Calibration of a Dobson Spectrophotometer
3.1 Relative Calibration
Calibration of a Dobson spectrophotometer on a relative scale involves careful execution of the operations described below.
3.1.1 Adjustment of the Optics
All optical components, i.e., lenses, prisms, slits, etc., must be in proper adjustment. Experience has indicated that optical components of Dobson spectrophotometers often exhibit varying degrees of maladjustment due either to errors committed during instrument manufacture, or to faults that develop in the instrument with time. Faults that have been observed in the past include the following:
(a) distortion of the instrument optic axis;
(b) crown glass lens, rather than a quartz lens, used in the path of the S2 and S3 wavelength beams;
(c) slit widths not set according to tolerance specifications;
(d) slit widths not set sufficiently parallel to each other;
(e) optical wedges loose in their holders;
(f) focussing lenses mounted in reverse positions;
(g) S2 and S3 optical beams partially obstructed by mechanical components of the instrument;
(h) mirror position influenced by temperature (with hysteresis) due to too close fit of mirror in its mount.
Details concerning adjustment of the spectrophotometer optics are provided elsewhere (Dobson, 1957b). Only a skilled technician should be permitted to perform optical alignment adjustments. For ozone observers needing assistance with such work, help is available through the WMO from experts in several countries.
3.1.2 Q Calibration
Ozone observations are made on wavelength pairs designated as A, B, C, C', and D. By rotating two levers, Q1 and Q2, on the instrument, the desired wavelengths are selected for observations. The Q (or wavelength) settings vary with temperature owing to the change of refractive index of quartz with temperature and the expansion or contraction of the metal of the instrument. (The Q settings are also pressure-dependent since the refractive index of quartz in air varies with air pressure (see Appendix B)). It is necessary then, to establish a correct table of Q settings vs. temperature for the spectrophotometer. The procedure for establishing the Table of Settings of Q is outlined in Appendix A.
3.1.3 Optical Wedge Calibration
The relative transmission along the spectrophotometer optical wedge must be known accurately in order to estimate, to a high degree of precision, the relative intensity of the two wavelength beams on which observations are made. Calibration of the optical wedge involves the determination of wedge density tables which relate instrument R-dial reading: to logarithmic ratios of pair wavelength beam intensities, or simply, tables of R vs. log (I/I') + K where K is an instrumental constant. The procedures to be followed in performing optical wedge calibrations by the two-lamp method, and in reducing the experimental data, are described in Appendix C.
3.2 Absolute Calibration
A spectrophotometer calibrated accurately on a relative scale cannot be used for measuring total ozone amounts correctly since it is necessary to know the "extra-terrestrial constant" for the instrument, i.e., the value of log (I0/I0') + K which would be found for each Dobson instrument wavelength pair if the measurements were made on sunlight outside the earth's atmosphere. Fixation of the "extra-terrestrial constant" to the optical wedge density table for A, B, C, and D wavelengths constitutes calibrating the spectrophotometer on an absolute scale. The modified wedge density tables are then called "N tables."* Values of N are simply values of log (I0/I0') - log (I/I').
*[For convenience, N values recorded in tables are often defined by 100[log (I0/I0') - log (I/I')].]
An absolute calibration of a spectrophotometer may be effected in three ways: by intercomparing two instruments directly; by making special types of direct sun observations (see Section 6.4.1); and by using lamps to transfer the calibration from a correctly calibrated instrument to one that is uncalibrated. A description of the three methods is provided in Appendix D.
3.3 Maintaining a Spectrophotometer in Calibration
Each calibrated spectrophotometer should have associated with it a set of reference tables entitled Table of Settings of Q, Standard Lamp Tests, Wedge Calibration Tests, and Sensitivity Tests. Whenever routine spectrophotometer tests described in Section 4 are performed, the experimental data obtained should be compared with the data presented in the reference tables. If the two sets of data are in agreement within prescribed tolerance limits, the implication is that the calibration level of the spectrophotometer has remained unchanged. If, however, some of the experimental and reference data are discordant, the interpretation should be that a change has occurred in either the spectral characteristics of the instrument, or in the testing apparatus. Steps must then be taken to discern the source of difficulty and correct it.
It is useful to maintain running plots of the routine spectrophotometer tests data since changes in instrument calibrations are readily discernable from such plots.
Remedial actions to be taken, when needed, are described in Section 4. Under no circumstances should the optics of the spectrophotometer be tampered with or major recalibrations undertaken by anyone other than a qualified technician.
Return to Table of Contents
Forward to Section 4. Routine Spectrophotometer Tests and Maintenance