NSA/SKYRAD/C1 - Frosting/Condensation
The down_long_hemisp1 varied up to 40 W/m^2 higher than down_long_hemisp2 during the affected time periods. Cold temperatures reported by the IRT suggest the issue is with PIR1. Icing or frost on PIR1 or within the ventilator may be to blame but ops reports can't substantiate this. Data from PIR1 should be avoided during these suspect time periods.
Use down_long_hemisp2 data during these time periods