6.5.2 Description of the data

Total solar irradiance, the integrated solar flux over the entire solar spectrum, has been measured from space for over a decade. The direct space observations of total solar irradiance began in late November of 1978 by the Earth Radiation Budget (ERB) experiment on board the Nimbus-7 satellite that was terminated in January of 1993 (Kyle et al., 1993). The ERB experiment was a multipurpose experiment that included an electrically self-calibrating active cavity radiometer with an estimated at-launch accuracy of 0.5%(Hickey et al., 1988). Limitations imposed on the ERB solar observations by the absence of solar pointing on the Nimbus platform sustained a noise level in the ERB results which was estimated as 0.03%(Hoyt et al., 1992) and this low precision inhibited the recognition of intrinsic solar variability until subsequent detection by the Active Cavity Radiometer Irradiance Monitoring (ACRIM) experiment on board the SMM satellite (Willson et al., 1981). The irradiance observations of the SMM/ACRIM radiometer started in February 1980. During the first 10 months of the ACRIM observations the measuring precision was so high (0.002%) that each observed event had a solar, rather than an instrumental, origin (Willson, 1984). The high precision and accuracy of the ACRIM I observations are attributed to the phased operation of its three independent ACR sensors and the full time solar pointing of the SMM satellite (Willson and Hudson, 1991). Unfortunately, in December 1980 the solar-pointing system of the SMM satellite failed and it was placed into the so-called ``spin-mode''. During this operational mode until May 1984, when the SMM was repaired during a Space Shuttle Mission, the measuring precision decreased by about a factor of 5 (from 0.002%to 0.01%) (Willson, 1984). After May 1984 the original measuring precision was restored and the original high quality ACRIM I total irradiance data were collected through November 1989 when the operation of the SMM satellite was terminated.

Extensive measurements of the solar ultraviolet irradiance have been made by the Solar Backscatter Ultraviolet (SBUV1) experiment on the Nimbus-7 satellite between 160-400 nm from November 1978 through 1987 (Schlesinger and Cebula, 1992; Cebula, Hilsenrath and Deland, 1994) and by the NOAA9 and NOAA11/SBUV2 instruments since 1985 (Donnelly, 1991; Donnelly, White and Livingston, 1994). Although the SBUV experiments suffer from a significant degradation in its diffuser reflectivity, the ratio of the irradiance in the core of the Mg 280 nm line to the irradiance at the neighboring continuum wavelengths (Mg II h &k core-to-wing ratio) can be used as an index of solar variability (Heath and Schlesinger, 1986). It has been demonstrated that the Mg II index is capable of measuring solar rotational modulation on both 27 and 13.5 days time scales and solar cycle variability on the 11-year time scale (Donnelly and Puga, 1990; Pap, Tobiska and Bouwer, 1990). Since the UV irradiance plays a significant role in the dynamics of the middle atmosphere of the Earth and establishing its chemical composition through photodissociation and photoionization processes, it is highly desirable to develop a proxy indicator for these variations. Heath and Schlesinger (1986) have derived ``scale factors'' to relate the Mg II index variations to changes at other UV wavelengths. The Nimbus-7 and NOAA9 Mg core-to-wing ratios (Mg c/w) were made consistent by means of their overlapping observational time period in 1986, and also by means of the full disk Ca II K index (Donnelly, 1991; Donnelly, White and Livingston, 1994). Two sets of the Mg c/w index have been available, derived at The NASA/Goddard Space Flight Center (GSFC) (Cebula, Hilsenrath, and Deland, 1994) and at the NOAA/Space Environment Laboratory (SEL) (Donnelly, White, and Livingston, 1994), respectively. In this paper the NOAA/SEL Mg c/w ratio has been used as a measure of UV irradiance.

In order to directly compare the variations in total solar irradiance with the changes in the UV irradiance at 280 nm represented by the Mg c/w, the effect of sunspots have been removed from total irradiance by means of the Photometric Sunspot Index (Hudson et al., 1982; Fröhlich, Pap and Hudson, 1994). The Photometric Sunspot Index (PSI) relates the observed sunspot areas and their contrast to a net effect on the radiative output of the observed solar hemisphere and it is corrected for the photospheric limb darkening. The original PSI function (Hudson et al., 1982) was calculated on the assumption that the area and temperature ratio of the umbra and penumbra are constant and therefore the PSI function was calculated on the basis of a constant contrast value of sunspots (details on the calculation of PSI are given by Hudson et al., 1982 and Fröhlich, Pap and Hudson, 1994). However, results of the photometry of sunspots have demonstrated that the contrast of sunspots is not a constant value: it changes as a function of their area (Steinegger et al., 1990) and the solar cycle (Maltby et al., 1986). Based on the results of photographic photometry published by Steinegger et al. (1990), the PSI function, used in this study, has been recalculated by Fröhlich, Pap and Hudson (1994). Since the solar magnetic activity is assumed to be the driving force for irradiance variability (e.g. Harvey, 1994), the full disk magnetic flux measured at the National Solar Observatory at Kitt Peak is included in our analysis and it has been directly compared to the irradiance data.