Meter-wavelength interplanetary scintillation (IPS) observations have been used for remote sensing of the solar wind in the inner heliosphere, in particular to study its density and velocity structure. The ability to probe the heliosphere out of the ecliptic (i.e., at high heliographic latitudes) is an extremely valuable characteristic of these observations.
Remote sensing observations of the heliospheric density and velocity structure (e.g. from K-coronal brightness and IPS measurements) have been primarily interpreted within a descriptive framework in which the heliospheric magnetic field (in particular the current sheet) is used as a reference standard: low density, high velocity solar wind emanates from open magnetic fields (coronal holes) in both hemispheres, with high density, low velocity solar wind flowing outward near the current sheet. In this picture active regions, under-lying closed magnetic structures in the low corona leave little or no imprint on the solar wind structure.
Since reliable direct measurements of the global magnetic field in the inner heliosphere are currently not available, this framework is supported primarily by indirect evidence, such as correlative studies between IPS velocity and K-coronal brightness, and synoptic comparisons of IPS and K-corona with model calculations of the heliospheric current sheet. Though this framework has certainly been very useful in unifying many different aspects of heliospheric observation, it must be regarded as incomplete and too simple to account for all observations. For instance, due to the increased complexity of the heliosphere near solar maximum the relation between high density, low velocity and proximity to the current sheet breaks down (Rickett and Coles, 1991). In this paper we present tentative evidence that active regions play a role in the modulation of the solar wind, and hence may be a contributing factor to the solar wind mass loss.
The array of the Mullard Radio Astronomy Observatory (Cambridge, UK) takes daily scintillation level measurements at 81.5 MHz (\lambda=3.7 m) for over 900 sources covering the celestial sphere between -10 and +75 degree declination. Each IPS measurement is given as a `disturbance factor', g, defined as the ratio between the observed degree of scintillation and that expected under quiet conditions (Woan, 1992). These data indicate a strong correlation between scintillation level and solar wind density, N, observed at 1 AU (Houminer and Hewish, 1974; Tappin, 1986).
We will use the Cambridge data set in the construction of Carrington maps. We have available daily maps of g-values on a hour angle-declination grid of 72 by 14 points (each value may be an average over several sources at nearby positions in the sky). Each line of sight measurement is assigned a heliographic location using a `trace-back' procedure: the bulk of the IPS signal is assumed to originate at the point nearest to the Sun and traced back from that location to the solar surface assuming a solar wind speed of 400 km/s. This method has been applied to IPS velocity and heliospheric Thomson scattering observations (Rickett and Coles, 1991; Hick et al., 1991) to study the global structure of the solar wind. The width of the IPS scattering function limits the usefulness of this method to structures showing little evolution over a significant fraction of a solar rotation. A Carrington map is generated by combining sufficient daily g-maps to cover a complete solar rotation.
Only data points with solar elongation angles between 30 and 80 degree (r~0.5-1.0 AU) are used. At smaller elongations strong scattering causes a strong decrease of scintillation. Larger elongations were omitted in an attempt to avoid data affected by ionospheric scintillation. In addition all daily g-maps were checked by eye and hand-edited to remove obvious contaminations, such as metric radiation of solar origin or thunderstorm activity.
The resolution of the IPS maps is 10 degree in heliographic longitude and latitude. The g-value in each 10 by 10 degree bin is an average of typically 30 back-projected IPS observations from daily g-maps obtained over about 10 days. By implication the emphasis in these maps is on quasi-permanent structures. We are primarily addressing the structure of the quiet solar wind, which is expected to be characterized by small variations in IPS levels as compared to strong (g>1.25) transient IPS events. Consequently contour levels in the IPS maps are selected to highlight structures in the range 0.8-1.2 of g.
Transient events, which are generally not mapped accurately to the solar surface using the present procedure, are a noticeable source of error in the current maps. Since the occurrence rate of transients decreases towards the minimum of the solar cycle (Webb and Howard, 1994), we expect to be able to produce more accurate maps in the coming years.
The Yohkoh soft X-ray maps are based central meridian strips extracted from full-Sun images made by the SXT instrument (Slater et al., 1993). We `degraded' the original high-resolution Yohkoh maps by averaging over 10 by 10 degree areas in heliographic longitude and latitude in order to better match the resolution of the IPS maps (Figure 1a)
Figure 1 shows the IPS and `degraded' Yohkoh maps (rotations 1853 through 1857) selected for this analysis. The most conspicuous feature in the X-ray maps is a persistent center of activity between 330-360 degree heliographic longitude, straddling the solar equator from -25 to +25 degree latitude, which can be identified in all maps shown (Figure 1a). A few other recurring features are seen between longitudes 0 and 60 degree first north of the equator (rotations 1853 and 1854), then south of the equator (1854 -1857). The central part of the map shows a band of activity south of the equator, which changes considerably from rotation to rotation. Obviously the degrading of the Yohkoh maps masks much of the detailed changes in the topology of the active regions. In the full resolution maps the central longitudes (~180 degree) in particular show several small active region complexes, appearing, evolving and disappearing from one rotation to the next. Also shown are coronal hole boundaries as determined from He I 10830 Å observations (P. McIntosh, private communication).
The density structure of the corona, in particular as derived from the observed brightness of the K-corona, has been connected to the heliospheric current sheet by various authors. For reference we added to the X-ray maps the position of the current sheet at 2.5 solar radii as calculated from Wilcox magnetograms from a potential field model (Hoeksema et al., 1983).
The IPS map for each rotation is shown next to the corresponding X-ray map. Several features can be identified in consecutive rotations. The most noticeable is the complex between heliographic longitudes 330 and 360 degree visible in all rotations. We take this internal consistency as a strong indication that the method used to construct the IPS maps is valid and leads to a useful representation of the heliospheric structure as observed in IPS. The large-scale structure in the IPS maps, most prominent in the maps for rotations 1853 and 1857, takes the form of a band of weak IPS (g> 0.9) just south of the solar equator. The picture is more fragmented in the intermediate maps, but the equatorial enhancement is still evident over at least half the longitude range in each map. Individual features in the IPS maps show a rough agreement with features in the X-ray maps: rotation 1853 through 1855 show an IPS enhancement (g>1.1) at the same location as the recurring feature between 330 and 360 degree in the X-ray maps. Not all IPS enhancements have an obvious counterpart in the X-ray maps. For instance, the strong IPS event evident in rotation 1857 at high positive heliographic latitude, is probably associated with a CIR. The enhanced IPS is located at the leading (west) edge of a coronal hole (identified from the 10830 Å observations) and is followed by a sustained region of low g-values over the coronal hole.
We made a 2D correlation study of the IPS and Yohkoh maps. The current sheet maps (i.e. the radial component of the magnetic field at the source surface at 2.5 solar radii) was also included in these correlations. For determining whether the IPS structures match the Yohkoh X-ray structures rather than the current sheet, the first three rotations (1853-1855) are particularly useful. In these rotations the current sheet shows substantial deviations from the active region belt. The correlation coefficients for the IPS-Yohkoh comparison for the rotations in Figure 1 are r = 0.45, 0.63, 0.64, 0.49 and 0.40, respectively, with an uncertainty of sigma=0.11. In all cases the 2D cross-correlation function has a well-defined peak at an offset in the longitudinal direction varying from 0 to 30 degrees and no offset in latitude. Part of the longitudinal offset is probably a result of fixing the solar wind velocity used for tracing back the IPS data at 400 km/s. Using the area in the Carrington map where the cross-correlation function exceeds r-sigma as a measure, we find that the probability that the observed correlation coefficients result by chance is between 0.02 (for rotation 1854) and 0.10 (for rotation 1857). The correlation between IPS and radial magnetic field [we used -abs(B_radial, i.e. a positive correlation would indicate a match between B_radial=0 and high G-values] gave 0.12, -0.48, -0.15, 0.21 and 0.50. With the possible exception of rotation 1857, the IPS maps (in particular rotations 1853-1855) correlate better with the X-ray maps than with the current sheet maps. The comparison of X-ray and source surface maps gave coefficients 0.15, 0.0, 0.30, 0.16 and 0.24.
The comparison of IPS and X-ray maps for a 5-month period in 1991-1992 leads to the tentative conclusion that regions of weakly enhanced g-values map to active regions. This suggests that active regions are an important source of the small-scale (100-1000 km) density variations in the quiet solar wind, which cause weak scintillation enhancements. Assuming that the observed correlation between g and solar wind density (Tappin, 1986) holds for the weak scintillation which determines the characteristics of the Carrington maps shown here, it follows that the active regions play a role in modulating and possibly adding to the mass flow in the quiet solar wind. Uchida et al. (1992) reached a similar conclusion from Yohkoh observations. They found that expansion of the corona above active regions is ubiquitous, even in the absence of flares, and suggested that the mass loss involved may be a non-negligible fraction of the solar wind. Conventional wisdom says that the dense, hot material seen emitting in X-rays, is contained by a closed magnetic field, which does not give the active region corona access to the solar wind. Uchida's finding suggest that the active region magnetic field is not rigidly closed, but rather that the active region corona regularly expands, `shedding' its top layers. Since the observed expansion velocities observed at the final phases of the expansion (typically 10 km/s) are far less than the solar wind velocity, there is no direct proof from Yohkoh observations that the expanding material is actually expelled into the solar wind. The correlation of IPS g-levels and X-ray intensity as evidenced by our comparison of IPS and Yohkoh maps, are a first corroboration that the expanding active region material is indeed released into the solar wind. An estimate of the mass flux involved is needed to establish the extent to which active regions, thus far not considered to be a factor, play a role in the generation of the solar wind.
Hick, P.,Jackson, B.V. and Schwenn, R., Synoptic maps for the heliospheric
Thomson scattering brightness as observed by the Helios photometers,
Astron. Astrophys. 244, 242, 1991
Hoeksema, J.T., Wilcox,
J.M. and Scherrer, P.H., The structure of the heliospheric current
sheet: 1978-1982, J. Geophys. Res. 88, 9910, 1983
Houminer,
Z. and Hewish, A., Correlation of inter-planetary scintillation and
spacecraft plasma density measurements, Planetary Space Sci. 22,
1041, 1974
Rickett, B.J. and Coles, W.A., Evolution of the solar
wind structure over a solar cycle: interplanetary scintillation velocity
measurements compared with coronal observations, J. Geophys. Res. 96,
1717, 1991
Slater, G., Linford, G.A. and 12 co-authors,
The dynamics of coronal holes as determined from X-ray synoptic maps derived
from SXT imagery, abstract for 24-th SPD meeting,
BAAS 25, No. 3, 1179, 1993
Tappin, S.J., Interplanetary scintillation and plasma
density, Planetary Space Sci. 34, 93, 1986
Uchida, Y., McAllister, A., Strong, K.T., Ogawara, Y., Shimizu, T.,
Matsumoto, R. and Hudson, H.S.,
Continual expansion of the active-region corona observed by the
Yohkoh Soft X-ray Telescope, Publ. Astron. Soc. Japan 44, L155,
1992
Webb, D.F. and Howard, R.A., The solar cycle variation of coronal mass
ejections and the solar mass flux, J. Geophys. Res. 99, 4201, 1994
Woan, G., Robust estimation of interplanetary scintillation, Mon. Not. R. Astr. Soc. 254, 273, 1992