A robust assessment of the local anisotropy of the Hubble constant

Magnitude predictions of $\Lambda$CDM, as parametrized by the Planck collaboration, are not consistent with the supernova data of the whole Pantheon+ sample even when, in order to take into account the uncertainty about its value, the Hubble constant is adjusted. This is a likely consequence of the increase of the number of low-redshift supernovae in the Pantheon+ sample, with respect to previous such samples. Indeed, when supernovae at redshifts below 0.035 are ignored, with H0=73.4 km/s/Mpc, $\Lambda$CDM predictions become consistent with Pantheon+ data. Interestingly, this is also the case when subsets of low-redshift supernovae roughly centered on the direction of the CMB dipole are considered, together with high-redshift ones, at least when CMB and peculiar velocities corrections are taken into account for the redshifts. These results seem robust, since they are also obtained with a simple, single-parameter tired-light model.