High-frequency oscillations in a solar active region coronal loop

D. R. Williams 2 P K. J. H. Phillips 1 P. Rudawy 0 M. Mathioudakis 2 P. T. Gallagher 3 E. O'Shea 2 F. P. Keenan 2 P. Read 1 B. Rompolt 0 0 Institute for Astronomy, University of Wrocaw , Wrocaw, Poland 1 Space Science & Technology Department, Rutherford Appleton Laboratory Chilton , Didcot, Oxon. OX11 0QX 2 Department of Pure and Applied Physics, The Queen's University of Belfast , Belfast BT7 1NN 3 Big Bear Solar Observatory, New Jersey Institute of Technology , Big Bear City, CA 92314, USA A B S T R A C T The Solar Eclipse Corona Imaging System (SECIS) was used to record high-cadence observations of the solar corona during the total solar eclipse of 1999 August 11. During the 2 min 23.5 s of totality, 6364 images were recorded simultaneously in each of the two channels: a white light channel, and the Fe XIV (5303 A ) 'green line' channel T , 2 MK . Here we report initial results from the SECIS experiment, including the discovery of a 6-s intensity oscillation in an active region coronal loop. I N T R O D U C T I O N The origin of the solar coronal temperature, which is of the order 106 K, is still the subject of much debate in solar and plasma physics (see review by Priest & Schrijver 1999). Efforts to establish the causes of coronal heating to date have produced two main classes of theory. The first holds that the corona is heated by current dissipation following myriad magnetic reconnections occurring throughout the corona in the form of flare and nano-flare activity (Parker 1988). The second class holds that the heating is driven by the damping of magnetohydrodynamic (MHD) waves propagating from the lower solar atmosphere and dissipating through ion viscosity and electrical resistivity (Hollweg 1981). MHD waves may be broken down into two main subcategories: Alfven waves and magneto-acoustic waves. Alfven waves may either be transverse and incompressible, propagating along the magnetic field, or compressional (compressing and rarefying the magnetic flux density perpendicular to the field). Magneto-acoustic (slow mode and fast mode) waves cause compression and rarefaction of the coronal plasma as they propagate. A notable observational difference between these two subcategories is that one might expect Alfven waves to cause only Doppler shifts in observed line measurements, whereas magneto-acoustic waves are expected to cause intensity variations, since the emission measure varies as the square of electron density ne2. Several previous studies have searched for intensity oscillations in the corona, by means of intensity, velocity and (to a lesser extent) linewidth fluctuations. Koutchmy, Z ugzda & Locans (1983), using coronagraph observations, found evidence of fluctuations in the velocity measurements of the coronal Fe XIV 5303-A -line, but did not detect intensity oscillations; however, Pasachoff & Landman (1984), using eclipse observations, did detect excess power in this line in the 0:52:0 Hz range using intensity fluctuation measurements. To date, the vast majority of studies have found oscillations with periods comparable to the photospheric 300-s oscillations (e.g. De Moortel, Ireland & Walsh 2000). Aschwanden et al. (1999) list all detected solar periodicities in the wavelength range 10 m . l . 10210 m. The detection of short-period oscillations are beyond the capability of most groundbased and space-borne instruments. Porter, Klimchuk & Sturrock (1994a) have investigated the feasibility of coronal heating by slow- and fast-mode highfrequency t , 100 s MHD waves. In a companion paper, (Porter, Klimchuk & Sturrock 1994b), they apply their model to the active region coronal loop conditions. One of the main findings of their work is that coronal heating by slow-mode waves is viable for periods of t # 100 s, and by fast-mode waves with periods of t # 1 s. These shorter periods are required to balance the radiative and conductive losses from the corona. Recent research has been encouraged by detection of wave motion in coronal loops, (McKenzie & Mullan 1997; Wood & Karovska 1998; Nakariakov et al. 1999). Here we describe results from the Solar Eclipse Corona Imaging System (SECIS; see Phillips et al. 2000) as a means of searching for oscillations in the solar corona. SECIS searches for periodicities as short as 4:5 1022 s (22.2 Hz), and so is well positioned to shed some light on the existence of coronal oscillations in a much faster, and hitherto unseen, frequency domain. An obvious consequence of its hightime resolution is that SECIS can also observe microflare activity in the solar corona. S O L A R E C L I P S E O B S E R VAT I O N S The SECIS instrument SECIS can record solar images at a rate of up to 70 Hz in each of the two channels. We used SECIS to obtain images of the 1999 August 11 eclipse at the Bulgarian Air Force Base, Shabla, Bulgaria. In the set up used on this occasion, one channel of SECIS was used to observe in the so-called green line of Fe XIV at 5303 A (hereafter referred to as the Fe XIV channel), while the other had no filter in place, and so was used to observe in white light. The green line was isolated using an interference filter with nominal bandwidth full width at half-maximum (FWHM) 2.6 A . (Later measurements indicated the bandwidth FWHM to be 4 A .) Under clear skies the data were recorded without interruption. The data consist of 6364 images in the Fe XIV channel, and the same number of white light images taken simultaneously. These images were recorded over the totality period of 143.5 s, yielding a cadence of 2:25 1022 s, or a frame rate of 44 Hz. Each image has a 512 512 pixel2 format, although the edge regions are unusable, so the analysed images are in a 504 504 pixel2 format. The image depth is 12 bits per pixel, with the two least-significant bits being treated as noise. The scale of the pixels is the same in both channels to within 1 per cent: each pixel in the green line channel corresponds to 4.07 arcsec (,8 arcsec resolution), whilst the white light pixels have been determined to be 4.04 arcsec wide (Rudawy et al. 2001) as confirmed by pre-eclipse measurements. For a detailed description of the instrument see Phillips et al. (2000). Additionally, sets of 500 images were taken for the purpose of calibration soon after the end of the eclipse, i.e. one set in each channel for dark-current calibration and two sets in each for flatfielding. This whole data set required approximately 7.5 GB of disk space. Fig. 1 illustrates the instrumental set up used for this experiment. Light from the Sun is reflected by a computer-controlled heliostat mirror onto a horizontally mounted 200 mm Schmidt Cassegrain telescope. This keeps the image of the Sun stationary in the focal plane. When the light emerges from the telescope, it is collimated before encountering a beam splitter mounted along the optical axis. The light is split into two beams, one beam continuing through the green line filter to the first charge-coupled devi (...truncated)