Last week we selected patches based on Foregrounds. Now select patches based on level of foregrounds as measured at 555 GHz.
Foregrounds separation still done over the full sky (*different than ground-based instruments*)
Model 98, small patch analysis - only with r=0; don't have r=0.003.
Kris asks - what limits the separation? Foregrounds or noise? Perhaps do noise-free and then separately foreground free? SH: Flauger had found r=2e-5 with Fisher analysis including simple foregrounds. So noise seems to not be the issue.