Verify noise levels and units

[evavagiakis]

We need to ensure that the generated gaussian noise is producing what we think it is for a given noise level in an ACT map

[evavagiakis]

I've looked into this and it seems like the noise function is the same as the white noise function in https://github.com/jeffmcm1977/CMBAnalysis_SummerSchool/blob/master/CMB_School_Part_03.ipynb, and from reading more about noise on these scales, including only white noise seems reasonable. Our original level of 30 uK$\cdot$arcmin may be a bit high, because most of the map used for the recent ACT cluster paper is < 30 uK$\cdot$arcmin. However, using a more realistic 20 uK$\cdot$arcmin still creates a submap in which you cannot pick out the cluster by eye. This might be fine, but I think this requires one more sanity check, which I've described in #160

[samueldmcdermott]

I think modern experiments aim to have noise levels between 1-10 uK*arcmin