Version 2 NSF network data:
A few errors in Version 2 data have been noted after the data had been published. These errors are listed below:
- Increased uncertainty of pre-Volume 1 data South Pole:
Pre-Volume 1 data from South Pole have an uncertainty of ±20% and should not be used.
- Incorrect solar zenith angle at McMurdo and South Pole:
The field "SZA at start" of "model_log"-databases includes the solar zenith angle at the end rather than at the start of a spectrum. Both data from McMurdo and the South Pole are affected. The problem has not been corrected.
- Incorrect solar zenith angle at Palmer:
Solar zenith angles reported in all databases and full-resolution spectra are only accurate to within ±0.3º. Data prior to Volume 14 are affected. The problem has not been corrected.
- Problems at start of spectra:
After wavelength correction, some spectra start at a wavelength λstart, where λstart > 280 nm. All published spectra were resampled to a uniform wavelength grid, starting at 280 nm. If a wavelength-corrected spectrum starts at λstart > 280 nm, irradiance values for wavelengths between 280 nm and λstart are meaningless in published spectra, and should be ignored. Only data from McMurdo and the South Pole prior to Volume 14 are affected. The problem has not been corrected.
- Problems in modeled clear-sky spectra:
Model calculations may include distorted spectra that are not labeled as such. The distortion is due to numerical instabilities of the model under certain conditions. Mostly data from McMurdo and the South Pole prior to Volume 14, which were processed with a single-precision model version, are affected. The problem has not been corrected.
- Incorrect time format:
Until 2/7/05, some South Pole data had an extra "0" in the "second" specification of the date/time field. For example, instead of "7/4/99 10:03:00", date/time was "7/4/99 10:03:000". Fields labeled "Time at 310 nm", "Time at 360 nm", "Time at 500 nm" of the following data bases of all volumes "x" were affected:
The problem has been corrected.
- Shading of Volume 15 Palmer data:
Starting in April 2005, a building ("Terra Lab") was constructed next to the SUV-100 spectroradiometer at Palmer Station.
The new facility obstructed the southerly horizon seen from the instrumentís collector after May 1, and also affected solar measurements.
Data were corrected for the shading effect and some data were flagged, but data measured between September 2005 and April 2006 should
be treated with caution. The spectroradiometer was moved to Terra Lab in May 2006.
- Use of incorrect ozone profiles in Volumes 14 and 15 data Summit:
An ozone profile with distorted altitude assignment was used in Volume 15 data from Summit for the
period 2-August-2005 - 14-October-2005. Volume 14 data were originally processed with the standard
AFGL profile for subarctic summer. Data of both volumes were reprocessed with profiles measured by NOAA
at Summit. The ftp-site was updated on 13-December 2007.
The problem has been corrected.
- Incorrect time assignment in Volume 1 data:
The time assignment of Volume 1 data is incorrect by up to 3 minutes. However, the bias is wavelength dependent and the times of some wavelengths may be correct.
The problem has not been corrected.
For more information contact us.