One picture is worth 1000 words:
Every photon in the image was simulated through the ab initio physics in PhoSim using its Monte Carlo methods. We performed 3 simulations using the Rubin (LSST) observatory PhoSim configuration for 15 second exposures using the y, i, and u filters. This 3-color image was produced by combining three images in the RGB channels corresponding with the y, i, and u filters (respectively). The images were corrected for basic calibration (gain, offset) and the overall levels were set to make the color of stars close to the Sun. The overall intensity is scaled using an inverse hyperbolic sine function (Lupton, Gunn, & Szalay 1999). We just show the top portion of a single chip, so the field covers roughly 13 by 6 arcminutes. The background was not subtracted, but some under-subtraction was performed in the blue and red channels to make the overall background have a grey color. No other corrections or calibrations were performed--intentionally in order to highlight the details. There are many physics details inside of PhoSim that are not directly observable without performing a detailed analysis on the image with sophisticated algorithms. On the other hand, there are many details that are visible by to the human eye, so here we show some of the detailed observable features in the image:
Objects: Hundreds of stars and hundreds of galaxies are visible in this image. The depth of this image corresponds to a magnitude of ~24, so at this depth you are largely seeing the full set of stars in our Galaxy for this particular part of the sky and are peering deep into the Universe and collecting a large number of distant galaxieswith this exposure. Thus, most of the brighter objects are likely stars, and the vast majority of the fainter objects are galaxies.
Colors: The observed colors of the stars and galaxies are a combination of the intrinsic true colors of those objects related to the stellar temperatures, metallicity, redshift,and foreground dust absorption, but also the distortion due to the effect of the atmosphere and telescope. In particular, the blue channel is affected strongly by Rayleigh scattering in the atmosphere, and the red channel is affected strongly by both water absorption in the atmosphere and the limited ability of red photons to create a photo-electron in the limited depth of the Silicon sensor. Note that the bands are more widely separatedin the EM spectrum than the human eye’s cones, meaning that this image is more colorful than the human eye would seepeering through a telescope.
Blurriness: The blurriness of the image (i.e. point-spread-function or PSF) is due primarily toatmospheric turbulence, the diffusion of electrons in the sensor, the slightly aberrated telescope design, and the mechanical and thermal distortions of the surfaces of the optics from their ideal shape. The sizes of most galaxies in the image are roughly an equal combination of their actual true size and this extra blurring. So this means everything looks more like a blurry blob than it is intrinsically. Yet the blurring isn’t large enough to completely erase some of the elliptical shapes of the galaxies and their bulge and disk components in some of the larger galaxies.
Positions: The positions of all the objects are slightly distorted from where they truly are on the sky (i.e. astrometry). The overall distortion cannot be determined from just looking at this single image, but you can notice that the red, green, and blue channels do not match perfectly and are off by different amounts in different parts of the image (i.e. differential astrometry). Most differential astrometry is due to a combination of chromatic dispersion in the atmosphere, imperfect averaging of the turbulent refraction of photons (image motion), non-isotropy of the field-lines in the Silicon due to doping impurity variations or other electric field variations, and higher frequency mechanical and thermal perturbations on the optical surfaces.
Bleeding & Diffraction: Several of the bright stars have obvious horizontal bleed trails. They have different colors simply depending on when there are enough photon-electrons of a particular exposure in a particular band to fill up the potential well of a pixel. There are also a pair diffraction spikes that are tilted relative to the bleed trail, due to the diffraction of the support structure of the telescope.
Red Edge: The background contains a red glow towards the outside. This is because near the edge of the image, electrons are kicked preferentially out of the active imaging region because of a slightly distorted electric field due to the guard rings that has a small lateral component. This effect is achromatic, but the red photons are more likely to convert deeper in the Silicon of the sensor making them less likely to be removed from the active imaging region because the net displacement is shorter since they have a shorter distance to travel to the readout. Then after making this color image the edge glows red.
Contamination: Dust and debris is on all optical surfaces. This mostly results in a loss of light in large circular patterns called dust rings (due to the optical design), but this is most visible on this image for light scattered or absorbed by dust on the surface of the detector. Several green circles can be seen on this image which are not galaxies, but lost light from debris on the sensor surface. It appears green not because it is a particularly chromatic effect, but because there is more background light relative to astrophysical sources in the blue (zodiacal light) and red (airglow). The fact that we subtracted light from the red and blue channels making the mean background grey then makes these patterns appear green. There is one prominent feature about 2/3 of the way down near the right of the image.
Structured Background: A faint hatched pattern appears blue under that background. This is from the laser annealing treatment on the back-side of the sensor. This results in a field-free region that is not uniform that causes electrons to be lost if the photon converts in the top of the device. Blue light is more likely to do that, hence resulting in a blue brick-likepattern.
Cosmic Rays: Cosmic rays result in mostly structured streaks due to ionization trail they leavealong the track of the particle, but it will only appear in one of the three images since they are taken at different moments in time. There are several cosmic ray tracks of various colors in this image. The most visible one is a blue streak right next to the bright orange galaxy 2/3 of the way down near the middle.