A MEASUREMENT OF THE BACKGROUND SOLAR VELOCITY SPECTRUM

Ten years of continuous measurements of the radial velocity of the Sun have been used to estimate the contribution of various aperiodic sola r surface phenomena to the observed background solar velocity spectrum (BSVS). The characteristics of this spectrum are of great importance, as they represent the ultimate limit on the sensitivity of measuremen ts of solar oscillations. A precise determination of this spectrum fro m the ground is invariably contaminated by the discontinuity in the ob servations, the unavoidable effect of the Earth's atmosphere and, poss ibly, by the instrumentation itself. The present analysis uses observa tions made with a very stable instrument to allow comparison of severa l sets of data, collected along different phases of the solar activity cycle. The results show a high stability of the BSVS throughout the f requency ranges of interest for helioseismology, and, moreover, its pr ofile cannot be modeled using a single exponential function, as the ex ponent changes with frequency. The roughness of the spectrum is calcul ated, allowing an estimate of the required signal-to-noise ratio in or der to detect an oscillation with a given amplitude. Finally, followin g the modeling of the expected background spectrum proposed by Harvey, the rms full disk velocity of the main solar atmospheric phenomena (g ranulation, mesogranulation, supergranulation, and active regions) are calculated. Despite the limitations of the procedure employed in this analysis, the main conclusion is the overall good agreement with the model, although mesogranulation seems to have been underestimated. Com parison with the different phases of the solar activity cycle reveals a different behavior of the mesogranulation, while granulation appears to be stable.