We have obtained structural parameters of about 340, 000 galaxies from the Kilo Degree Survey (KiDS) in 153 square degrees of data release 1, 2 and 3. We have performed a seeing convolved 2D single Sérsic fit to the galaxy images in the 4 photometric bands (u, g, r, i) observed by KiDS, by selecting high signal-to-noise ratio (S/N > 50) systems in every bands.We have classified galaxies as spheroids and disc-dominated by combining their spectral energy distribution properties and their Sérsic index. Using photometric redshifts derived from a machine learning technique, we have determined the evolution of the effective radius, R e and stellar mass, M ⋆ , versus redshift, for both mass complete samples of spheroids and disc-dominated galaxies up to z∼ 0.6.Our results show a significant evolution of the structural quantities at intermediate redshift for the massive spheroids (Log M * /M ⊙ > 11, Chabrier IMF), while almost no evolution has found for less massive ones (Log M * /M ⊙ < 11). On the other hand, disc dominated systems show a milder evolution in the less massive systems (Log M * /M ⊙ < 11) and possibly no evolution of the more massive systems. These trends are generally consistent with predictions from hydrodynamical simulations and independent datasets out to redshift z ∼ 0.6, although in some cases the scatter of the data is large to drive final conclusions.These results, based on 1/10 of the expected KiDS area, reinforce precedent finding based on smaller statistical samples and show the route toward more accurate results, expected with the the next survey releases.

Context. The Kilo-Degree Survey (KiDS) is an ongoing optical wide-field imaging survey with the OmegaCAM camera at the VLT Survey Telescope. It aims to image 1500 square degrees in four filters (ugri). The core science driver is mapping the large-scale matter distribution in the Universe, using weak lensing shear and photometric redshift measurements. Further science cases include galaxy evolution, Milky Way structure, detection of high-redshift clusters, and finding rare sources such as strong lenses and quasars. Aims. Here we present the third public data release and several associated data products, adding further area, homogenized photometric calibration, photometric redshifts and weak lensing shear measurements to the first two releases. Methods. A dedicated pipeline embedded in the Astro-WISE information system is used for the production of the main release. Modifications with respect to earlier releases are described in detail. Photometric redshifts have been derived using both Bayesian template fitting, and machine-learning techniques. For the weak lensing measurements, optimized procedures based on the THELI data reduction and lensfit shear measurement packages are used. Results. In this third data release an additional 292 new survey tiles (≈300 deg 2 ) stacked ugri images are made available, accompanied by weight maps, masks, and source lists. The multi-band catalogue, including homogenized photometry and photometric redshifts, covers the combined DR1, DR2 and DR3 footprint of 440 survey tiles (44 deg 2 ). Limiting magnitudes are typically 24.3, 25.1, 24.9, 23.8 (5σ in a 2 aperture) in ugri, respectively, and the typical r-band PSF size is less than 0.7 . The photometric homogenization scheme ensures accurate colours and an absolute calibration stable to ≈2% for gri and ≈3% in u. Separately released for the combined area of all KiDS releases to date are a weak lensing shear catalogue and photometric redshifts based on two different machine-learning techniques.

We present a machine-learning photometric redshift (ML photo-z) analysis of the Kilo-Degree Survey Data Release 3 (KiDS DR3), using two neural-network based techniques: ANNz2 and MLPQNA. Despite limited coverage of spectroscopic training sets, these ML codes provide photo-zs of quality comparable to, if not better than, those from the Bayesian Photometric Redshift (BPZ) code, at least up to zphot ≲ 0.9 and r ≲ 23.5. At the bright end of r ≲ 20, where very complete spectroscopic data overlapping with KiDS are available, the performance of the ML photo-zs clearly surpasses that of BPZ, currently the primary photo-z method for KiDS. Using the Galaxy And Mass Assembly (GAMA) spectroscopic survey as calibration, we furthermore study how photo-zs improve for bright sources when photometric parameters additional to magnitudes are included in the photo-z derivation, as well as when VIKING and WISE infrared (IR) bands are added. While the fiducial four-band ugri setup gives a photo-z bias 〈δz/(1 + z)〉 = −2 × 10−4 and scatter σδz/(1+z) < 0.022 at mean 〈z〉 = 0.23, combining magnitudes, colours, and galaxy sizes reduces the scatter by ~7% and the bias by an order of magnitude. Once the ugri and IR magnitudes are joined into 12-band photometry spanning up to 12 μm, the scatter decreases by more than 10% over the fiducial case. Finally, using the 12 bands together with optical colours and linear sizes gives 〈δz/(1 + z)〉 < 4 × 10−5 and σδz/(1+z) < 0.019. This paper also serves as a reference for two public photo-z catalogues accompanying KiDS DR3, both obtained using the ANNz2 code. The first one, of general purpose, includes all the 39 million KiDS sources with four-band ugri measurements in DR3. The second dataset, optimised for low-redshift studies such as galaxy-galaxy lensing, is limited to r ≲ 20, and provides photo-zs of much better quality than in the full-depth case thanks to incorporating optical magnitudes, colours, and sizes in the GAMA-calibrated photo-z derivation.

A variety of fundamental astrophysical science topics require the determination of very accurate photometric redshifts (photo-z's). A wide plethora of methods have been developed, based either on template models fitting or on empirical explorations of the photometric parameter space. Machine learning based techniques are not explicitly dependent on the physical priors and able to produce accurate photo-z estimations within the photometric ranges derived from the spectroscopic training set. These estimates, however, are not easy to characterize in terms of a photo-z Probability Density Function (PDF), due to the fact that the analytical relation mapping the photometric parameters onto the redshift space is virtually unknown. We present METAPHOR (Machine-learning Estimation Tool for Accurate PHOtometric Redshifts), a method designed to provide a reliable PDF of the error distribution for empirical techniques. The method is implemented as a modular workflow, whose internal engine for photo-z estimation makes use of the MLPQNA neural network (Multi Layer Perceptron with Quasi Newton learning rule), with the possibility to easily replace the specific machine learning model chosen to predict photo-z's. We present a summary of results on SDSS-DR9 galaxy data, used also to perform a direct comparison with PDF's obtained by the Le Phare SED template fitting. We show that METAPHOR is capable to estimate the precision and reliability of photometric redshifts obtained with three different self-adaptive techniques, i.e. MLPQNA, Random Forest and the standard K-Nearest Neighbors models.

Photometric redshifts (photo-z's) are fundamental in galaxy surveys to address different topics, from gravitational lensing and dark matter distribution to galaxy evolution. The Kilo Degree Survey (KiDS), i.e. the ESO public survey on the VLT Survey Telescope (VST), provides the unprecedented opportunity to exploit a large galaxy dataset with an exceptional image quality and depth in the optical wavebands. Using a KiDS subset of about 25, 000 galaxies with measured spectroscopic redshifts, we have derived photo-z's using i) three different empirical methods based on supervised machine learning, ii) the Bayesian Photometric Redshift model (or BPZ), and iii) a classical SED template fitting procedure (Le PHARE). We confirm that, in the regions of the photometric parameter space properly sampled by the spectroscopic templates, machine learning methods provide better redshift estimates, with a lower scatter and a smaller fraction of outliers. SED fitting techniques, however, provide useful information on the galaxy spectral type which can be effectively used to constrain systematic errors and to better characterize potential catastrophic outliers. Such classification is then used to specialize the training of regression machine learning models, by demonstrating that a hybrid approach, involving SED fitting and machine learning in a single collaborative framework, can be effectively used to improve the accuracy of photo-z estimates.