LISA pathfinder appreciably constrains collapse models

LISA Pathfinder's measurement of a relative acceleration noise between two free-falling test masses with a square root of the power spectral density of $5.2 \pm 0.1 \mbox{ fm s}^{-2}/\sqrt{\rm{Hz}}$ appreciably constrains collapse models. In particular, we bound the localization rate parameter, $\lambda_{\rm CSL}$, in the continuous spontaneous localization model (CSL) to be at most $\left( 2.96 \pm 0.12 \right) \times 10^{-8} \mbox{ s}^{-1}$. Moreover, we bound the regularization scale, $\sigma_{\rm DP}$, used in the Diosi-Penrose (DP) model to be at least $40.1 \pm 0.5 \mbox{ fm}$. These bounds significantly constrain the validity of these models. In particular: (i) a lower bound of $2.2 \times 10^{-8\pm2} \mbox{s}^{-1}$ for $\lambda_{\rm CSL}$ has been proposed in (although a lower bound of about $10^{-17} \mbox{s}^{-1}$ is sometimes used), in order for the collapse noise to be substantial enough to explain the phenomenology of quantum measurement, and (ii) 40 fm is larger than the size of any nucleus, while the regularization scale has been proposed to be the size of the nucleus.