THE ASTROPHYSICAL JOURNAL, 491:L111–L114, 1997 December 20
© 1997. The American Astronomical Society. All rights reserved. Printed in U. S. A.
Up: Issue Table of Contents
Go to: Search Page | Previous Article | Next Article
Other formats: HTML (small files) | PDF (115 kb) | PostScript pages (154 kb)

Ultraviolet Coronagraph Spectrometer Observations of Density Fluctuations
in the Solar Wind

L. OFMAN

Hughes STX and NASA Goddard Space Flight Center, Laboratory for Astrophysics and Solar Physics, Code 682, Greenbelt, MD 20771

M. ROMOLI

Department of Astronomy and Space Science, University of Florence, Largo Fermi 5, 50125 Florence, Italy

G. POLETTO

Arcetri Observatory, I-50125 Florence, Italy

G. NOCI

Department of Astronomy and Space Science, University of Florence, Largo Fermi 5, 50125 Florence, Italy

AND

J. L. KOHL

Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138

Received 1997 August 29; accepted 1997 October; published 1997 November 6


ABSTRACT

     Recent Ultraviolet Coronagraph Spectrometer (UVCS) white-light channel (WLC) observations on board the Solar and Heliospheric Observatory (SOHO) indicate quasi-periodic variations in the polarized brightness (pB) in the polar coronal holes. This is the first observation of possible signatures of compressional waves high above the limb (at heliocentric distances in the range 1.9–2.45 R⊙). The Fourier power spectrum of the pB time series at 1.9 R⊙ shows significant peak at about 6 minutes and possible fluctuations on longer timescales (20–50 minutes). The observation at 1.9 R⊙ is the only currently available WLC data set with sufficient cadence to resolve the 6 minute period. These preliminary observations may result from density fluctuations caused by compressional waves propagating in polar coronal holes. We stress that our results are preliminary, and we plan future high-cadence observations in both plume and interplume regions of coronal holes. Recently, Ofman & Davila used a 2.5 D MHD model and found that Alfvén waves with an amplitude of 20–70 km s-1 at the base of the coronal hole can generate nonlinear, high-amplitude compressional waves that can contribute significantly to the acceleration of the fast solar wind. The nonlinear solitary-like waves appear as fluctuations in the density and the radial outflow velocity and contribute significantly to solar wind acceleration in open magnetic field structures. The motivation for the reported observations is the MHD model prediction.

Subject headings: MHD—solar wind—Sun: corona—Sun: magnetic fields—waves

CONTENTS

§1.  INTRODUCTION

     It has been known since the early 1970s that coronal hole regions are the sources of high-speed solar wind (Krieger, Timothy, & Roelof 1973; Neupert & Pizzo 1974; Wagner 1976; Nolte et al. 1976). Continuous, fast solar wind speed in the range 700–800 km s-1 was recently encountered by the Ulysses spacecraft above the polar coronal holes at a distance of ∼1.6 AU (Phillips et al. 1995). However, the exact mechanism that provides the additional momentum input necessary to obtain the high-speed solar wind streams is still unknown. Thermal conduction alone is not sufficient to explain the observed flow speed of the high-speed streams (see, e.g., Kopp & Holzer 1976; Holzer & Leer 1980; Leer & Holzer 1980; Davila 1985) or the observed, large-amplitude Alfvénic fluctuations in the solar wind (Smith et al. 1995). Possible sources of the additional acceleration in coronal holes are MHD waves that must get their energy from photospheric motions (see Ofman & Davila 1997a and references therein).

     Recently, Ofman & Davila (1997a, 1997b, 1997c) have developed a self-consistent, nonlinear 2.5 D MHD model of solar wind acceleration by waves and found that nonlinear compressional MHD waves are generated in a model coronal hole via Alfvén waves. The nonlinear wave shape and phase-amplitude relations are similar to that of sound solitons. These waves contribute to solar wind acceleration and can account for the additional energy input required to obtain the high-speed solar wind streams. A viable observational goal is to test the above model by detecting the presence (or absence) of compressional waves in the solar wind. The detection of waves can be accomplished by observing time-dependent density fluctuations and determining whether the amplitude and propagation speed of these waves are consistent with the model predictions.

     Here, we present the first results from the Ultraviolet Coronagraph Spectrometer (UVCS) on the Solar and Heliospheric Observatory (SOHO) white-light channel (WLC) (Kohl et al. 1995, 1997) that indicate the presence of density fluctuations in a polar coronal hole—a possible signature of compressional waves. This is the first observation of the millihertz-order frequency density fluctuations far from the solar limb (in the range 1.9–2.45 R⊙) consistent with the wave-acceleration solar wind model. We plan to perform additional observations to determine the propagation and the phase speed of these fluctuations.

§2.  OBSERVATIONAL METHOD

     The detection of compressional waves in the polar coronal holes can be accomplished by UVCS through time-resolved observations with the UVCS WLC. The WLC is a polarimeter that measures the polarized brightness (pB) in the 450–600 nm band over a 14×14 arcsec2 area located at the center of the instantaneous UVCS field of view. The coronal pB can be directly related to the electron density integrated over the line of sight. Because the UVCS field of view can be rotated about the center of the Sun, the WLC can look at different position angles. This allows us to test adjacent regions, investigating both the high- and low-density coronal structures (i.e., plumes and interplumes).

     The WLC measurements and the most recent calibration are described in Romoli et al. (1997). The purpose of the measurements described in this paper is to investigate low-frequency (millihertz) electron density fluctuations observed for several hours. The determination of the absolute magnitude of the electron density is left for future studies. The instrumental polarized stray light, which is usually removed from the pB measurement, constitutes a constant and negligible contribution to the total pB at the heights of interest. The time resolution depends on the heliocentric height of observation (it decreases with the height) and on the required accuracy of the pB measurement. In order to improve the time resolution, we assume that the pB polarization plane is tangent to the solar limb. This allows us to reduce from three to two the number of polarizer positions necessary to obtain a pB measurement. The pB counts are obtained by taking the difference between two successive positions of the polarizer.

     From the MHD model, we expect the density fluctuations due to nonlinear solitary-like compressional waves to become most evident when the solar wind is supersonic and exhibits large-amplitude, parallel velocity fluctuations modulated on top of an average solar wind velocity, in phase with significant density fluctuations. For typical parameters at the base of the coronal hole in the model (magnetic field strength B=7 G and falls off radially; temperature T=1.4×10$\mathstrut{^{6}}$ K; density n=10$\mathstrut{^{8}}$ cm-3; and the Alfvén wave amplitude is ∼40 km s-1), we get ρ$\mathstrut{_{{\rm max}}}$/ρ$\mathstrut{_{{\rm min}}}$∼2.5 at 5 R⊙. Below the sonic point, the effect of solitary waves is less apparent, but still significant with a predicted ρ$\mathstrut{_{{\rm max}}}$/ρ$\mathstrut{_{{\rm min}}}$∼v$\mathstrut{_{{\rm max}}}$/v$\mathstrut{_{{\rm min}}}$∼1.2 at ∼2 R⊙ (with the above model parameters), where ρ$\mathstrut{_{{\rm max},{\rm min}}}$ and v$\mathstrut{_{{\rm max},{\rm min}}}$ are the maximum and minimum densities and velocities, respectively. These fluctuations compare favorably with the signal-to-noise ratio of the pB measurements.

     The power spectrum of the compressional waves at the observation point will depend on the spectrum of driving Alfvén waves, on the physical parameters of the ambient plasma, and on the geometry of the structure in which the waves are propagating. In the case of a thin flux tube, it is possible to have wave-guided modes that will appear as a discrete set of frequencies in the power spectrum (see, e.g., Davila 1985; Roberts 1986; Ofman & Davila 1997a).

     The effect of the line-of-sight averaging of the white-light data is important and must be taken into account when determining the amplitude of the waves from the observations in future studies. However, the dilution due to line-of-sight integration may be reduced by looking at narrow structures such as plumes, in which, according to the theory (see, e.g., Davila 1985; Roberts 1986; Ofman & Davila 1997a), the waves propagate in phase, minimizing line-of-sight integration effects on the signal. When interplume regions are concerned, only the largest amplitude fluctuations in the line of sight (i.e., localized density enhancements) will contribute to the pB intensity, minimizing the line-of-sight dilution.

§3.  OBSERVATIONAL RESULTS

     We have made a set of preliminary observations at several heights in the range of 1.9–2.45 R⊙ in the south coronal hole, with time resolutions ranging from 1 to 5 minutes. Table 1 lists the parameters of each observation: date, starting time and length, position angle (counterclockwise from the north pole), heliocentric height, cadence of pB measurements, and the type of coronal structure (plume or interplume). The target—plume or interplume—was determined by measuring a spatial intensity profile in Lyα at 1.5 R⊙ at the beginning of the observation and extrapolating the observed structure radially outward to the heliocentric height of observation.

     The observations at 1.9 and 2.45 R⊙ exhibit spectra of fluctuations in pB with distinct peaks in the power spectrum (see below), while at 2.1 R⊙ there are no significant peaks in the pB fluctuation spectrum. The fluctuations in pB may indicate possible periodicities consistent with the presence of nonlinear compressional waves propagating in the coronal hole. However, to establish the wave origin of these fluctuations with higher confidence, we need measurements of longer duration for power spectral analysis (in particular for fluctuations in the submillihertz range), and we need to determine their phase speed.

     In Figure 1 we show the observations made at a height of 1.9 R⊙ taken with an average count rate of approximately 300 counts s-1. The top panel shows significant fluctuations of pB on a timescale of about 6 minutes, with additional fluctuations on longer and shorter timescales. The dashed lines indicate the error bars derived from Poisson statistics. In Figure 2 we show the observations taken on 1997 February 26 at 2.45 R⊙ with an average count rate of ∼130 counts s-1. It is interesting to note that the magnitude of the relative fluctuations in pB agrees well with the magnitude of ρ$\mathstrut{_{{\rm max}}}$/ρ$\mathstrut{_{{\rm min}}}$ derived from the MHD model. This is surprising, since the coronal parameters used in the model are only estimates of typical coronal values.

Fig. 1 Fig. 2

     We use the fast Fourier transform (FFT) spectral data analysis of the time series (with the DC component removed) to determine the frequency content of the fluctuations in pB. In the middle panel, we show the raw power spectrum of the pB time series. In Figure 1, the largest peak appears at a frequency of $\mathstrut{\left(2.7{\pm}0.1\right)}$×10$\mathstrut{^{-3}}$ Hz or a period of 6.2±0.3 minutes, with additional, smaller peaks at about 20 and 50 minutes. The low-frequency peaks are also apparent in Figure 2 and in both data sets for 2.45 R⊙ (Table 1). The observation at 1.9 R⊙ is the only currently available WLC data set with sufficient cadence to resolve the 6 minute period (additional, high-cadence WLC observations are planned for the near future).

     In order to test the statistical significance of these peaks, we have applied a running average of 5 points to the power spectrum and determined the plus or minus standard error interval (lower panels in Figs. 1 and 2). The 6.2 minute peak is still apparent in the lower panel of Figure 1. The longer timescale (lower frequency) fluctuations have lower statistical significance, and they appear as a general increasing trend of the power toward lower frequencies. However, the fact that the 20 and 50 minute peaks appear in the raw power spectra of the two data sets at 2.45 R⊙ is encouraging. We hope that observations of longer duration will allow us to establish the lower frequency peaks with higher statistical confidence.

§4.  CONCLUSIONS

     Preliminary results from the UVCS WLC indicate that the density in a coronal hole at 1.9 R⊙ fluctuates on a timescale of about 6 minutes, with possibly longer timescale fluctuations between 1.9 and 2.45 R⊙. This is the first observation of such fluctuations at a considerable distance above the solar limb. The temporal evolution and the corresponding power spectrum suggest that these fluctuations might be generated by compressional waves propagating from the Sun. The fact that we get a density fluctuation spectrum peaked in a narrow frequency band in the 3 mHz range strongly suggests that we have a wave-related phenomenon. The fluctuations can be generated by the nonlinear compressive effects driven by Alfvén waves. Other reasonable sources of these fluctuations might be the possible remnant of the solar p-mode oscillations that propagate into the corona in the form of fast or slow magnetosonic waves and the Alfvénic fluctuations of a thin flux tube due to footpoint motions (which could possibly be detected if the WLC is pointed near the flux-tube boundary). Additional information on the phase speed of the waves will help to determine the nature of these fluctuations.

     The observations are consistent with the predictions of the nonlinear, solitary-like wave model developed by Ofman & Davila (1997a, 1997b, 1997c). However, it is not possible to determine, based on present observations alone, whether there are nonlinear compressional waves in solar coronal holes as predicted by the model. For this purpose, we need to establish whether the fluctuations are propagating, their phase speed, and preferably the relation between the phase speed and other parameters of the plasma. The phase speed is an important indicator of the nature of the waves and could be compared to theoretical predictions if the plasma parameters (such as density, temperature, and magnetic field) are known (or if a reasonably narrow range could be estimated).

     We plan to perform additional observations with the UVCS that will help to establish the nature of the density fluctuations reported in this Letter. We hope that this Letter will stimulate coordinated observations with other SOHO instruments, such as the Large-Angle and Spectrometric Coronagraph (LASCO) and the Extreme-ultraviolet Imaging Telescope (EIT), that might help to determine whether compressional nonlinear waves are present in the solar coronal holes and, ultimately, the energy source of the acceleration of the fast solar wind.

ACKNOWLEDGMENTS

     This work is supported by the National Aeronautics and Space Administration under grant NAG5-3192 to the Smithsonian Astrophysical Observatory; by Agenzia Spaziale Italiana; by Swiss funding through ESA's PRODEX programs, national funds, and the Swiss Federal Institute of Technology Zurich; and by NASA SOHO Guest Investigator Program grant W-91558. L. O. would like to thank J. M. Davila for helpful discussions. We would like to thank the referee for helping to improve this Letter and Sarah Gibson for critical reading of the manuscript.

REFERENCES

FIGURES


Full image (67kb) | Discussion in text

     FIG. 1.—Time variation of the polarized brightness observed with the UVCS WLC in the south polar coronal hole at 1.9 R⊙ on 1997 February 24 (top) with error bars (dashed line). The integration time was 60 s for each exposure. The polarized brightness units are given relative to the Sun center brightness integrated over the WLC wavelength bandpass (450–600 nm). The raw power spectrum of the pB (middle) and the smoothed power spectrum showing the plus or minus standard error interval (bottom).


Full image (62kb) | Discussion in text

     FIG. 2.—Same as Fig. 1, but for observations taken on 1997 February 27 at 2.45 R⊙ with 300 s integration time.

TABLES

TABLE 1
PARAMETERS OF THE UVCS WLC OBSERVATIONS OF PB FLUCTUATIONS
DateStart
(UT)		Duration
(hr)		P. A.
(deg)		Distance
(R⊙)		Cadence
(s)		Location
1997 Feb 24...16:1641802.45180Plume
21:283.51841.960Interplume (Fig. 1)
1997 Feb 26...18:556.51802.45300Interplume (Fig. 2)
1997 Feb 27...17:3361872.1300Interplume

Image of typeset table | Discussion in text
Up: Issue Table of Contents
Go to: Top of This Article | Search Page | Previous Article | Next Article