Title:Can the sound horizon-free measurement of $H_0$ constrain early new physics?

Abstract:The sound horizon-independent $H_0$ extracted by using galaxy clustering surveys data through, e.g., EFTofLSS or ShapeFit analyses, is considered to have the potential to constrain the early new physics responsible for solving the Hubble tension. Recent observations, e.g. DESI, have shown that the sound horizon-independent measurement of $H_0$ is consistent with $\Lambda$CDM. In this work, we clarify some potential misuses and misinterpretations in these analyses. On the one hand, imposing some prior from other cosmological probes is often used to strengthen the constraints on the results, however, these priors are usually derived using the assumption of $\Lambda$CDM, it is not suitable to apply these so-called $\Lambda$CDM priors (e.g., the $n_s$ prior from CMB), which would bias the results, to early new physics because these early new physics are usually accompanied by shifts of the $\Lambda$CDM parameters. On the other hand, the constraints on $H_0$ in the sound horizon-independent EFTofLSS analysis arise from not only the shape of the power spectrum ($k_\text{eq}$-based $H_0$), but also the overall amplitude (when combined with CMB lensing observations) and the relative amplitudes of the BAO wiggles, thus besides $k_\text{eq}$ other information may also play a role in constraining $H_0$. We also make forecasts for an Euclid-like survey, which suggest that ongoing observations will also have difficulty ruling out early new physics.