Prospects of using closure traces directly for imaging in Very Long Baseline Interferometry

The reconstruction of the polarization of a source in radio interferometry is a challenging calibration problem since the reconstruction strongly depends on the gains and leakages that need to be inferred along with the image. This is particularly true for the Event Horizon Telescope (EHT) due to its small number of antennas, small signal-to-noise ratio and large gain corruptions. To recover linear polarization, one either has to infer the leakages and gains together with the image structure, or rely completely on calibration independent closure quantities. While the first approach has been explored in Very Long Baseline Interferometry (VLBI) for a long time, the later one has been less studied for polarimetry. Closure traces are a recently proposed concept of closure quantities that, in contrast to closure phases and closure amplitudes, are independent against both gains and leakages and carry the relevant information about the polarization of the source. Here we explore, how closure traces could be directly fitted to create an image and point out an imaging pipeline that succeeds in the direct imaging from closure traces. Since closure traces have a number of inherent degeneracies, multiple local image modes that can fit the data are detected. Therefore, a multiobjective imaging technique is needed to correctly sample this multimodality. Closure traces are not constraining enough for the EHT configuration in 2017 to recover an image directly, mainly due to the small number of antennas. For planned successors of the EHT however (with a significantly larger number of antennas), this option becomes feasible and performs competitive to the imaging with residual leakages.