preparation

To date, galaxy image simulations for weak lensing surveys usually approximate the light profiles of all galaxies as a single or double Sérsic profile, neglecting the influence of galaxy substructures and morphologies deviating from such a simplified parametric characterisation. While this approximation may be sufficient for previous data sets, the stringent cosmic shear calibration requirements and the high quality of the data in the upcoming Euclid survey demand a consideration of the effects that realistic galaxy substructures and irregular shapes have on shear measurement biases. Here we present a novel deep learning-based method to create such simulated galaxies directly from Hubble Space Telescope (HST) data. We first build and validate a convolutional neural network based on the wavelet scattering transform to learn noise-free representations independent of the point-spread function (PSF) of HST galaxy images. These can be injected into simulations of images from Euclid’s optical instrument VIS without introducing noise correlations during PSF convolution or shearing. Then, we demonstrate the generation of new galaxy images by sampling from the model randomly as well as conditionally. In the latter case, we fine-tune the interpolation between latent space vectors of sample galaxies to directly obtain new realistic objects following a specific Sérsic index and half-light radius distribution. Furthermore, we show that the distribution of galaxy structural and morphological parameters of our generative model matches the distribution of the input HST training data, proving the capability of the model to produce realistic shapes. Next, we quantify the cosmic shear bias from complex galaxy shapes in Euclid-like simulations by comparing the shear measurement biases between a sample of model objects and their best-fit double-Sérsic counterparts, thereby creating two separate branches that only differ in the complexity of their shapes. Using the Kaiser, Squires, and Broadhurst shape measurement algorithm, we find a multiplicative bias difference between these branches with realistic morphologies and parametric profiles on the order of (6.9 ± 0.6)×10−3 for a realistic magnitude-Sérsic index distribution. Moreover, we find clear detection bias differences between full image scenes simulated with parametric and realistic galaxies, leading to a bias difference of (4.0 ± 0.9)×10−3 independent of the shape measurement method. This makes complex morphology relevant for stage IV weak lensing surveys, exceeding the full error budget of the Euclid Wide Survey (Δμ1, 2 < 2 × 103).