Photospheric flows in Sun-as-a-star EUV Doppler spectra
| Nugget | |
|---|---|
| Number: | 536 |
| 1st Author: | Hugh HUDSON |
| 2nd Author: | Anne-Marie BROOMHALL and Lyndsay FLETCHER |
| Published: | September 28, 2026 |
| Next Nugget: | |
| Previous Nugget: | Why do only some fast CMEs produce sustained gamma-ray emissions? |
Introduction
This Nugget describes a new application of the SDO/EVE Sun-as-a-star EUV spectroscopy. These very precise data allow us to get a first look at the mean Doppler variability revealed by many emission lines covering chromospheric to coronal temperature ranges. A first study of this kind (Ref. [1]) averages Doppler measurements for the entire year 2011, and finds characteristic and unexpected patterns. These features relate to mean properties, blithely ignoring all of the beautiful structure and dynamics of solar EUV images. Note that the EUV spectra do not show the photosphere directly, with dominant contributions from the transition region and low corona.
What do power spectra show us?
EVE operational patterns give us 3-hour contiguous data runs for most days of 2011. In each 10-s sample, one can get a robust Doppler measurement, and over a 3-h span this corresponds to the frequency range 0.1-50 mHz. This frequency range includes the p-modes, standing waves in the solar interior; the chromospheric 3-minute oscillations, a feature of solar sunspot activity, and photospheric flows due to convection, prominently the granulation in the photosphere.
At the highest frequencies (above 10 mHz), we do not have much stellar Doppler observation yet, basically because the stars are faint and require longer intgration times for such measurements. Accordingly the EVE Sun-as-a-star spectra are pioneering. What we can see in stellar observations is "microturbulence", a catchall phrase that describes the excess widths of spectral lines. This is an integral property of the power spectrum, present even in non-convective stars, and the significance of the EVE spectra is that one can get a detailed spectrum of the motions. This in principle can help to understand their physics.
EVE power spectra
To analyze the EVE spectra, we first find the line centroid wavelength in each 10-s integration. This is done with a six-parameter function consisting of a Gaussian plus a quadratic background continuum. In practice this effectively eliminates the effects of line blends in the EVE spectra, which do not have high spectral resolution. Figure 1 shows a month of EVE data in the Lyman-beta line, one of the EVE lines with chromospheric properties. Note however that it is optically thick, and that its contributions to Sun-as-a-star signals may actually show a range of heights, especially towards the limb. Because of projection effects, horizontal motions would have very strong limb brightening.
Each 3-hour data series, one per day, then yields a power spectrum via Fourier transforms of residuals against median values, and the mean of these spectra (an "incoherent sum", since phase information is not retained) results in a very clean spectrum, based on typically 250 good daily spectra from a single year. Figure 2 shows the 2011 spectra for oxygen and magnesium ions at different ionization states.
The photospheric motions and the Harvey function
All of the lines in Figure 2, and indeed the 20 other lines in the sample analyzed in Ref. [1] show similar features: they are broad continua, monotonically decreasing with frequency, and they show the characteristic Harvey profile (Ref. [2]) that nicely describes the continuum of photospheric motions. The Harvey spectrum is a Lorentzian, closely related to shot noise: white at low frequencies, and rolling over at a characteristic frequency (the inverse of the mean time scale of the source motions. The basic Harvey spectrum needs a generalization, since the EVE data clearly do not show a f-2 tail at high frequencies, but a much steeper power law that is quite well defined. In no case do the EVE spectra show broad spectral peaks either in the photospheric p-mode 5-minute range, or the chromospheric 3-minute range. The spectra also show no trace of a Kolmogorov f-5/3 turbulence.
Conclusions
The EVE spectra look entirely photospheric and probably result from basically horizontal motions, which couple to MHD modes in the solar atmosphere. Notably the Harvey spectrum near 8 mHz, which must reflect the granulation, becomes less important for the lines with coronal opacity. The amplitude of the flows is consistent with the "microturbulent" excess line widths, but the EVE data show that the Doppler motions are not turbulent in the sense of a Kolmogorov model.
References
[1] "Signatures of photospheric convection throughout the solar atmosphere"
[2] "Chromospheric Oscillations and the Background Spectrum"