Astron. Astrophys. 333, 1053-1068 (1998)

4. Development history of an active region

The unusually complete temporal sequence of the emergence of Region 3 of the period of six days allows us to place the signatures of the emerging vector field as outlined in the previous section into the context of the overall evolution of the active region.

4.1. Overview of the development

Fig. 6 illustrates the development of the magnetic field in Region 3 during six days of its disk passage. This sequence reveals the evolution of Feature GP which grew into the dominant leader sunspot, with a penumbra clearly visible along the northern sector of the spot in the first observations on the . During this entire sequence, emerging (horizontal) flux was visible between the two opposite polarities, as identified by weaker field strengths that are prominent in the field strength images at the right, but not apparent in the flux images of the middle column. The eventual leader and follower sunspots separated along an east-west direction during this sequence.

Fig. 6. A time sequence of inversion results for Region 3, NOAA 7781, is represented by gray-scale images. The columns, from the left, are images of continuum intensity (Ic), mean vertical field or flux F, and magnetic field strength . Images are presented in the local solar frame, with solar longitude increasing (westward) to the right and latitude (north) up. The flux image is scaled from -1500 to G (negative = dark), and G (white = 2000 G, background = 1500 G). Poor seeing and clouds cause the vertical stripes at the center of the image for the , and faculae associated with plage are visible in the continuum image for the when the region was at longitude W . Times represent approximately the midpoint of the spatial map.

The spatial distribution of regions of weaker field strength fragments with age of the emerging region, such that by Sept. 27 the flux emergence zone between the two spots has a mottled appearance in the image of . The emergence zone evolves from a rather uniform region occupied primarily by weak, horizontal fields, to a mixture of weak fields and regions of stronger (kiloGauss), more vertical fields.

An interesting aspect of the apparent evolution of the emergence zone is revealed by Fig. 7, where the kiloGauss fields are inferred to be more horizontally oriented as the date of the observation increases. The emergence zones defining the selected data points of Fig. 7 were isolated in the same way as was done for Fig. 3. We believe this to be the influence of a low canopy of arched magnetic fields that traverses the emergence zone: a behavior that is more pronounced as the active region approaches the limb where the surface is higher in the atmosphere. The polarization signals from both the canopy and the stronger, more vertical flux elements become mixed substantially towards the limb, resulting in inferred field strengths that are large, but also highly inclined to the surface. We have not seen this effect in any of the active regions observed near the center of the disk, including Region 2 which was more highly evolved than the others. Furthermore, even near the limb, plage regions away from the emergence zone are inferred to be nearly vertical, both in the present data and in previous studies using data from the ASP (Martinez Pillet et al. 1997a). Therefore, several indicators suggest that the phenomenon apparent in Fig. 7 is due to a low, unidirectional magnetic canopy over the emergence zone.

Fig. 7. Scatter plots are presented for inferred zenith angle vs. field strength for the emergence zone of Region 3 during the period 1994 Sept. 24 - 27. The zenith angles for the higher field strengths in this zone become systematically closer to as the region is viewed closer to the solar limb. As in the scatter plots of Figs. 3 and 4, the emergence zone was chosen visually as the region between the opposite polarity spots where there are measured vector fields. See caption of Fig. 3 for further details.

The original flux emergence zone associated with the emerging Feature GP on the was oriented such that the field azimuth was directed from southwest toward northeast, with value (see the weak field region just to the north of the emerging pore in Fig. 8). Towards the end of our six-day observing period the field azimuth within the emergence zone became oriented closer to the west east orientation, but never surpasses in our observing time frame. Thus this emerging region never reaches the azimuth value of expected of "Joy's law".

Fig. 8. A time series of the development of the newly-emerging flux is presented for a subarea of the data from 1994 Sept. 23. As in Fig. 6, the images of , F, and are shown, except that the flux image is scaled from -1000 to G. The acute field angle indicates clearly where the field vector is oriented nearly horizontal (dark) or nearly vertical (white) to the local solar surface. The tick mark interval is 5 Mm. Times represent approximately the midpoint of the spatial map.

The original leading sunspot (westernmost spot at the far right of the image on the - see Fig. 6) migrated slowly in a southeasterly direction (relative to Feature GP) and gradually disappeared, contributing flux to both Feature GP and the surrounding positive polarity plage.

4.2. An emergence event: the birth and growth of a sunspot by both bipolar flux emergence and moving magnetic features

The sequence of Fig. 8 shows the early development of the main flux emergence zone during eight hours on the . most of the flux emerges as horizontal, weak fields at the immediate northeast edge of Feature GP, as may be seen in the images of (low field strengths = dark) and ( = dark). Fig. 8 displays only every image recorded, but the full video loop of the the observations on this day clearly show that positive (leading) polarity flux emerged very close to Feature GP and became immediately encorporated therein at the high field strengths typical of pores. In contrast, the following polarity flux migrated some distance northward and eastward to form the assemblage of very small following-polarity plage elements and pores (see flux image sequence of Fig. 8.)

Feature GP also grew as a result of assemblage of flux from pre-existing flux to its west. Fig. 9 presents the and F images separated by seven hours on the . During the entire time sequence recorded on this and the following day "moving magnetic features" (MMFs) emanate in every direction from the decaying westernmost sunspot, ⁴ which continued to reduce in size, and to drift in a southeasterly direction. Although some of these MMFs migrate into plage surrounding their parent sunspot (i.e., at the outer edge of this sunspot's "moat"), it is also clear that some of them (e.g., flux element 'c') also migrate toward and are assimilated into the growing pore to its east. The growing pore also assimilates magnetic flux from elements not clearly of MMF origin (flux element 'b'). Thus, it appears that Feature GP grew as a result of both new emerging flux and acquisition of flux from a decaying spot and other nearby flux.

Fig. 9. Migration and coalescence of magnetic flux is apparent in the and F images for Region 3, 1994 Sept. 23. Flux emergence zone is indicated by 'a'. Flux elements 'b' and 'c' retain their identity during this 7-hour interval, and may be followed in the time sequence as they coalesce into the growing pore near the center of the image. Initially separate flux 'd' coalesces into a single bundle by the end of the sequence. Numerous "moving magnetic features" may be seen emanating from the large pore/sunspot at the right (west).

In contrast to the decaying sunspot, Feature GP does not show any sign of MMF expulsion during these two days of its active growth when the decaying spot to the west is continually active in ejecting MMFs. By the Feature GP (now a sunspot with a partial penumbra) began to eject MMFs itself. Another aspect of the MMF phenomenon is that, on the while Feature GP is still rapidly growing, some of the MMFs ejected from the decaying spot appear to flow past and around the outer perimeter of Feature GP then become incorporated into the leading polarity plage. They appear to be diverted in their movement so as to bypass Feature GP. This suggests that those particular MMFs are not connected immediately below the surface to the flux system that makes up Feature GP, otherwise they would have been assimilated into that pore.

4.3. Evolution of plage field strengths during emergence

These data have neither the time resolution nor the continuously high-quality spatial resolution to discern positively the flux history of individual, small (arcsecond or less in linear dimension) flux elements. However, it is clear from the development of Region 3 (Fig. 8) that at least some of the emerging flux rapidly evolves from weak and horizontal to strong and vertical once it has emerged. This is especially evident for the following polarity flux which moves only a short distance northeastward from the emergence zone before increasing to typical plage field strength of 1300 G. The following polarity flux gradually assembles itself into a sunspot on the .

4.4. Quantitative flux history of Region 3

Fig. 10 plots the net observed magnetic flux within the FOV for Region 3 during the six-day sequence shown in Fig. 6. The observations are presented as a function of time, illustrating the limitations of observing from only one site even if those observations cover most of the available nine-hour observing day. Through the , the leading ( 's) and following (-'s) polarity fluxes are nearly balanced. After that time, a considerable ( %) flux imbalance persists until the end of our observations. Also, we note on the a much larger spread in observed negative polarity than on previous days, while the spread in observed net positive polarity flux remained comparable to that of the previous days.

Fig. 10. The evolution of net positive () and negative (-) flux is shown for the 6 days of evolution of Region 3. Symbols are placed on this plot at the fraction of day corresponding to the UT of the observation. Fluxes represent only those field elements within the observed field-of-view.

Both the flux imbalance and the spread in the observed flux measures may result from observational selection effects. Our FOV is not large enough to encompass all of the flux of this region, as is especially evident from the onward. The following (negative polarity) flux, primarily in the form of plage, might be migrating outward from the region at a higher rate than that confined to sunspots or pores. This could lead to an observational bias in favor of the leading polarity flux. Seeing is also a major factor when the flux of each polarity is dominated by different structures. When the seeing worsens, the observed polarization from small plage flux elements may drop below our observational threshold, thus leading both to reduced total negative flux and to a larger observed scatter, especially during the later evolution of this region. We illustrate the likelihood of this latter effect in Fig. 11, where we plot separately the total negative and total positive flux as a function of the observed rms granular contrast for each map observed on the . Because is an excellent indicator of seeing quality, it helps us to explore the relationship between measured flux and seeing quality. The granular contrast was computed for a rectangular area within the FOV, away from pores or sunspots, but still containing some plage as the active region covers most our mapped area. Bearing in mind that the seeing quality varies during one of our spatial maps, the only represents some sample of the seeing during that portion of each map. Fig. 11 also shows LSQ linear fits to the flux of each polarity. The negative flux, being composed mainly of plage, shows a stronger dependence upon than does the positive flux, which is dominated by the sunspots and pores. This relationship is more distinct for the frames having larger , i.e. those frames having better seeing. On the condition that complete flux balance between the two polarities exists within our FOV, as the angular resolution increases one would expect these relationships to approach asymptotically the true flux. It is interesting to note that the two extrapolated linear relationships in Fig. 11 intersect at a value of typical of that obtained at very high angular resolution (Lites et al. 1988). A substantial fraction of the total flux may be below the adopted polarization threshold, perhaps approaching 50% of the measured flux in maps having the worst seeing.

Fig. 11. The net magnetic flux of each polarity is plotted as a function of the observed rms continuum contrast of granulation for each of the maps of NOAA active region 7781 observed on 1994 Sept. 24. Linear LSQ fits to the positive (negative) flux distributions are shown by the solid (dashed) curves. The relationship of observed flux vs. rms granular contrast is more distinct for the higher contrast images, presumably indicating a relationship between observed net flux and the quality of seeing (see text).

The presence of such flux, as indicated by substantial Stokes V signals (accompanied by Stokes Q, U signals that are not sufficiently large to allow a reliable inference of the vector field), has been well-known to us (cf. Lites 1996; Lites et al. 1996). If such weak polarization arises from significantly inclined magnetic fields, Fig. 11 suggests both that higher angular resolution and better polarimetric precision would be needed to properly account for most of the flux of active regions. However, the present work indicates that emerging horizontal fields rapidly orient themselves vertically, so the issue of flux accounting in mature active regions near disk center might be reasonably addressed by assuming most of the weaker polarization arises from vertically oriented fields. In this case, their Stokes V profiles, which are well-observed with the ASP, could be used to estimate the total flux. In the future we plan to explore this possibility.

We speculate on the origin of the observed drop in flux of both polarities after the . On the one hand, it is possible that this decrease is also an observational selection effect due to the viewing angle. Near the limb we see higher in the atmosphere where the intrinsic field strength, hence line polarization, is weaker, thus being more subject to a threshold selection effect. On the other hand, this behavior may be a valid property of the photospheric field, in which case we need to determine if the loss in flux represents a diffusion of flux out of our FOV, or if it could be a result of loss of flux upward into the corona (e.g. Lites et al. 1994; Low 1994).

© European Southern Observatory (ESO) 1998

Online publication: April 28, 1998

helpdesk.link@springer.de