It is common to model a deterministic response function, such as the output of a computer experiment, as a Gaussian process with a Mat\'ern covariance kernel. The smoothness parameter of a Mat\'ern kernel determines many important properties of the model in the large data limit, such as the rate of convergence of the conditional mean to the response function. We prove that the maximum likelihood and cross-validation estimates of the smoothness parameter cannot asymptotically undersmooth the truth when the data are obtained on a fixed bounded subset of $\mathbb{R}^d$. That is, if the data-generating response function has Sobolev smoothness $\nu_0 + d/2$, then the smoothness parameter estimates cannot remain below $\nu_0$ as more data are obtained. These results are based on a general theorem, proved using reproducing kernel Hilbert space techniques, about sets of values the parameter estimates cannot take and approximation theory in Sobolev spaces.

Asymptotic Bounds for Smoothness Parameter Estimates in Gaussian Process Regression