2. Data Analysis and Results

[6]  Coronal holes are usually related to the regions of open configurations of the magnetic field [Levine, 1982; Obridko and Shelting, 1999] what are sources of the solar wind high-speed streams [Nolte et al., 1976]. However, not every passage of an isolated coronal hole over the solar disk center causes high-speed solar wind stream. Probably this may be explained by the fact that not all coronal holes have the open configuration of magnetic fields. For example, Stepanyan [1995] and Stepanyan and Malanushenko [1995] studied 51 coronal holes and showed that 39 (76%) out of them have coinciding signs of the magnetic field at the levels of the photosphere and source surface ( 2.5 Rodot ), whereas 12 holes have opposite signs of the field. Analyzing variations in the coronal holes area, they show also that the area of some coronal holes decreases with height.

[7]  In this paper the period of 1998-1999 is analyzed using the Kitt Peak full-disk daily solar magnetograms, maps of coronal holes locations in the He I 10830 Å line from the same observatory, and the satellite database of solar wind parameters OMNI. Coronal holes in X ray are identified as dark regions with decreased density in images obtained by Yohkoh satellite in soft X-ray wavelength. The comparison of the values of the speed, density, and other parameters of the solar wind for the 1996-2002 period shows that solar wind parameters vary within the solar cycle. The character of these variations is different at different stages of the growth phase of solar activity and manifests the dynamics of the global magnetic fields of the Sun [Bilenko, 2002]. Detailed consideration of particular events does not always reveal the relation between solar wind parameter variations at the Earth orbit and particular coronal holes, active regions, or other phenomena in the solar atmosphere. The general variations in daily parameters of the solar wind better correspond to the variations in coronal holes and the absolute value of the average magnetic field of the Sun. The number of coronal holes (the source regions of the solar wind high-speed streams) increases with an increase of solar activity [Bilenko, 2001].

[8]  Ivanov [1987] revealed a relation between the large-scale magnetic field on the Sun and the interplanetary magnetic field at the Earth's orbit. Localization of the source regions of the solar wind high-speed (as well as of low-speed) streams is difficult because the real variations in the speed during flow motion from the Sun to the Earth are not known. Moreover, the plasma outflow in the corona is not radial. Neither currently existing model provides the localization accuracy better than 10o. There are several models, which are used to predict solar wind parameters at the Earth orbit using the solar photospheric magnetic fields observations. In such models magnetic field distribution is calculated at the so-called solar surface and solar wind speed is considered a constant and radial [Schatten et al., 1969]. Another model used heliospheric current sheet [Schatten, 1971]. Detailed analyses of different models and comparison with observations were carried out by Neugebauer et al. [1998]. Investigation of discrepancies in the predictions of solar wind parameters with the use of potential field source surface model was performed by Poduval and Zhao [2004]. Different authors used different transit time to map the source of the solar wind speed. Poduval and Zhao show that location of solar wind source can differ by 25o in longitude, depending on the value of the used speed. The accuracy is influenced by the parameters of the model ( N max, for example). Source locations calculated using different N max differ by about 36o in longitude and 18o in latitude. All these can result in the fact that the calculated location on photosphere would not have any relation to the location of a source of the given stream of the solar wind. The next rough assumption is that the photospheric magnetic fields are radial. Near-Sun flow of solar wind is seldom radial [Neugebauer et al., 1998].

[9]  However, the studies of the character and features of the coronal hole location on the basis of the Yohkoh data and observations in the He I 10830 Å line and also of the coronal hole relations to the photosphere magnetic field and to parameters of the solar wind at the Earth orbit make it possible to reveal some regularities.

[10]  Statistical studies of correlations between the solar wind stream arrival time and coronal holes distribution over the solar disk show that about 3-4.5 days are needed. Arrival time depends on the speed of the solar wind stream. Sheeley et al. [1976] found that the high-speed solar wind streams registered near the Earth orbit are connected with equatorial coronal holes. Time of arrival of the streams is equal approximately to three days concerning date of coronal hole central meridian passage.

[11]  Figures 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 show the solar images. Figures 1a-10a are full-disk magnetograms combined with coronal holes maps in the He I 10830 Å line. White and black color indicate the positive and negative polarity of photospheric magnetic fields, respectively. Coronal holes marked as CH for each day. Active regions, the bipolar regions in magnetograms with increased magnetic field strength, are marked using maps of the Mess Solar Observatory according to their NOAA number. Images in soft X ray are taken from Yohkoh data (Figures 1b-10b). Figures 1c-10c show parameters of the solar wind at the Earth orbit: the flow speed (Figures 1c-10c, top), density (Figures 1c-10c, middle) and magnetic field strength (Figures 1c-10c, bottom) corresponding to the time period of particular coronal hole influence.

[12]  It follows from the analysis of the mutual position of coronal holes according to the Yohkoh X-ray measurements and observations in the He I 10830 Å line that the coronal holes registered in the He I 10830 Å line always are located in unipolar magnetic regions. They are able to cover both the entire unipolar region on the photosphere (Figures 2 and 10) or only part of it (Figures 6 and 8). The areas of coronal holes in the He I line, as a rule, are less than or equal to isolated Yohkoh data coronal holes. The shape and area of coronal holes in the He I line vary considerably from one day to another. However, they do not exceed the coronal holes registered in X ray and stay within the same unipolar region. Coronal holes are often observed only in the X-ray images and only some time after they are registered in the He I line.

[13]  Figure 1 shows a large X-ray coronal hole passage through the solar disk center. A small part of it coincides with a coronal hole identified in He I line (CH1). The He I coronal hole is associated with the negative polarity magnetic field of the enhanced bipolar magnetic field structure. Solar wind speed rises up to 700 km s-1. X-ray coronal hole passed the central meridian approximately from 29 September to 5 October, but the high-speed solar wind stream was observed from 1 October to 4 October. The second increase in solar wind corresponds to the magnetic cloud ( Bx, By, and Bz components of magnetic field have a typical behavior). Figure 2 also shows an X-ray coronal hole passage through the solar disk center. Also a small part of it coincides with a coronal hole identified in He I line (CH1). The He I coronal hole is associated with the negative polarity photospheric magnetic field. CH2 is the second coronal hole, which is located near the south pole. It is also associated with the negative polarity photospheric magnetic field. Both these coronal holes are covered by one X-ray coronal hole which is connected with the south negative polarity solar pole. Variations of the solar wind parameters show that not all X-ray coronal holes but only part of it, which correspond to the CH1, may cause the high-speed solar wind stream observed during 28-31 October. The magnetic field strength also increases. Slight increase in density on 28 October manifests an interaction of high and low solar wind streams. There were no active regions at that time that can produce such changes.

[14]  Within one coronal hole observed in X ray, several coronal holes in the He I line may be observed (Figures 3, 4, 5, 6, 8, and 9). In such a case they, as a rule, correspond to the same unipolar region of the photosphere magnetic field. Coronal holes detected in X ray have more stable regular form and, as a rule, larger area than the holes detected in the He I 10830 Å line. In the cases when the areas according to the X ray and He I line almost coincide, an increase of the solar wind speed up to the maximum values and a depletion of the flux density are observed during the entire period of the coronal hole passage. For example see CH2 in Figure 10.

[15]  Some authors paid a special attention to the boundaries of coronal holes because they play an important role in the process of formation of both high-speed and low-speed solar wind streams. According to Kozlova and Somov [1998, 2000] a prevailing upward flows in a coronal hole and at the boundaries of the quiet network with a velocity up to 2.3 km s -1 are observed. According to their observations, "dark points" with a shape of chains surrounding coronal holes are also detected in the He I line. The observations in the Ha line show an increase in the radial velocities of the substance ascent in the vicinity of the dark points as compared to the velocities within the coronal hole and in quiet regions. Kozlova and Somov [1998, 2000] assumed that the increase and acceleration of the flows arise as a result of the reconnection of oppositely directed magnetic fields of dipoles of the chromospheric network and the unipolar magnetic field within the coronal hole.

[16]  Wang [1994] indicated two types of the low-speed solar wind. The first type is related to a sharp change in the magnetic field values at the boundaries of large coronal holes. The second type is formed over small coronal holes. According to this study, high-speed solar wind streams are generated not at coronal holes boundaries, but in the places where locations of coronal holes in He I line and X ray coincide.

[17]  Analyzing the Yohkoh satellite data, Kahler and Hudson [2002] detected three types of coronal hole boundaries in the soft X-ray images: diffusive boundaries not determined clearly; sharp and clear boundaries of the coronal holes located near matching-polarity active regions and adjoining magnetic field of these active regions with coinciding polarity; and loopy boundaries of coronal holes observed near opposite-polarity magnetic field of active regions.

[18]  The consideration of the features of the solar wind parameters and the structure of the magnetic field during the periods of observations on the disk of the equatorial large-area coronal holes and longitude-aligned and also of the transitional processes corresponding to the above mentioned three types of coronal hole boundaries shows that the solar wind parameters in a certain degree manifest these features.

[19]  Smooth increase and decrease of the solar wind speed and lower values of the latter are typical for the regions of irregular boundaries. For example, in Figure 2, X-ray coronal hole boundaries are diffusive and solar wind speed increases and decreases smoothly. It reaches ~580 km s-1. In the cases when the X-ray coronal hole coincide with He I coronal hole and is adjacent to active regions (in Figure 4, X-ray coronal hole, He I coronal hole CH2, and active region NOAA 8540) an increase of the solar wind speed up to values of 700 km s -1 and a sharp increase of the flux density and magnetic field strength are observed. Similar evolution is observed when X-ray coronal hole is adjacent to active region. In Figure 8, X-ray coronal hole is adjacent to NOAA 8627 and NOAA 8631. In such cases solar wind can rise as a result of the reconnection of open magnetic field lines with closed magnetic loops. It this case the model proposed by Fisk [2003] can be realized. The speed values of 300-500 km s-1 and enhanced values of the flux density are typical for the loop-like boundaries of X-ray coronal hole (Figures 7 and 9). This corresponds to the Wang et al. [1998] model in which the most low-speed and dense streams are formed at the top of the streamer arcs mainly from the loop substance at the boundaries of coronal holes. Schwadron et al. [1999] and Wang et al. [1996] proposed models where the low-speed streams are formed from the substance accumulated in the loops of active regions. The energy is released due to the mechanism of reconnection of the closed field of these loops and the regions of open configurations. If a coronal hole or its part is covered by arcs high-speed streams are not formed. One can see this in the X-ray image in Figure 7. High-speed streams are observed outside the arch structures, whereas in the arch zone some decrease of the speed and increase of the density are observed.

[20]  One can see that the regions of the coronal holes registered in the He I 10830 Å line correspond to the zones of increased magnetic field strength of the polarity dominating on the photosphere for the given coronal hole. The following parameters are typical for these regions. The magnitude of the mean magnetic field strength for the elements of the dominating and opposite polarity is about 20-40 G and 5-10 G, respectively. This magnitude for the magnetic elements with the magnetic field strength above 20 G is of about 60-70 G and 30-40 G for the dominating and opposite polarity, respectively. The maximum values of the magnetic field strength in these regions are about 300-500 G. For the regions of coronal holes registered only by Yohkoh the disbalance of magnetic fields is either less pronounced or is not observed at all. For example, for the coronal hole observed on 28 February 1999 (Figure 5) three separated coronal holes in the He I line are located within one unipolar region. Also, the ratio of the magnetic field strengths of the dominating (in this case, negative) polarity to the strength of the opposite (positive) polarity in the regions registered only in X ray is by the absolute value approximately 10-47 G to 7-10 G. For the magnetic elements with the magnetic field strength above 20 G the ratio is 45-65 G to 40-50 G. The maximum values of the magnetic field strength in this regions is of the order of 200-300 G. For some coronal holes no disbalance of magnetic field is observed for the X-ray coronal hole part.

[21]  The studies performed earlier [Bilenko and Kononovich, 1999] showed that in the regions of the coronal hole location registered in the He I 10830 Å line the structure of the photosphere magnetic field differs considerably from the structure of quiet regions, i.e., regions outside coronal holes. For example, the dimensions of photosphere magnetic network elements in the regions of location of coronal holes in the He I line exceed dimensions of the corresponding elements in quiet regions by a factor of 2-4. The mean dimensions of the network elements of the dominating polarity exceed the dimensions of elements of the opposite polarity by a factor of 5-10, the difference increasing with an increase of solar activity. The total magnetic flux in the regions of location of the coronal holes registered in the He I line is by a factor of 2-3 higher than in the adjoining quiet regions, the disbalance of magnetic fluxes also increasing with solar activity. In the regions of location of the coronal holes in the He I line a constant upward component of the radial velocity of the network elements substance of the order of tens kilometers per second is observed at the photosphere level.

[22]  Quasiperiodic line-of-sight velocity variations were also detected by Kobanov and Makarchik [2003] at the bases of the solar polar coronal holes. Upward flows with radial velocities reaching 3 km s-1 in the photosphere and 12-15 km s-1 in the chromosphere were observed near dark points at the boundaries of the chromospheric network.

[23]  All this testifies that all layers of a solar atmosphere are involved in process of formation and acceleration of high-speed streams of the solar wind down to the photosphere. According to investigations of Grall et al. [1996], acceleration of the polar solar wind is almost completed to 10 Rodot.


AGU

Powered by TeXWeb (Win32, v.2.0).