How to Optimize?



We can afford to follow up all peak above 5 sigma ourselves at 90 GHz
Confirm detection and refine mass estimate

Crucial thing is how many clusters of given mass missed
False detections are OK in intiial 30 GHz survey