Atmospheric sanity check
August 28, 2020 - Reijo Keskitalo
In this post we perform a quick comparison of the atmospheric noise in the reference design simulations and SPT 2013-2015 maps shown in arXiv:1707.09353.
Here is the SPT temperature map (top) and null map showing the noise (bottom):
We use the MFL2 (150GHz) South Pole LAT simulation of the atmospheric component, observed with a single optics tube over 10 independent scans. Each scan covers the 1000 square degree patch in 101 elevation steps in 10 hours. Each azimuth sweep is filtered with a 15th order polynomial (SPT uses 5th order). The filtered signal is binned into a single IQU map.
Top row here shows the CMB and the atmospheric component in the reference design simulation after filtering but without any scaling:
The bottom row shows two different versions of the scaled atmospheric fluctuation. Bottom left is scaled to match the telescope time per square degree in the SPT map (100 hours vs. 9087 hours, 1000 sq.deg vs. 500 sq.deg). Bottom right is scaled to match the number of independent visits (10 vs. 3491). While the reference simulations contain slightly more atmospheric noise, the order of magnitude is in good agreement. Small differences are expected due to the atmospheric realization being simulated and due to there being no data cuts applied in the simulation.