Time Variability of the “Quiet” Sun Observed with TRACE. I. Instrumental Effects, Event Detection, and Discrimination of Extreme-Ultraviolet Microflares

The Transition Region and Coronal Explorer (TRACE) observed a "quiet-Sun" region on 1999 February 17 from 02:15 UT to 3:00 UT with full resolution (05 pixel size), high cadence (125 s), and deep exposures (65 and 46 s) in the 171 A and 195 A wavelengths. We start our investigation of the time variability of "quiet-Sun" images with a detailed analysis of instrumental and nonsolar effects, such as orbital temperature variations, filtering of particle radiation spikes, spacecraft pointing drift, and solar rotation tracking. We quantify the magnitude of various noise components (photon Poisson statistics, data digitization, data compression, and readout noise) and establish an upper limit for the data noise level, above which temporal variability can safely be attributed to solar origin. We develop a pattern recognition code that extracts spatiotemporal events with significant variability, yielding a total of 3131 events in 171 A and 904 events in 195 A. We classify all 904 events detected in 195 A according to flarelike characteristics and establish a numerical flare criterion based on temporal, spatial, and dynamic cross-correlation coefficients between the two observed temperatures (0.9 and 1.4 MK). This numerical criterion matches the visual flare classification in 83% of the cases and can be used for automated flare search. Using this flare discrimination criterion we find that only 35% (and 25%) of the events detected in 171 (and 195) A represent flarelike events. The discrimination of flare events leads to a frequency distribution of peak fluxes, N(ΔF) ∝ ΔF-1.83±0.07 at 195 A, which is significantly flatter than the distribution of all events. A sensitive discrimination criterion of flare events is therefore important for microflare statistics and for conclusions on their occurrence rate and efficiency for coronal heating.