Closed gvdsteene closed 11 months ago
You generally shouldn't, but it is likely to occur when using velocities. The equation for moment-2 is:
$\frac{\Sigma (y -
To add to this, a mixture of negative and positive intensities (i.e., noise) can also give you a negative variance because it's an intensity-weighted variance. When this occurs it's a strong indication that you should mask out the noise before calculating the moment maps (and the moment 1 centroid is also likely to be very noisy).
There's an example of this issue in one of our spectral-cube tutorials here: https://mybinder.org/v2/gh/radio-astro-tools/tutorials/master?filepath=casa_to_spectralcube_guide%2Fcasa_to_spectralcube_guide.ipynb
A moment 2 map is a variance map. How come that I obtain negative values ?