Randomly sampling the chromospheric peak power distribution

We have analyzed the UV continuum light curves of 2535 pixels obtained during fifteen separate observations of active regions with the Ultraviolet Spectrometer and Polarimeter (UVSP) on board the solar maximum mission in 1989. Specifically we have looked for periodicities. In the power spectra of 738 light curves (i.e., 29%), evidence was found for periodicities at multiple frequencies. For each power spectrum, containing significant power, we determined the frequency at which the maximum power is found. The distribution N (ν), which describes the number of pixels having maximum power at frequency ν, is strongly concentrated in the 2–5mHz band with a distinct maximum at 3–3.5mHz. No pixels had their maximum power above 10mHz and only a few had their maximum power in the 5–10mHz range. The oscillations in the 2–5mHz range are probably related to evanescent acoustic waves driven by the photospheric five minute oscillations. In the 2–5mHz band, the distribution N (ν) resembles very much the power spectra as have been observed near the temperature minimum. This suggests that the photospheric power distribution can be interpreted as a probability distribution for finding a given frequency in the UV continuum light curves. In the 2–5 mHz band the spread of the maximum powers and the spread of the count rates, at any frequency, is much larger than above 5mHz where both are relatively constant. No clear correlation is found between the maximum power and the count rate in a pixel. Furthermore, no evidence is found for emission from the shocks which have recently been invoked to explain the formation of Ca II grains. Whether this is due to temperature effects is unclear. Our general procedure adopted to estimate the confidence level in the power spectrum of a light curve containing photon noise can be equally applied to SUMER and CDS data from SOHO.