Beginning with August 19, 2019 at 13:00 UTC, noise in the form of random horizontal lines over the Eastern Pacific began appearing, with most of the problems appearing each following day in the 12:00, 12:30, and 13:00 half-hours, with at least one instance of 13:30.  This was traced to noise in the GOES-W Tb’s (as the result of the faulty sensor cooling system), which made it past the filtering at CPC, where the Merged Global 4-km IR dataset is assembled.  The Early and Late datasets contain these errors, which will be dealt with in the Final computation.