2022A&A...668A..50S

Context. Distinguishing the signal from young gas-rich circumstellar disks from the stellar signal in near-infrared (NIR) light is a difficult task. Multiple techniques have been developed over the years of which angular differential imaging (ADI) and polarimetric differential imaging (PDI) have been most successful. However, both techniques cope with drawbacks such as self-subtraction. To address these drawbacks, we explore iterative ADI (IADI) techniques to increase signal throughput in total intensity observations. Aims. The aim of this work is to explore the effectiveness of IADI in recovering the self-subtracted regions of disks by applying ADI techniques iteratively. Methods. IADI works by feeding back all positive signal of the result from standard ADI over multiple iterations. To determine the effectiveness of IADI, a model of a disk image is made and post-processed with IADI. We explored two versions of IADI, classical IADI, which uses the median of the data set to reconstruct the point spread function (PSF), and PCA-IADI, which uses principal component analysis to model the PSF. In addition, we explored masking based on polarimetric images and a signal threshold for feeding back signal. Results. Asymmetries are a very important factor in recovering the disk because these lead to less overlap of the disk in the data set. In some cases, we were able to recover a factor ∼75 more flux with IADI than with ADI. The Procrustes distance is used to quantify the impact of the algorithm on the scattering phase function. Depending on the level of noise and the ratio between the stellar signal and disk signal, the phase function can be recovered a factor 6.4 in Procrustes distance better than standard ADI. Amplification and smearing of noise over the image due to many iterations did occur. By using binary masks and a dynamic threshold this feedback was mitigated, but it is still a problem in the final pipeline. Finally, observations of protoplanetary disks made with VLT/SPHERE were processed with IADI giving rise to very promising results. Conclusions. While IADI has problems with low-signal-to-noise-ratio (S/N) observations due to noise amplification and star reconstruction, higher S/N observations show promising results with respect to standard ADI.