SGP/SWS/C1 - Instrument offline for annual calibration

SWS instrument was taken offline for an annual calibration.  

Comparision of the December 19, 2007 annual calibration with the previous annual 
calibration done on November 8, 2006 shows very little change in the instrument sensitivity over 
2007.  From our analysis, there is effectively no change in the silicon spectrometer 
sensitivity from 400 to 981 nanometers.  The InGaAs spectrometer shows a slight decrease in 
sensitivity of 1 to 2% over the wavelength range of 982 to 2150 nanometers, which is within 
our expected error range.  Characterization of the change in sensitivity from 350 to 400 
nanometers over the year is a bit more difficult to determine and maybe greater than our 
expected error of 2%.  The science team at the University of Colorado is currently 
working on this problem.

• time
• base_time
• shutter_closed
• time_offset
• Wavelength of spectrum data(wavelength)
• lon
• alt
• Calibrated spectrum(zen_spec_calib)
• lat

• SI_spect_temp
• pc104_neg12v
• internal_temp
• ps2_12v
• pc104_pos12v
• alt
• lon
• time_offset
• pc104_5v
• base_time
• time
• ps2_5v
• lat
• InGaAs_spect_temp

SGP/SWS/C1 - Temperature instability of InGaAs detector

During this time period the SWS experienced a problem with the temperature control InGaAs 
detector (for wavelength from 900-2200 nm).  The problem was initially infrequent but 
gradually increased in frequency and duration until it was chronic and nearly continual.  
All data from the InGaAs during affected period is bad and unusable, but data from the Si 
detector (300 nm to 900 nm) is unaffected.  Affected periods are easily discerned from the 
InGaAs_spect_temp (InGaAs spectrometer temperature) field in the sgpswsauxC1.b1 
datastream. When the spectrometer is behaving properly this temperature will be held steady to a 
value between -9.5 and -10.7 C.  When the temperature behaves erratically the InGaAs data 
is unusable. However the Si data is unaffected.

The spectrometer connectors were disconnected and all contacts cleaned. The problem has 
not reappeared.

• Calibrated spectrum(zen_spec_calib)

• InGaAs_spect_temp

SGP/SWS/C1 - Contamination by scattered light on optics.

It has been observed that the zenith radiance measurement is susceptible to contamination 
from light scattered off of the internal optics. This problem is most serious when 
measuring clear sky radiance in the presence of direct sun.  Under this worst case condition, 
it affects short wavelengths (less than 500 nm) at a few percent level, moderate 
wavelengths (500-800nm) between a few percent and 10-20%, and longer wavelengths up to 80% or more.

Chui and Marshak have demonstrated some success at identifying isolated affected data 
using the TSI to find cases of scattered clouds.  In general though, one must treat the 
zenith radiance data with considerable caution whenever direct solar radiance is likely to be 
present.  It is anticipated that the SWS zenith radiance for overcast skies is not 
significantly affected.

The instrument has been renovated by adding a sun shade to prevent direct sunlight from 
striking the optics.  An additional set of baffles further restricts the field of view, 
further reducing the potential for even off-axis bright clouds to affect the zenith 
measurement.  Data taken after June 2009 is considered to be free of contamination.

• Calibrated spectrum(zen_spec_calib)