Two-Dimensional Solar Wind Speeds from 6 to 26 Solar Radii in Solar Cycle 24 by Using Fourier Filtering

Measurement of the solar wind speed near the Sun is important for understanding the acceleration mechanism of the solar wind. In this Letter, we determine 2D solar wind speeds from 6 to 26 solar radii by applying Fourier motion filters to SOHO/LASCO C3 movies observed from 1999 to 2010. Our method successfully reproduces the original flow speeds in the artificially generated data as well as streamer blobs. We measure 2D solar wind speeds from one-day to one-year timescales and their variation in solar cycle 24. We find that the solar wind speeds at timescales longer than a month in the solar maximum period are relatively uniform in the azimuthal direction, while they are clearly bimodal in the minimum period, as expected from the Ulysses observations and interplanetary radio scintillation reconstruction. The bimodal structure appears at around 2006, becomes most distinctive in 2009, and abruptly disappears in 2010. The radial evolution of the solar wind speeds resembles the Parker’s solar wind solution.

Figure 1

Height-time images of artificially generated moving intensity features [white in panel (a)], streamer blobs observed by LASCO C3 from September 26, 2007 to September 28, 2007 at an azimuthal angle equal to 334° [white in panel (c)], and comparisons of the estimated speeds by applying a Fourier motion filter with true speeds [(b) and (d)]. In the right panels, the estimated speeds are indicated by black solid lines. The true speeds are indicated by blue, green, yellow, and red lines. The grey vertical bars indicated in the right panels are the median absolute deviations during three days at each height.

Figure 2

Comparisons of the solar wind speeds obtained from LASCO C3 with IPS speeds (a) and radial speeds of protons given by a Doppler dimming technique below “six” solar radii (b). In panel (a), the IPS speeds are interpolated from the projected speeds [65] based on the position and time of our data sets covering whole latitudes. In panel (b), the radial speeds of protons are measured in the extended corona above coronal holes that have been taken from [38], and the estimated electron speed in 6–26 solar radii is taken from the North and South Poles. The grey area indicates $1\text{−}\sigma$ interval.

Figure 3

Maps of median solar wind speeds in 6–26 solar radii over several time periods (one day to one year) in 2000 (a)–(d) and 2009 (e)–(h). The one-day maps are constructed by taking the median solar wind speed at each position during 00–24 UT on October 4, 2000 (a) and March 7, 2009 (e). The one-week maps are constructed by taking the median solar wind speed at each position during October 4, 2000 12 $\mathrm{UT}\pm 3.5$ days (b) and March 7, 2009 12 $\mathrm{UT}\pm 3.5$ days (f). The one-month maps are constructed from maps of daily solar wind speed in July 2000 (c) and February 2009 (g). The one-year maps are constructed from maps of monthly solar wind speed in 2000 (d) and 2009 (h).

Figure 4

Maps of yearly median solar wind speeds from 1999 to 2010. The two dotted lines in 2000 and 2009 indicate $\pm 20°$ from the North and South Poles. The two dashed lines in 2000 and 2009 indicate $\pm 20°$ from the equator.

Figure 5

Radial profiles of equatorial and polar solar wind speeds in the one-year maps of 2000 and 2009. The black lines represent the observed speeds averaged over $\pm 20°$ from the equator and poles indicated by dashed and dotted lines in Fig. 4, respectively. The vertical bars are the corresponding standard deviations. The red lines represent the model speeds obtained from the least squares fit by the function defined as $v_0\sqrt{1-e^{-(r-r_1)/r_0}}$.