We have attempted in this paper to measure and quantify the lifetime of mesocells (4- ) from a long time sequence (6 h 40 min) obtained a the Pic du Midi Observatory. These mesocells were detected using the Local Correlation Technique (LCT) developed by L. November (1988). The adjustment of the pixel size with respect to the derived displacements and the comparison of horizontal flow measurements provided by the application of two different LCT algorithms (November 1988, Shine in Simon et al. 1988) to our data set make us confident in the reliablity of our results. We found that November's LCT, used in this study, underestimated high velocity flows when compared with Shine's LCT, but that the relative orientations and amplitudes of the velocity vectors are almost identical in most of the field of view, where the flows are well defined. The differences in the divergence maps derived by the two methods (November and Shine's LCTs) can be attribuated to a less accurate determination of the small velocities (moduli and orientations), in the less well defined convergent flows. Our detailed comparison of the flows calculated by both LCT algorithms (November and Shine) make us confident in the divergence locations and evolution.
Direct tracking of cells during their evolution seems to be a suitable method in order to monitor individual mesocells and quantify their lifetime.
The mesocell lifetime histograms derived from our measurements reveal a continuous distribution between 10 min and 2 h 40 min, peaked at around 30-40 minutes. This result is not significantly influenced by the choice of the temporal window in the LCT. 70% of the mesocells have a lifetime less than or equal to 50 minutes. Nevertheless, there is a clear component in the histogram which extends up to 2 h40 min, related to the long living mesoscale features. The short lifetime mesocell component is probably related to the family of exploding granules described by Kawagushi (1980). It could also represent the contribution of the exploders in the mesogranules, described by Simon et al. (1991), or else a component of scatterred exploding granules. From the continuous histograms of mesocell lifetimes, it is quite difficult to separate the granulation component from the mesogranulation component. Mesogranulation appears as an extended part of the granulation phenomena.
Mesocell lifetime has also been determined by the temporal correlation coefficient approach. This method, applied to the and flow components, yields results which are at least two to five times smaller than the previously published values (Brandt et al. 1994, Darvann 1991). The high scatter of the correlation coefficient values, described by Darvann (1991) and Brandt et al. (1994), is the major uncertainty in the determination of mesocell lifetime. Our observation was taken at higher altitude (40 km above ) than the previous measurements, but we do not believe this factor could explain the differences. We think rather that the correlation coefficient method is too sensitive to the temporal window of the LCT and does not take into account proper mesocell motions.
Our estimation the mesocell lifetime is all smaller, by a factor 2 to 3, than Muller's observation (Muller et al. 1992). These differences may be explained by a mesocell selection arising from the use of a different definition, in particular for the direct measurement of lifetimes and motions. They only consider those divergences whose flows are circular around a central point while in the present study we use the divergence field which corresponds to an exploding activity.
We observe a variation in the number of new mesocells with a period of about 90-100 minutes, although the contribution of the positive divergences in the total field of view is quasi constant at 49% throughout the time sequence. This result does not seem to derive from our reduction method and, is likely to have a solar origin, but it has to be confirmed by other observations.
The measured horizontal velocity histogram of these mesoscale cells lies between 0.1 and 0.9 km/sec and peaks at 0.5 km/sec, which is commensurate with previous determinations (Muller et al. 1992, Brandt et al. 1988, Simon et al. 1994) In the previous determinations (Muller et al. 1992) mesogranule motions were observed to converge toward the supergranule boundaries with a mean velocity of 0.3-0.4 km/sec. In our sequence, the proper motions of mesogranules with a lifetime greater than or equal to 70 min seem more random than those described by Muller et al. 1992. The different organization of the proper mesocells motions as described by Muller et al. (1992) and by the present study can be attributed to the different locations of the mesocell field of view in the supergranular network. Muller's observation is centered on a supergranular cell while our study's field of view is located at the corner of 3 supergranular cells.
We conclude that mesocell lifetime determination yields disparities that can be attributed to the mesocell definition or to the method of flow measurement (LCT methods, temporal correlation, direct tracking, etc..). They may also be due to the different location of the observed field of view with respect to the supergranular network.
© European Southern Observatory (ESO) 1998
Online publication: January 27, 1998