The penumbra of a sunspot consists of bright filaments separated by dark fibrils. Bright filaments are, in fact, chains of elongated bright features called penumbral grains (PGs, Muller 1973a,b). They have comet-like shapes with "heads" pointing towards the umbra. Their width is 0:004 on average, and their length ranges from 0:004 to more than 2". The brightness of the "heads" approaches that of the photosphere. PGs have not been studied by many authors because successful observations require an extended period of excellent seeing.
Schröter (1962) noted that "Bright knots as seen in the penumbral filaments show ... a systematic motion (1-2 km/s) away from the center of the spot." Muller (1973a) used an excellent series of 34 white-light photographs, taken at Pic du Midi and covering an interval of 3 hours, to identify and track visually 220 PGs in order to determine both velocities and lifetimes as functions of position within the penumbra. He claimed that PGs moved toward the umbra; their horizontal speed was maximum (0.5 km s-1) at the penumbra-umbra (P/U) border and zero at the penumbra-photosphere (P/Ph) boundary. The lifetimes were about 3 hours in the middle part of the penumbra, and 50 and 40 minutes in the inner and outer parts, respectively. Tönjes & Wöhl (1982), using a similar method, in general confirmed the results of Muller but found lower horizontal velocities with a maximum of 0.33 km s-1 located in the middle penumbra. Lifetimes ranged from 1 to 3 hours.
The horizontal motions in penumbrae can be measured using local correlation tracking (LCT, November & Simon 1988). Because of LCT's inherently poor spatial resolution (typically 1" -2"), and its inability to distinguish between motions of bright or dark features, it is not clear that LCT of a sunspot penumbra is tracking solely PGs, or also other penumbral features. Wang & Zirin (1992) applied LCT to series (duration 2-3 hours) of digitized video images of five sunspots. In four cases they observed that "both bright grains and dark fibrils in the inner part of the penumbra move toward the umbra at a speed of about 0.5 km s-1" and "the elements in the outer part of the penumbra move outward at about the same speed as the inflow." In one case the inflow occurred over the whole penumbra. Also applying LCT, Denker (1998) reported that a line of positive divergence divides the penumbra, implying opposite directions of motion in the inner and outer parts.
The average observed brightness of PGs ( nm) and dark spaces between them is 0.95 and 0.60 of the mean photospheric intensity (), respectively (Muller 1973b, Collados et al. 1988). Recently, Sütterlin & Wiehr (1998) presented a temperature map derived from speckle-reconstructed three-color photometry of a large sunspot. The sunspot penumbra shows typical spatial temperature fluctuations of 700 K. The mean temperature of PGs is about 6250 K, hotter by 100 K than the mean photosphere. Dark penumbral fibrils have an average temperature of 5650 K, corresponding to an intensity of 0.64 .
A theoretical explanation for PGs has been suggested by Schlichenmaier et al. (1998a,b). According to their model, a PG is the intersection (footpoint) of an inclined thin magnetic flux tube with the visible surface. Hot sub-photospheric plasma flows upward along this tube. As the flux tube rises, its inclination decreases, so that the footpoint exhibits a horizontal motion toward the umbra. The model predicts a gradual decrease of horizontal velocity from 2 km s-1 at the beginning, to 0.1 km s-1 at the end, of a PG's lifetime. Bright filaments are interpreted as dimmer and thinner tails of PGs. Their width is very small, only 50 km. The extreme narrowness of penumbral filaments was confirmed by Sánchez Almeida & Bonet (1998) who found that the spatial spectrum of penumbral intensity fluctuations, corrected for the instrumental modulation transfer function, was flat up to the highest observable spatial frequency. This implies that the true structure of penumbral filaments has not yet been resolved.
In 1993 an 11-hour time series of high-resolution white-light images of solar granulation was obtained by Simon et al. (1994). A medium-size sunspot was present in the field of view for about 4.5 hours. Sobotka et al. (1997a,b - Papers I and II of this series) analyzed temporal variations of sunspot fine structure in the central umbral core, and applied a feature tracking algorithm to umbral dots. We now use this algorithm to measure proper motion, brightness, and lifetime of PGs in the penumbra of this sunspot. Preliminary results have been published by Sobotka et al. (1999). We have made substantial improvements in the algorithm, and here present our revised results. We hope these will spur refinement of theoretical models, and stimulate further discussion of the earlier measurements.
© European Southern Observatory (ESO) 1999
Online publication: July 26, 1999