Mid-Term Quasi-Periodicities and Solar Cycle Variation of the White-Light Corona from 18.5 Years (1996.0 – 2014.5) of LASCO Observations

Abstract

We report on the analysis of the temporal evolution of the solar corona based on 18.5 years (1996.0 – 2014.5) of white-light observations with the SOHO/LASCO-C2 coronagraph. This evolution is quantified by generating spatially integrated values of the K-corona radiance, first globally, then in latitudinal sectors. The analysis considers time series of monthly values and 13-month running means of the radiance as well as several indices and proxies of solar activity. We study correlation, wavelet time-frequency spectra, and cross-coherence and phase spectra between these quantities. Our results give a detailed insight on how the corona responds to solar activity over timescales ranging from mid-term quasi-periodicities (also known as quasi-biennial oscillations or QBOs) to the long-term 11 year solar cycle. The amplitude of the variation between successive solar maxima and minima (modulation factor) very much depends upon the strength of the cycle and upon the heliographic latitude. An asymmetry is observed during the ascending phase of Solar Cycle 24, prominently in the royal and polar sectors, with north leading. Most prominent QBOs are a quasi-annual period during the maximum phase of Solar Cycle 23 and a shorter period, seven to eight months, in the ascending and maximum phases of Solar Cycle 24. They share the same properties as the solar QBOs: variable periodicity, intermittency, asymmetric development in the northern and southern solar hemispheres, and largest amplitudes during the maximum phase of solar cycles. The strongest correlation of the temporal variations of the coronal radiance – and consequently the coronal electron density – is found with the total magnetic flux. Considering that the morphology of the solar corona is also directly controlled by the topology of the magnetic field, this correlation reinforces the view that they are intimately connected, including their variability at all timescales.

Appendix: LASCO-C2 Data Preprocessing

The routine observational program is basically composed of unpolarized white-light images of $1024\times1024~\mbox{pixels}$ every 20 min and a daily polarization sequence all taken with an orange filter (bandpass of 540 – 640 nm); in addition, there have been specific campaigns usually coordinated with other SOHO instruments.

The present analysis makes use of these polarization sequences since their temporal cadence is sufficient for our purpose and since they allow extracting the K-corona following a classical procedure to be described below. A polarization sequence is composed of three linear polarized images of the corona obtained with three polarizers oriented at $60^{\circ}$, $0^{\circ}$, and $-60^{\circ}$ and an unpolarized image, all taken with the orange filter in the binned format of $512\times512~\mbox{pixels}$.

All LASCO images, either polarized or unpolarized, undergo a standard preprocessing that corrects for instrumental effects. The in-flight performances of C2 are continuously monitored so as to update these corrections and the absolute calibration. Our analysis also benefits from recent improvements introduced in the procedures. The preprocessing performs the following tasks:

• 1. Bias correction. The bias level of the CCD detector evolves with time; it is continuously monitored using specific blind zones, and is systematically subtracted from the images.

• 2. Exposure-time equalization. Small random errors in the exposure times are corrected using a method developed by Llebaria and Thernisien (2001) and refined by Llebaria, Loirat, and Lamy (2010) in which relative and absolute correction factors are determined.

• 3. Missing-block correction. Telemetry losses result in blocks of $32 \times32~\mbox{pixels}$ sometime missing in the images. Different solutions are implemented to restore the missing signal depending upon the location of these blocks (Pagot et al., 2013).

• 4. Cosmic-ray correction. The impact of cosmic rays (and stars) is eliminated from the images using the procedure of opening by morphological reconstruction developed by Pagot et al. (2013).

The polarized images also undergo the following processes:

• 1. Correction for the global transmission of the polarizers.

• 2. Correction for the spatial variation of the transmission of the polarizers (Llebaria and Lamy, 2008).

• 3. Correction for the circular polarization introduced by the two folding mirrors.

• 4. Polarimetric analysis based on the Mueller procedure. This is applied to each triplet of polarized images and returns images of the total radiance, the polarization, and the angle of polarization.

• 5. Vignetting correction. This instrumental effect is removed from the radiance images using a geometric model of the two-dimensional vignetting function of C2 (Llebaria, Lamy, and Bout, 2004).

• 6. Absolute calibration of the total radiance derived from the photometric measurements of thousands of observations of stars present in the C2 field of view (Llebaria, Lamy, and Danjard, 2006; Gardès, Lamy, and Llebaria, 2013).

At this stage, the images represent the sum of the radiances of the K and F coronae and of the instrumental stray light. The separation of the K component is performed on the basis of the classical assumptions that the F corona and the stray light are unpolarized based upon the fact that both mostly result from diffraction effects. This is indeed a valid assumption for the F corona inside $6.5~R_{\odot}$. Using these assumptions, the polarized radiance $pB$ is equal to ${p}_{\mathrm{K}}{B}_{\mathrm{K}}$. At this point, we use a model of $\mathrm{p}_{\mathrm{K}}$ calculated for a standard spherical corona (Allen, 1973), which is justified by the robust asymptotic behavior of ${p}_{\mathrm{K}}$ beyond $2.2~R_{\odot}$. The radiance maps of the K corona can then be derived.