The calibrations used in the determination of this data set are a combination of those obtained during PEM-
Tropics B and TRACE-P. It was found that there was a leak in the calibration assembly inside of the pylon. 
Normally this space is flushed with N2. During flight, however, ambient air flushed inside of the assembly. 
This leak introduces a problem in that the amount of light exiting the calibration assembly has been 
attenuated by absorptions from O2 and H2O inside the assembly. We have recently measured the O2 
absorption cross-section at 184.9 nm using the same path lengths as found in the calibration assembly. 
Using these values we were able to back correct calibrations for the attenuation occurring inside the 
assembly and obtain a better estimate of the OH produced during calibration. This approach produced an 
altitude dependent calibration curve with 30% higher sea level values than those obtained during PEM-
Tropics B. These calibration values are expected to (and have) remain constant over time. Because of the 
larger uncertainties introduced by correcting the TRACE-P calibrations, it was decided to average the sea 
level values from both PEM-Tropics B and TRACE-P and reduce the TRACE-P values by this amount. 
The additional reported error (+/- 60% as opposed to +/- 40% which we typically quote) reflects the 
additional uncertainty in the use of the average of the calibrations from the two studies.