Angular rotation rates exhibit no significant variation over the five years of the data set. We did not detect any strong north-south asymmetry in rotation rates in either the cycle 19 or 21 sunspots.
Our m.c. study confirms cycle 21 sunspot results (Paper II), namely a poleward/equatorward m.c. pattern with latitude-dependence. We also computed sunspot covariance and found it to be very small. The differences observed in Paper II between leader and follower behaviors, and between old and young sunspots, are not present in cycle 19, and probably reflect some internal variability specific to cycle 21. So the main conclusion of this paper together with Paper II is the following. The covariance exhibited by sunspots seems to be very small. Recent modeling of the differential rotation (Küker et al. 1993) requires a smaller covariance than previous models, which would be in agreement with our observations. A discussion of these results as well as facula covariances is given in Meunier et al. (1996b).
Two main sunspot data sets are available for the same period (cycle 19): the Greenwich photoheliographic plates and the Mount Wilson observations. The rotation obtained from these data sets agrees fairly well with the Meudon sunspot rotation. However, there are strong discrepancies in the covariances. The Greenwich covariance was calculated from the motions of the barycenter of the whole sunspot groups (Ward, 1965; Balthasar et al., 1986). So a direct comparison with individual sunspot motions is not straightforward, as discussed at length in Paper I. The motions of individual sunspots are available from the Mount Wilson data set. However, the tracking procedure used by Howard et al. (1984) is subject to biases mentioned in Papers I and II.
In Collin et al. (1995), we used helioseismology results (Gough et al. 1993) to deduce tracer anchorage depth. In the present work, we obtain similar rotation rates for old and young sunspots, so we confirm our conclusion about the deep sunspot anchorage. However, seismic observations of internal rotation are not available for cycle 19, and it might be unsafe to extrapolate cycle 21 data to cycle 19.
© European Southern Observatory (ESO) 1997
Online publication: June 30, 1998