Atmospheric Absorption in Laboratory Measurements: A Comparison Between modtran3 and Measurement

Errors can occur in laboratory measurements when the response of a bandpass-filtered radiometer extends into an atmospheric absorption region. Atmospheric models, such as modtran3, can be valuable tools that permit the optical measurements in these regions to be accurately analyzed, provided the models themselves are accurate. Comparisons of modtran3-predicted and laboratory-measured atmospheric transmittance have been made to help establish the validity of modtran3 for use in modeling short-path-length (a few meters or less), low-resolution optical effects over a wavelength range of 700-5700 nm. Comparisons include percentage differences, range of differences, and band-averaged differences. Good agreement is shown for all absorption bands except for the CO(2) band near 4300 nm. In general, the band-average differences are less than ~0.75% for all water bands. For the CO(2) band the modtran3 default CO(2) level is found to be inappropriate, and 610 parts per million in volume is found to give much better agreement with measurement.