Spectral and angular differential imaging with SPHERE/IFS. Assessing the performance of various PCA-based approaches to PSF subtraction

Angular differential imaging (ADI) and spectral differential imaging (SDI) are commonly used for direct detection and characterisation of young, Jovian exoplanets in datasets obtained with the SPHERE/IFS instrument. We compare the performance of ADI, SDI, and three combinations of ADI and SDI to find which technique achieves the highest signal-to-noise ratio (S/N), and we analyse their performance as functions of integration time, field rotation, and wavelength range. We analyse SPHERE/IFS observations of three known exoplanets, namely Beta Pictoris b, 51 Eridani b, and HR 8799 e, with five differential imaging techniques. We split the datasets into subsets to vary each parameter before the data are processed with each technique. The differential imaging techniques are applied using principal component analysis (PCA). The tests show that a combination of SDI and ADI consistently achieves better results than ADI alone, and using SDI and ADI simultaneously (combined differential imaging; CODI) achieved the best results. The integration time test shows that targets with a separation larger than 0.24 arcsec observed with an integration time of more than 10$^3$s were photon-noise limited. Field rotation shows a strong correlation with S/N for field rotations up to 1 full width at half maximum (FWHM), after which no significant increase in S/N with field rotation is observed. Wavelength range variation shows a general increase in S/N for broader wavelength ranges, but no clear correlation is seen. Spectral information is essential to boost S/N compared to regular ADI. Our results suggest that CODI should be the preferred processing technique to search for new exoplanets with SPHERE/IFS. To optimise direct-imaging observations, the field rotation should exceed 1 FWHM to detect exoplanets at small separations.