Influence of Polar Coronal Holes on Solar Wind Characteristics at the Activity Minimum between Solar Cycles 24 and 25

Abstract

The SDO/AIA 193 Å and ACE/SWEPAM data acquired in 2019 are used to compare the effects of polar and equatorial coronal holes (CHs) on solar wind (SW) characteristics under conditions of low solar activity. As expected, most geomagnetic storms in this period were caused by high-speed SW streams (${>}500{-}600$ km s${}^{-1}$) originating from equatorial CHs. At the same time, it has been shown that at a deep solar activity minimum polar CHs can exert a noticeable influence on SW characteristics at the Earth’s orbit. A correlation with a correlation coefficient of 0.8 has been found for the integrated polar CH area and the SW speed in the investigated period. The southern polar CH, for which a correlation with the SW speed with a correlation coefficient of 0.82 was found in the spring of 2019 (in the period when the south solar pole was maximally tilted to the Earth), exerted a particularly significant influence on the SW speed. The northern polar CH had virtually no effect on the SW speed. An anticorrelation of the polar CH area with the SW speed at the Earth’s orbit was found in the fall of 2019, in the period when the north solar pole was tilted to the Earth. We discuss a possible mechanism for the influence of polar CHs on SW characteristics and propose an interpretation of the results obtained.

APPENDIX

The question about the determination of the CH boundaries in solar images is very difficult. As a rule, CHs have a complex configuration and blurred boundaries with a smooth gradient between the reduced brightness of the emission inside the CH and its brightness in the surrounding corona. As a result, both the CH boundaries and, as a consequence, their areas depend significantly on the method being applied, especially on the choice of the brightness threshold separating the CH from the surrounding corona.

Since during our study we needed to process more than a thousand SDO/AIA 193 Å images, we decided to develop our own automatic algorithm with a high data processing speed. Note that the possible inaccuracies in determining the CH boundaries, which are characteristic of any method, were partially averaged in our study precisely owing to the large number of processed frames.

The principle of image histogram analysis, which has found wide application in analyzing solar images (see, e.g., the paper by Krista and Gallagher (2009) on the detection of CHs based on SOHO/EIT 195 Å data), underlies the algorithm used by us. The software of the algorithm was written in the IDL programming language.

The procedure for finding the boundaries and measuring the CH area consisted of the following steps shown in Figs. 5–7.

Fig. 7

The CH detection method, step 3: the solar disk with the superimposed CH boundaries (left); the arbitrary regions for the CH division into polar and equatorial ones (right) (for more details, see the text).

(1) In the first step shown in Fig. 5, we aligned the images based on the histogram by varying their brightness and contrast. The necessity of the procedure stems from the fact that the SDO level 1 FITS files are initially subjected to minimal processing and, as a consequence, can differ greatly in brightness and contrast, making a further image processing with automatic algorithms difficult. In this step, we also cut out the solar disk by superimposing a mask that filled the entire space outside the solar disk with the black color (the intensity is 0). The solar disk radius was extracted from the header of the corresponding FITS file. The result of applying the corresponding procedures can be seen by comparing the left (the original image) and right (the resulting image) panels in Fig. 5.

(2) In the second step, we produced the sample of pixels belonging to the CHs. For this purpose, we first separated out the part of the histogram corresponding to intensities above zero and below some threshold T. The corresponding threshold value was selected by an empirical method, including a comparison of the results with other known CH detection algorithms, for example, CHIMERA (solarmonitor.org) and SPOCA (helioviewer.org). Next, we assigned 255 (white color) to the pixels corresponding to this part of the histogram and 0 to the remaining pixels. In this way the original image was converted to the binary black-and-white image shown on the left panel in Fig. 6. Thereafter, this image was improved by applying the mathematical expansion and contraction operations, which allowed most of the small ‘‘noise features’’ to be removed. Then, we finally processed the mask from which the remaining small features unrelated to the CHs were removed. The result is shown on the right panel in Fig. 6.

(3) In the third step, we divided the CHs into polar and equatorial ones and excluded the CHs far from the central solar meridian from consideration, because they do not affect the SW parameters. For this purpose, the CH boundaries found in the previous steps were superimposed back on the solar disk (the left panel in Fig. 7) and their positions were compared with the calibration grid shown on the right panel. The inner squares of the grid have horizontal and vertical sizes of 2/3 and 3/4 of the solar radius, respectively. The CH or CH fragments falling into this region were deemed equatorial. The CHs falling into the upper and lower regions (correspond to the latitudes northward and southward 70${}^{\circ}$–75${}^{\circ}$ of the solar equator) were deemed belonging to the northern and southern CHs. The regions on the right and the left (correspond to longitudes ${\pm}60^{\circ}$ from the central meridian) were excluded from consideration, because a preliminary analysis showed that the CHs located here do not affect the SW.

(4) In the last step, for each CH we calculated its area in sq. km by taking into account the disk projection onto the hemisphere. In addition, for each CH we calculated the ‘‘center-of-mass’’ position. The results of our CH area measurements and other information were written into files for a further analysis.