Front Page
 Scientific Ops
  » Latest
  » Science
  » Calibration
  » Operations
  » Past Issues
Issue 1: 10th September 2001
A Note from the Editor and Contents

Welcome to the first issue of the CDS Newsletter. The goal of this Newsletter is to inform the CDS user community of
  • current CDS science topics
  • developments in CDS data analysis
  • instrumental matters
  • operational issues
The Newsletter will be updated monthly and initially will cover one main topic per issue. We invite your contributions on CDS-related matters: data analysis, science results, instrument calibration, software and questions on these topics. Your responses will influence the content of future issues. Please send newsletter inputs and correspondence to:

This Month's Topics:

Measuring irradiance with CDS, and implications for the CDS calibration
William Thompson

The current calibration of the CDS Normal Incidence spectrometer is tied to two sounding rocket measurements made in 1997. A description of the current calibration used for CDS/NIS can be found at ../software/calibration.shtml

Currently, the same calibration is used for both the data taken before SOHO's accident in the summer of 1998, and for the data taken after the accident. Coordinated observations with sounding rockets have also been taken after the accident, but the data is not yet available for comparison.

The second-order calibration is not embedded within the software. For the second order lines of He II and Si XI at 304 A, I've been using an efficiency 25.9 times less than the corresponding first-order calibration, i.e. about 1.64e-5.

One area where the calibration is important is in measuring the solar irradiance. Comparing the full disk irradiance measurements made by CDS against observations made by other instruments tests both the absolute calibration of CDS, and how well the calibration was maintained over that period.

Irradiance measurements have been made with CDS since March 1997, using the monthly study USUN_6. In order to make these measurements, 69 individual rasters are taken to cover the entire Sun. The entire spectral range in both NIS channels is brought down, so that the irradiance can be measured for every line observable by CDS. However, in order to be able to bring down this much data in a reasonable amount of time, a large amount of compression is needed. Two steps are taken to compress the data. One is to move the 4 arcsecond slit by 24 arcseconds between exposures, so that the Sun is undersampled by a factor of six. Although this means that the entire Sun is not observed, tests have shown that a representative enough sample of the Sun is taken, so that the result should be correct to within a few percent.

The other method used to compress the data is to sum pixels along the slit. The first few observations added up all the pixels along the slit into a single number. Since then, pixels have been summed vertically in groups of four into larger pixels, so that low resolution images of the Sun can be built up, as in the following figure.

Comparison with other instruments
There are two other instruments on SOHO which are also capable of measuring the solar irradiance in the same wavelength regime as CDS. One is the Solar EUV Monitor (SEM), which is part of the CELIAS package. SEM measures the integrated irradiance between 260-340 A. It takes three separate CDS components to match the SEM bandpass:
  1. First-order NIS-1 from 310-340 A.
  2. Second-order NIS-2 He II and Si XI lines at 304 A.
  3. Synthetic spectrum from DEM analysis for 260-300 A, based on both NIS-1 and NIS-2 lines.
Another instrument which also observes in this wavelength regime is EIT. Each EIT bandpass is actually a combination of several emission lines. However, Jeff Newmark has developed software to apply DEM analysis to EIT data, and generate synthetic spectra over the SEM bandpass. Thus, all three instruments can be compared directly against each other. The following figure shows a comparison for the period up until the summer of 1998.

Because the SEM instrument does not have a flat response over its bandpass, there is a certain level of interpretation which needs to be applied to the SEM data to convert them to physical units. In the above plot, the black line represents the results returned by the standard SEM software, using a constant reference spectrum. The colored dashed lines, on the other hand, represent the SEM results when the spectra of either CDS or EIT are used.

The CDS irradiance is slightly lower than that measured by either SEM or EIT. When the CDS data is compared against the SEM results as modified by the CDS spectral response (the red dashed curve above), the CDS/SEM varies from about 95% at the beginning to 85% just before the accident. There is a slight downward slope to the ratio, which may indicate a small (10%) uncorrected loss of sensitivity in CDS.

Since the uncertainty in the CDS irradiance values is about 25%, the CDS measurements agree with the SEM and EIT values within the error bars.

Post-recovery changes
In the summer of 1998, attitude control was temporarily lost with the SOHO spacecraft, and for several months the CDS instrument was subjected to extremely high temperatures. One question is whether the calibration of the instrument used for the first two years of the mission can also be applied to the data taken after control of the spacecraft was recovered. There are three regular sets of CDS observations which can be used to monitor the instrument response: SYNOP_F, NISAT, and NIMCP. A difficult thing to do is to distinguish changes in the instrument from those which are solar in origin. In particular, the Sun was much more active after the accident than it was before. This seems to be influencing the SYNOP_F data, where the He 584 line is slightly brighter after the accident than before. However, the NISAT and NIMCP observations are specifically made in quiet Sun areas; in those data, there is no perceptible change in the 584 intensity.

On the other hand, there does appear to be a change in the sensitivity for the second-order He II line at 304 A. This is most evident in the ratio of the two helium lines at 304 and 584 A, as shown below. From these data, one can infer that the He II intensity should be increased by about 17% (1.167) after the accident.

An even larger loss of sensitivity can be seen in the Mg 368 A intensity, as shown below. These data suggest that the NIS-1 intensity needs to be increased by a factor of about 45% (1.447) after the accident.

When these "fudge factors" are applied to the post-recovery data, one gets the following comparison between the CDS, SEM, and EIT irradiances.

When the ratio of the CDS irradiance (solid red) to the adjusted SEM data (dashed red) is examined, it appears that there is a jump up just after the accident to almost 100% of the SEM signal, followed by a slow drop back down to 85% again.

The "fudge factors" discussed above have not yet been implemented in the CDS calibration software. The plan is to wait for the sounding rocket analysis to be finished before making any changes in the CDS software tree.

A more complete analysis of this data is being prepared for the Bern calibration meeting.

From the CDS Operations Management Team in the Space Science & Technology Department at CCLRC Rutherford Appleton Laboratory
Site maintained by John Rainnie.
Last revised on Tuesday (22/Jan/2019) at 15:07.