Yep, doing the math helps here, to some degree: the convolution of a Gaussian with a Gaussian is still a Gaussian (without that, there wouldn’t be the central limit theorem, i.e. the reason why ...

Gaussian fitting via maximum-likelihood estimation: I determined a maximum-likelihood estimate for a symmetric, two-dimensional Gaussian function as described in ref. 10 .

An analytical solution is shown to exist for the specific case of the temporal broadening of a Gaussian-shaped pulse by a thin scattering screen. The solution is applied to multi-frequency ...