Evidence of Solar Oscillations in Rayleigh-Scattered Light

A new instrument has been developed for making unimaged whole-disk observations of low-degree solar normal -mode oscillations. The apparatus, referred to as the sky monitor, does not track the solar disk, but instead measures the radiant flux scattered by the Earth's atmosphere at 1.6 and 0.5 mu. The expected diurnal flux variation was obtained from a detailed radiative transfer calculation. Data were acquired for 15.5 months during 1991-92. Seventy-two days of data were analyzed for evidence of solar p modes in the frequency range 1800-4776 muHz. Noise generated by the Earth's atmosphere was reduced by scaling the Fourier amplitude computed from the infrared signal and subtracting it from the Fourier amplitude computed from the visible signal. A superimposed frequency analysis was then performed which revealed ~2 sigma peaks within 0.3 muHz of symmetry-allowed locations, while no excess power was detected at the symmetry-forbidden frequencies. The probability of obtaining by chance the observed excess power density at symmetry-allowed frequencies and deficit of power density at symmetry-forbidden frequencies was computed to be 6.9 times 10^{ -3}. Correcting the frequencies for solar-cycle variations, the probability was reduced to 2.9 times10^{-4}. These results indicate that it is quite unlikely that the observed symmetry properties have occurred by chance, and support the hypothesis that solar normal-mode signals are manifested in the data. The amplitudes {cal I}^'/ {cal I}_{rm o} averaged over radial order n of the l = 0 and l = 2, m = 0 modes were found to be (7.54 +/- 0.54) times 10 ^{-7} and (7.68 +/- 0.56) times 10^{-7}, respectively. These results are about a factor of two smaller than the amplitude of total irradiance oscillations measured from space. While the rotational splitting of the l = 2 multiplet appears to be consistent with that reported by Hill (1985a), results for l = 1 and l = 3 are inconclusive.