Maybe they were not using ata at all. We don't know.
What instruments did they use? They only had a 3% instrument error which is very good.
Most bourdon tube or bellows instruments found in the panel are ~5%.
What is a fact is that the data is not converted to standard conditions.