi-band optical (UBVR) and near-infrared (JHK) photometry of the cluster field, as well as multi-object and long-slit optical spectroscopy for a large sample of stars in the field. We applied classical photometric analysis, as well as more sophisticated methods using the CHORIZOS code, to determine the reddening to the cluster. We then plotted dereddened HR diagrams and determined cluster parameters via isochrone fitting. We have identified a large population of luminous supergiants confirmed as cluster members via radial velocity measurements. We find nine red supergiants (plus one other candidate) and two yellow supergiants. We also identify a large population of OB stars. Ten of them are bright enough to be blue supergiants. The cluster lies behind 7.5 mag of extinction for the preferred value of Rv of 2.9. Isochrone fitting allows for a narrow range of ages between 12 and 16 Ma. The cluster radial velocity is compatible with distances of 6 and 10 kpc. The shorter distance is inconsistent with the age range and Galactic structure. The longer distance implies an age of 12 Ma and a location not far from the position where some Galactic models place the far end of the Galactic Bar. VdBH 222 is a young massive cluster with a likely mass greater than 20000 Msolar . Its population of massive evolved stars is comparable to that of large associations, such as Per OB1. Its location in the inner Galaxy, presumably close to the end of the Galactic bar, adds to the increasing evidence for vigorous star formation in the inner regions of the Milky Way.

The VIMOS Public Extragalactic Redshift Survey (VIPERS)

We describe the construction and general features of VIPERS, the VIMOS Public Extragalactic Redshift Survey. This `Large Programme' has been using the ESO VLT with the aim of building a spectroscopic sample of ~100,000 galaxies with i_{AB}<22.5 and 0.5<z<1.5. The survey covers a total area of ~24 deg^2 within the CFHTLS-Wide W1 and W4 fields. VIPERS is designed to address a broad range of problems in large-scale structure and galaxy evolution, thanks to a unique combination of volume (~ 5 x 10^7 h^{-3} Mpc^3) and sampling rate (~ 40%), comparable to state-of-the-art surveys of the local Universe, together with extensive multi-band optical and near-infrared photometry. Here we present the survey design, the selection of the source catalogue and the development of the spectroscopic observations. We discuss in detail the overall selection function that results from the combination of the different constituents of the project. This includes the masks arising from the parent photometric sample and the spectroscopic instrumental footprint, together with the weights needed to account for the sampling and the success rates of the observations. Using the catalogue of 53,608 galaxy redshifts composing the forthcoming VIPERS Public Data Release 1 (PDR-1), we provide a first assessment of the quality of the spectroscopic data. Benefiting from the combination of size and detailed sampling of this dataset, we conclude by presenting a map showing in unprecedented detail the large-scale distribution of galaxies between 5 and 8 billion years ago. [abridged]

CfA4: LIGHT CURVES FOR 94 TYPE Ia SUPERNOVAE

We present multi-band optical photometry of 94 spectroscopically-confirmed Type Ia supernovae (SN Ia) in the redshift range 0.0055 to 0.073, obtained between 2006 and 2011. There are a total of 5522 light curve points. We show that our natural system SN photometry has a precision of roughly 0.03 mag or better in BVr'i', 0.06 mag in u', and 0.07 mag in U for points brighter than 17.5 mag and estimate that it has a systematic uncertainty of 0.014, 0.010, 0.012, 0.014, 0.046, and 0.073 mag in BVr'i'u'U, respectively. Comparisons of our standard system photometry with published SN Ia light curves and comparison stars reveal mean agreement across samples in the range of ~0.00-0.03 mag. We discuss the recent measurements of our telescope-plus-detector throughput by direct monochromatic illumination by Cramer et al (in prep.). This technique measures the whole optical path through the telescope, auxiliary optics, filters, and detector under the same conditions used to make SN measurements. Extremely well-characterized natural-system passbands (both in wavelength and over time) are crucial for the next generation of SN Ia photometry to reach the 0.01 mag accuracy level. The current sample of low-z SN Ia is now sufficiently large to remove most of the statistical sampling error from the dark energy error budget. But pursuing the dark-energy systematic errors by determining highly-accurate detector passbands, combining optical and near-infrared (NIR) photometry and spectra, using the nearby sample to illuminate the population properties of SN Ia, and measuring the local departures from the Hubble flow will benefit from larger, carefully measured nearby samples.

Empirical Optical k-Corrections for Redshifts

ABSTRACTThe Smithsonian Hectospec Lensing Survey (SHELS) is a magnitude-limited spectroscopically complete survey for R ≤ 21.0 covering 4 deg2. SHELS provides a large sample (15,513) of flux-calibrated spectra. The wavelength range covered by the spectra allows empirical determination of k-corrections for the g and r bands from z = 0 to ∼0.68 and 0.33, respectively, based on large samples of spectra. We approximate the k-corrections using only two parameters in a standard way: Dn4000 and redshift, z. We use Dn4000 rather than the standard observed galaxy color, because Dn4000 is a redshift-independent tracer of the stellar population of the galaxy. Our approximations for the k-corrections using Dn4000 are as good as those based on observed galaxy color (g - r) (σ of the scatter is ∼0.08 mag). The approximations for the k-corrections are available in an online calculator. Our results agree with previously determined analytical approximations from single stellar population (SSP) models fitted to multiband optical and near-infrared photometry for galaxies with a known redshift. Galaxies with the smallest Dn4000 —the galaxies with the youngest stellar populations—are always attenuated and/or contain contributions from older stellar populations. We use simple single SSP fits to the SHELS spectra to study the influence of emission lines on the k-correction. The effects of emission lines can be ignored for rest-frame equivalent widths (REWs) ≲100 Å, depending on required photometric accuracy. We also provide analytic approximations to the k-corrections determined from our model fits for z ≤ 0.7 as a function of redshift and Dn4000 for ugriz and UBVRI (σ of the scatter is typically ∼0.10 mag, and the rms is typically ∼0.15 mag). Again, the approximations using Dn4000 are as good as those based on a suitably chosen observed galaxy color. We provide all analytical approximations in an online calculator.

A MULTI-COLOR OPTICAL SURVEY OF THE ORION NEBULA CLUSTER. II. THE H-R DIAGRAM

We present a new analysis of the stellar population of the Orion Nebula Cluster (ONC) based on multi-band optical photometry and spectroscopy. We study the color-color diagrams in BVI, plus a narrow-band filter centered at 6200A, finding evidences that intrinsic color scales valid for main-sequence dwarfs are incompatible with the ONC, while a better agreement is found employing synthetic intrinsic colors obtained constraining the typical lower surface gravity of young stars. We refine these model colors even further, empirically, by comparison with a sample of ONC stars with no accretion and no extinction. We consider the stars with known spectral types from the literature, and add 65 newly classified stars from slit spectroscopy and 182 M-type from narrow-band photometry; in this way we isolate a sample of about 1000 stars with known spectral type. We introduce a new method to self-consistently derive reddening and accretion excess from the location of each star in the BVI color-color diagram. This enables us to accurately determine the extinction of the ONC members. We adopt a lower distance for the ONC than previously assumed, based on recent parallax measurements. With a careful choice also of the spectral type-temperature transformation, we produce the new H-R diagram of the ONC population, more populated than previous works. With respect to previous works, we find higher luminosity for late-type stars and a lower luminosity for early types. We determine the age distribution of the population, peaking at 2-3 Myr, a higher age than previously estimated. We study the distribution of the members in the mass-age plane, and find that taking into account selection effects due to incompleteness removes an apparent correlation between mass and age. We derive the IMF for low- and intermediate-mass members of the ONC, which turns out to be model-dependent, and shows a turn-over at ~<0.2Msun.

On the Stellar Content of the Carina Dwarf Spheroidal Galaxy11Based on images collected with the MOSAIC II camera available at the CTIO 4 m Blanco telescope, La Serena; (2003B-0051, 2004B-0227, 2005B-0092, PI: A. R. Walker), and in part with the WFI available at the 2.2 m MPG/ESO telescope (A064.L-0327), and images obtained from the ESO/ST-ECF Science Archive Facility.

ABSTRACTWe present deep, accurate, and homogeneous multiband optical (U, B, V, I) photometry of the Carina dwarf spheroidal galaxy, based on more than 4000 individual CCD images from three different ground-based telescopes. Special attention was given to the photometric calibration, and the precision for the B, V, and I bands is generally better than 0.01 mag. We have performed detailed comparisons in the V, B - V, and V, B - I color-magnitude diagrams (CMDs) between Carina and three old, metal-poor Galactic globular clusters (GGCs, M53, M55, M79). We find that only the more metal-poor GCs (M55, [Fe/H] = -1.85; M53, [Fe/H] = -2.02 dex) provide a good match with the Carina giant branch. We have performed a similar comparison in the V, V - I CMD with three intermediate-age clusters (IACs) of the Small Magellanic Cloud (Kron 3, NGC 339, Lindsay 38). We find that the color extent of the subgiant branch (SGB) of the two more metal-rich IACs (Kron 3, [Fe/H] = -1.08; NGC339, [Fe/H] = -1.36 dex) is smaller than the range among Carina’s intermediate-age stars. Moreover, the slope of the RGB of these two IACs is shallower than the slope of the Carina RGB. However, the ridge line of the more metal-poor IAC (Lindsay 38, [Fe/H] = -1.59 dex) agrees quite well with the Carina intermediate-age stars. These findings indicate that Carina’s old stellar population is metal-poor and appears to have a limited spread in metallicity (Δ[Fe/H] = 0.2–0.3 dex). The Carina’s intermediate-age stellar population can hardly be more metal-rich than Lindsay 38, and its spread in metallicity also appears modest. We also find that the synthetic CMD constructed assuming a metallicity spread of 0.5 dex for the intermediate-age stellar component predicts evolutionary features not supported by observations. In particular, red clump stars should attain colors that are redder than red giant stars, but this is not seen. These results are at odds with recent spectroscopic investigations suggesting that Carina stars cover a broad range in metallicity (Δ[Fe/H] ∼ 1–2 dex). We also present a new method to estimate the metallicity of complex stellar systems using the difference in color between the red clump and the middle of the RR Lyrae instability strip. The observed colors of Carina’s evolved stars indicate a metallicity of [Fe/H] = -1.70 ± 0.19 dex, which agrees quite well with spectroscopic measurements.

V1647 ORIONIS: OPTICAL PHOTOMETRIC AND SPECTROSCOPIC MONITORING THROUGH THE 2003-2006 OUTBURST

We present results from an optical imaging and spectroscopic monitoring campaign on the young, low-mass eruptive variable star V1647 Orionis. The star and associated nebulosity (McNeil's Nebula) were observed over the period February 2004 to February 2006 with observations commencing a few months after the original outburst event occurred. Using the Gemini North telescope, we obtained multi-band optical imaging photometry and medium-resolution long-slit spectroscopy of V1647 Ori on an approximately monthly interval. During this period, V1647 Ori remained at, or close to, peak brightness and then faded by 5 magnitudes to close to its pre-outburst brightness. This implies an outburst timescale of around 27 months. Spectral features seen in both emission and absorption varied considerably during the monitoring period. For example, the Halpha line changed significantly in both intensity and profile. We present and discuss the observed photometric and spectroscopic changes and consider how this eruptive event relates to the early formative stages of low-mass stars.

Very low-mass members of the Lupus 3 cloud

We report on a multi-band survey for very low-mass stars and brown dwarfs in the Lupus 3 cloud with the Wide Field Imager (WFI) at the ESO/MPG 2.2m telescope on La Silla Observatory (Chile). Our multiband optical photometry is combined with available 2MASS JHK photometry to identify 19 new young stars and 3 brown dwarf candidates as probable members of this star forming region. Our objects are mostly clustered around the cloud core. Stars and brown dwarfs have similar levels of H-alpha emission, probably a signature of accretion. One object, a brown dwarf candidate, exhibits a near-infrared excess, which may indicate the presence of a disk, but its H-alpha emission cannot be confirmed due to its faintness in the optical passbands. We also find two visual pairs of probable Lupus 3 members that may be wide binaries.

Multiband optical photometry and bolometric light curve of the Type Ia supernova 2004S

We present BVRCIC broad-band CCD photometry of the Type Ia supernova SN 2004S, which appeared in the galaxy MCG–05-16-021, obtained during 2004 February 12 to March 22. We present multiband and bolometric light curves constructed using our data as well as other available data. The time of the B-band maximum and the peak magnitudes in different bands are obtained using fits of light-curve and colour templates. We clearly see a strong shoulder in the RC band and a second maximum in the IC band. SN 2004S closely resembles SN 1992al after maximum. From the peak bolometric luminosity, we estimate the ejected mass of 56 Ni to be 0.41 M⊙.

The Mass Assembly and Star Formation Characteristics of Field Galaxies of Known Morphology.

We discuss a new method for inferring the stellar mass of a distant galaxy of known redshift based on the combination of a near-IR luminosity and multiband optical photometry. The typical uncertainty for field galaxies with I<22 in the redshift range 0<z<1 is a factor of 2. We apply this method to a newly constructed sample of 321 field galaxies with redshifts and Hubble Space Telescope morphologies enabling us to construct the stellar mass density associated with various morphologies as a function of redshift. We find a marked decline with time in the stellar mass associated with peculiar galaxies accompanied by a modest rise in that observed for elliptical galaxies. The result suggests that peculiar galaxies decline in abundance because they transform and merge into regular systems. The star formation rate per unit stellar mass indicates that massive systems completed the bulk of their star formation before redshift 1, whereas dwarf galaxies continue to undergo major episodes of activity until the present epoch.