2021 occultations and transits of Linus orbiting (22) Kalliope: I. Polygonal and `cliptracing' algorithm

The satellite Linus orbiting the main-belt asteroid (22) Kalliope exhibited occultation and transit events in late 2021. A photometric campaign was organized and observations were taken by the TRAPPIST-South, SPECULOOS-Artemis, OWL-Net, and BOAO telescopes, with the goal to constrain models of this system. Our dynamical model is complex, with multipoles (up to the order $\ell = 2$), internal tides, and external tides. The model was constrained by astrometry (spanning 2001--2021), occultations, adaptive-optics imaging, calibrated photometry, as well as relative photometry. Our photometric model was substantially improved. A new precise (${<}\,0.1\,{\rm mmag}$) light curve algorithm was implemented, based on polygon intersections, which are computed exactly -- by including partial eclipses and partial visibility of polygons. Moreover, we implemented a `cliptracing' algorithm, based again on polygon intersections, in which partial contributions to individual pixels are computed exactly. Both synthetic light curves and synthetic images are then very smooth. Based on our combined solution, we confirmed the size of Linus, $(28\pm 1)\,{\rm km}$. However, this solution exhibits some tension between the light curves and the PISCO speckle-interferometry dataset. In most solutions, Linus is darker than Kalliope, with the albedos $A_{\rm w} = 0.40$ vs. $0.44$. This is confirmed on deconvolved images. A~detailed revision of astrometric data allowed us to revise also the $J_2 \equiv -C_{20}$ value of Kalliope. Most importantly, a~homogeneous body is excluded. For a differentiated body, two solutions exist: low-oblateness ($C_{20} \simeq -0.12$), with a~spherical iron core, and alternatively, high-oblateness ($C_{20} \simeq -0.22$) with an elongated iron core. These correspond to the low- and high-energy collisions, respectively, studied by means of SPH simulations in our previous work.