An Erupting Active Region Filament: Three-Dimensional Trajectory and Hydrogen Column Density

From 15:33 through 16:02 UT on 13 June 1998, observations of an erupting filament as it crossed solar disk center were obtained with the NSO/KPVT and SOHO/CDS instruments as part of the SOHO Joint Observing Program 70. Context observations show that this event was the eruption of the north-east section of a small active region filament associated with NOAA 8237, that the photospheric magnetic field was changing in this active region between 12–14 June 1998, and that a coronal Moreton-wave disk event occurred, as well as a white-light CME off the south-west solar limb. The NSO/KPVT imaging spectroscopy data covered 512 × 512 arc sec of the disk center and were spectrally centered at the Hei 1083 nm line and captured ±1.0 nm of surrounding solar spectrum. The Hei absorption line is seen blue-shifted to velocities of between 200 and 300 km s^–1. The true solar trajectory of the eruption is obtained by using the projected solar coordinates and by integrating the Doppler velocity. The filament travels with a total velocity of about 300 km s^–1 along a path inclined roughly 49 deg to the solar surface and rises to a height of just over 1.5 solar radii before it becomes too diffuse to follow. The filament also shows internal motions with multiple Doppler components shifted by ±25 km s^–1. Finally, the KPVT data show no Stokes V profiles in the Doppler-shifted Hei 1083.03 nm absorption to a limit of roughly 3×10^–3 times the continuum intensity. The SOHO/CDS scanned the center of the KPVT FOV using seven EUV lines; Doppler-shifted filament emission is seen in lines from Hei 58.4 nm, Heii 30.4 nm, Oiv 55.5 nm, Ov 63.0 nm, Nevi 56.3 nm, and Mgx 61.0 nm representing temperatures from about 2×10⁴K through 1×10⁶K. Bound-free continuum absorption from Hi, without confusion from foreground emission and line emission, is seen as the filament obscures underlying chromospheric emission. A fit to the wavelength dependence of the absorption from five lines between 55.5 to 63.0 nm yields a column density _H I =4.8±2.5×10¹⁷ cm^–2. Spatial maps show that this filament absorption is more confined than the regions which show emission.     

Comparison of Prominences in Hα and He II 304 Å

In this letter, we bring attention to prominences which show different morphology in H and Heii 304 Å, as observed simultaneously by BBSO and EIT on board SOHO. Those two lines have been thought to represent similar chromospheric structures although they are formed at significantly different temperatures. We give two examples representing two kinds of anomaly: (1) prominences showing strong H emissions in the lower part and strong Heii emissions in the upper part, and (2) erupting prominences showing extensive Heii emission, but nothing in H. Our results indicate that a part or the whole of a prominence may be too hot to emit H radiation, possibly due to heating or thermal instability. Please note that these are not just two isolated cases, many other prominences show the similar differences in H and Heii 304 Å.     

CORONAL RADIATION AND HELIUM λ584 EMISSION IN ACTIVE REGIONS

Based on EUV observations of eleven sunspot regions obtained with the Coronal Diagnostic Spectrometer, CDS, on SOHO we have studied the spatial distribution, temporal variation and wavelength shift of the Hei 584 line. We find a relatively high spatial correlation between the coronal line Fexvi 360 and the Hei 584 line. This points to coronal back-radiation as an important contributor to the formation of the Hei line in active regions. However, contribution to the line formation from another source is suggested by the following two findings: First, the red-shifted line profiles of both Hei 584 and the transition region lines tend to be more intense than blue-shifted profiles. Second, the Hei 584 emission changes significantly faster than the coronal line emission.     

Multi-Wavelength Observation of the 3n/x3.3 Flare of 28 November 1998

The 3N/X3.3 flare of 28 November 1998 was observed in multiple wavelength simultaneously. The available data include H images, spectra in Hei 1083 nm and Caii 854.2 nm from Purple Mountain Observatory (PMO), soft X-ray (SXR) and hard X-ray (HXR) images and flux from Yohkoh. Morphological relationship investigation and spectral analysis of these data show: (1) The sudden brightening at loop top above the active region and the steep increase of SXR flux before flare onset suggest that the corona there had already been heated to some extent in the preflare phase. (2) The scales of the Caii 854.2 nm emission areas are very similar to those of the H line, but the emission profiles look like those of the Caii K line. Most of the Hei 1083 nm emissions exist in the bright H kernels and can last to the decay phase. (3) Flare spectra show that line shift and asymmetry are very common in this flare not only in the impulsive phase but also in the decay phase. The difference in the line shifts or asymmetry between Caii 854.2 nm and Hei 1083 nm, as well as the difference between the line center and wings of Caii 854.2 nm imply the existence of a velocity gradient in the line-of-sight direction. (4) Post-flare loops with very deep absorption (70%) and very-high-velocity red shifts (30–90 km s^–1) were observed in Hei 1083 nm during the decay phase. However, only a slight dip can be found in the Caii 854.2 nm profile.     

EUV Full-Sun Imaged Spectral Atlas Using the SOHO Coronal Diagnostic Spectrometer

The Coronal Diagnostic Spectrometer (CDS) aboard the Solar and Heliospheric Observatory (SOHO) carries out a regular program of measuring the full-disk irradiance using the Normal Incidence Spectrograph (NIS). The full-disk solar spectrum is returned in the wavelength bands 308–379 Å and 513–633 Å, with a spectral resolution between 0.3 and 0.6 Å. A recent modification to the CDS on-board software allows simultaneous moderate resolution monochromatic images to be made of the stronger lines in these wavelength ranges. We report on observations made 23 April 1998, 21 May 1998, and 22 June 1998. A total of 69 monochromatic full-Sun images are extracted from the spectral line data. For the first time, spectrally resolved images of the full Sun in Heii 303.8 Å and Sixi 303.3 Å are presented and compared. Velocity maps of the Sun in singly ionized helium are presented. Correlations of intensity to velocity over a wide range of transition region and coronal temperatures are shown. Lines from Hei to Fexiv show statistical red shifts of 1–7 km s^–1 between active regions and quiet Sun areas. Velocity maps of Mgix andx are presented, showing strong upflow and downflow regions associated with active regions, but not correlated with the brightest emission. Changes in line width are also presented in Hei, with discussion of similar features in other lines of comparable temperature. Corrections which need to be applied to CDS/NIS data to extract meaningful velocities and line widths are presented and discussed. The identifications of the lines in the CDS spectrum are examined. The spatial and spectral variation of the background component of the CDS spectrum is examined.     

Filament channel structures in a Si iv line related to a 3d magnetic model

A recent 3D magnetic model of filament support (Aulanier and Démoulin, 1998) has shown that specific morphologies derived from the model, based on SOHO/MDI magnetograms, match quite well with the observations of a filament observed in H and Caii lines with the German telescope VTT in Tenerife on 25 September 1996 (Aulanier et al., 1998, 1999a). Some predictions of this model concern the filament channel. To continue the comparison of model and data, we have investigated the same filament region observed in ultraviolet by the SOHO spectrometers SUMER and CDS. The elongated EUV fine structures in the filament channel observed in the Siiv 1393.76 Å line by SUMER have similar orientations and locations to features predicted by the model of Aulanier et al. (1999a). These regions are near the bases of field lines which tangentially join to the photosphere in so called 'bald patches' and are parts of large arcades above the filament channel. In addition, we consider the Siiv Doppler shifts in these structures and compare them to what might be expected from the model field structure. Our study also suggests that the filament has a very low opacity in Siiv, lower than that of the Ov line observed by CDS.     

Infrared Spectroscopy from San Fernando Observatory: He I 1083 nm, O I 1316 nm, and Fe I 1565 nm

Imaging spectroscopy of the Sun was carried out at the California State University Northridge San Fernando Observatory using an InGaAs near-IR video camera. Using the Sii 1082.71 nm and Hei 1083.03 nm lines the Evershed effect is measured simultaneously in the photosphere and the chromosphere for three sunspots; the speed of the Evershed flow is measured to be between 3 to 8 times greater in the Hei line than in the Sii line, and the direction is radially inward in the chromosphere and outward in the photosphere. Telluric absorption lines prevented a meaningful measurement of Oi 1128.7 nm limb emission, but an upper limit of 20×10^–3 B is measured for chromospheric limb emission at Oi 1316.3 nm. Zeeman splitting in Fei 1564.9 nm was observed in six sunspot umbrae, and a linear relationship between magnetic field and umbral continuum intensity is confirmed.     

Earth Tides and Lense–Thirring Effect

The general relativistic Lense—Thirring effect can be measured by inspecting a suitable combination of the orbital residuals of the nodes of LAGEOS and LAGEOS II and the perigee of LAGEOS II. The solid and ocean Earth tides affect the recovery of the parameter by means of which the gravitomagnetic signal is accounted for in the combined residuals. Thus an extensive analysis of the perturbations induced on these orbital elements by the solid and ocean Earth tides is carried out. It involves the l=2 terms for the solid tides and the l=2,3,4 terms for the ocean tides. The perigee of LAGEOS II turns out to be very sensitive to the l=3 part of the ocean tidal spectrum, contrary to the nodes of LAGEOS and LAGEOS II. The uncertainty in the solid tidal perturbations, mainly due to the Love number k ₂, ranges from 0.4% to 1.5%, while the ocean tides are uncertain at 5–15% level. The obtained results are used in order to check in a preliminary way which tidal constituents the Lense-Thirring shift is sensitive to. In particular it is tested if the semisecular 18.6-year zonal tide really does not affect the combined residuals. It turns out that, if modeled at the level of accuracy worked out in the paper, the l=2,4 m=0 and also, to a lesser extent, the l=3, m=0 tidal perturbations cancel out.This revised version was published online in October 2005 with corrections to the Cover Date.     

Solar Wind Velocity and Anisotropic Coronal Kinetic Temperature Measured with the O Vi Doublet Ratio

Doppler dimming of the Ovi resonance lines (1032 Å, 1037 Å) in an expanding corona is calculated including the pumping effect on the Ovi 1037.61 Å of both Cii lines at 1036.34 Å and 1037.02 Å, and the effect of the width of the absorption profiles of the coronal oxygen ions along the incident radiation. The pumping effect of the Cii line at 1036.34 Å allows us to extend to approximately 450 km s–1 the measurement of solar wind velocities with the Ovi line ratio technique. Since the emissivity ratio of the Ovi doublet depends on the width of the oxygen coronal absorbing profiles, this ratio can provide an accurate measurement of the solar wind velocity in the case that the width of the absorbing profile along the direction of the incident radiation is independently determined. However, if on the one hand the ratio of the emissivities of the Ovi doublet has limitations in probing the wind velocity, on the other hand it can be used as a diagnostics for inferring the velocity distribution of the coronal Ovi ions along the radial, and detecting possible velocity anisotropies. This diagnostics, applied to recent observational results, allows us to infer that the velocity distribution of the oxygen ions is much broader in the direction perpendicular to the magnetic field direction, and that the acceleration of the fast solar wind in the first 2 solar radii is high.     

Crossing of Various Cantori

We find the form of cantori surrounding an island of stable motion in the standard map for various values of the nonlinearity parameter K near the value K=5 (much larger than the critical value K _cr=0.971635...). The asymptotic curves of unstable periodic orbits inside the cantorus cross it after a certain time and then escape to the large chaotic sea. For K=5 the crossing time (in appropriate units) is t=1 and the escape time is t=2. For K=4.998 the crossing time is t=7 and the escape time t=23000. This delay of escape is due to the existence of higher order cantori, with very small gaps. We found that, as K increases the noble torus [2,4,1,1,..] is destroyed before the destruction of the higher order tori [2,4,1,1,1,1,2,1,...] and [2,4,1,1,1,1,3,1,...]. Thus the torus with the simplest noble number is not the last KAM curve to be destroyed. Then we find that nearby orbits deviate considerably, but the average times spent near various resonance before escape are very similar.     

Structure and Evolution of the Transition Region Network Observed in he ii λ304

The length scale and life time of the transition region network cells were studied using Heii 304 filtergrams. The temporal structure function was calculated from spatially aligned Heii 304 images. The estimated life time of the network cell was about 27 hr. We compared this life time with the life time of photospheric magnetic network and of the extrapolated magnetic network. The spatial structure function was calculated from the Heii 304 filtergrams. The calculated spatial structure function saturates at 25000 km. The transition region network elements are bigger in size than the photospheric magnetic network element. The magnetic network element equals the size of the Heii 304 network element when the photospheric magnetic field is extrapolated to a height of 3000 km above the photosphere where the magnetic fluxes are deployed. The derived value of the diffusion speed of the network elements was 0.098 km s^–1.     

Time Evolution of low-Frequency Periodicities in Cosmic ray Intensity

The long-time series of daily means of cosmic-ray intensity observed by four neutron monitors at different cutoff rigidities (Calgary, Climax, Lomnický tít and Huancayo/Haleakala) were analyzed by means of the wavelet transform method in the period range 60 to 1000 days. The contributions of the time evolution of three quasi-periodic cosmic-ray signals (150 d, 1.3 yr and 1.7 yr) to the global one are obtained. While the 1.7-yr quasi-periodicity, the most remarkable one in the studied interval, strongly contributes to the cosmic ray intensity profile of solar cycle 21 (particularly in 1982), the 1.3-yr one, which is better correlated with the same periodicity of the interplanetary magnetic field strength, is present as a characteristic feature for the decreasing phases of the cycles 20 and 22. Transitions between these quasi-periodicities are seen in the wavelet power spectra plots. Obtained results support the claimed difference in the solar activity evolution during odd and even solar activity cycles.     

Investigation of He I 10 830 Å `Dark Points' at the Sayan Solar Observatory and the Baikal Astrophysical Observatory

A new two-bandpass birefringent filter has been produced at ISTP, Irkutsk for the investigation of the fine structure of the chromosphere. One filter passband is centered on the Hei 10830 Å line, the second one is centered on H. The FWHM of the Hei 10830 Å passband is 0.46 Å and of the H passband is 0.3 Å. A large number of filtergrams were obtained with the filter at the Sayan observatory. At the same time, spectral observations with high spatial and spectral resolution were carried out by the large solar vacuum telescope at the Baikal Observatory. We selected 29 `dark point' spectra with sizes from 2 to 13, as well as `dark points' on the filtergrams. Comparison of spectrograms and filtergrams has shown a good agreement of their size and intensity in relation with the surrounding chromosphere as well as the absence of primary line-of-sight velocities in both observation types. From spectral observations, the depth of 10830 Å is over 30% for some `dark points', and the FWHM is more than 1 Å. Hei 10830 Å line profiles in `dark points' are more deep and wide than in quiet regions. The optical depth of the chromosphere in `dark points' is estimated. Comparison with the unperturbed chromosphere showed that `dark points' in Hei 10830 Å are more optically thin than the nearby chromosphere.     

Wavelength Shifts in the Solar Photospheric Spectrum

Wavelength shifts converted to velocities between solar lines observed at disc center and laboratory wavelengths of Fei, Feii, Tii, Nii, and Fei lines in the near infrared are plotted as a function of the logarithm of their solar equivalent width in milliångstroms. The need for wavelengths based on the wavelength standards is stressed. A comparison of photographic Fei solar wavelength is shown to agree, on the average, with Fourier Transform Spectrometer solar wavelengths within less than 0.5 milliångstroms. Using Balthasar's limb effect tables we convert the disc center velocities to limb velocities and find, though the scatter is large, that there is little evidence for a super-gravitational red shift.     

Statistical Characteristics of CMEs and Flares Associated with Solar Type II Radio Bursts

We present results from a study of sunspots and faculae on continuum and Caii K images taken at the San Fernando Observatory (SFO) during 1989–1992; a total of approximately 800 images in each bandpass were used. About 18000 red sunspots, 147000 red faculae, and 800000 Caii K faculae were identified based on their contrasts. In addition, we computed the contrasts of pixels on the red images cospatial with Caii K faculae. Sunspot contrasts show a strong dependence on size but no dependence on heliocentric angle. There are continuous but systematic differences among facular regions. We find that the contrast of Caii K faculae is relatively insensitive to heliocentric angle, but is a strong function of facular size, in the sense that larger Caii K faculae are always brighter. The contrast of red faculae is a function of both heliocentric angle and size: the contrast functions show that larger regions contain larger flux tubes, contain deeper flux tubes, and have larger filling factors than small facular regions. Comparisons of cospatial pixels on red and Caii K images show a tight correlation between the average contrast of a region in the continuum and its size and heliocentric angle in the Caii K images. The average contrast of all facular regions is positive everywhere on the disk, even though the largest regions contain flux tubes which appear dark at disk center.     

Variation of Acoustic Power with Magnetic Field as Seen in Gong+ Data

The acoustic spectra in sunspots are known to be richer in higher frequency power. We have attempted a generalized study of the effect of magnetic fields on the shape of the acoustic spectrum using GONG+ bread-board data (spatial scale of 2 arc sec per pixel) of 11 May 2000 and 12 June 2000. The mean power spectra of the velocity oscillations were obtained by averaging over several spectra for different values of the magnetic field. With increasing magnetic field, the acoustic power increases at higher frequencies and decreases at lower frequencies with a transition at 5 mHz. This behavior is slightly different from earlier results obtained from SOHO/MDI data.     

Multiresolution wavelet analysis of SUMER/SOHO observations in a solar prominence

We have studied through a multiresolution wavelet analysis the oscillations in a limb prominence. Intensity fluctuations in time and height corresponding to different lines of Siiv and Oiv observed with SUMER on board SOHO have been analyzed in the wavelet bands of J₃= 1 min 36 s to 3 min 12 s and J₄=3 min 12 s to 6 min 24 s. For all species, oscillations in the J₄ band were dominant. We found relevant differences between the behavior of line D₁ (1393.76 Å) corresponding to Siiv and the set D₂ (1401.16 Å), D₃ (1404.81 Å), D₄ (1402.77 Å) corresponding to Oiv, Oiv and Siiv respectively. We also report the identification of a pulse in the intensity of the line D₁ that appears in the range of 15–20 min. This disturbance seems to travel with a speed of about 170 km s^–1.     

Observational Study of the Three-Dimensional Magnetic Field Structure and Mass Motion in Active Regions

Spectro-polarimetric observations of active regions were carried out in the spectral lines of Sii 10827.1 Å and Hei 10830 Å to study the three-dimensional magnetic field structure and associated plasma flow properties. Comparison of Sii and Hei magnetograms with the potential field model shows that a large fraction of the magnetic field is consistent with the potential field structure, by assuming that the height difference between the origin of the two lines is about 1200 km. The slope of the scatter plot between Sii and Hei magnetograms is 0.5, 0.76 in an emerging flux and a larger active region, respectively. These values are lower than the scatter plot slopes obtained from Kitt Peak photospheric and chromospheric magnetograms, in which case the corresponding values are 0.83 and 0.9, respectively. Considering the height difference between these two sets of chromospheric magnetograms, this implies that the magnetic field spreads out faster near the transition region heights. Dopplergrams obtained by determining the centroid of the asymmetric line profiles show that, in case of emerging flux region, the chromospheric upflow regions are located in the magnetic neutral line areas.     

A Study of the Rotation of the Solar Corona

Synoptic photoelectric observations of the coronal Fexiv and Fex emission lines at 530.3 nm and 637.4 nm, respectively, are analyzed to study the rotational behavior of the solar corona as a function of latitude, height, time and temperature between 1976 (1983 for Fex) and 2001. An earlier similar analysis of the Fexiv data at 1.15 R over only one 11-year solar activity cycle (Sime, Fisher, and Altrock, 1989, Astrophys. J. 336, 454) found suggestions of solar-cycle variations in the differential (latitude-dependent) rotation. These results are tested over the longer epoch now available. In addition, the new Fexiv 1.15 R results are compared with those at 1.25 R and with results from the Fex line. I find that for long-term averages, both ions show a weakly-differential rotation period that may peak near 80° latitude and then decrease to the poles. However, this high-latitude peak may be due to sensing low-latitude streamers at higher latitudes. There is an indication that the Fexiv rotation period may increase with height between 40° and 70° latitude. There is also some indication that Fex may be rotating slower than Fexiv in the mid-latitude range. This could indicate that structures with lower temperatures rotate at a slower rate. As found in the earlier study, there is very good evidence for solar-cycle-related variation in the rotation of Fexiv. At latitudes up to about 60°, the rotation varies from essentially rigid (latitude-independent) near solar minimum to differential in the rising phase of the cycle at both 1.15 R and 1.25 R . At latitudes above 60°, the rotation at 1.15 R appears to be nearly rigid in the rising phase and strongly differential near solar minimum, almost exactly out of phase with the low-latitude variation.     

Radial Flows in Supergranules

We determine the radial component of the supergranular flow velocity by examining the center-to-limb variation of the Doppler velocity signal. We acquire individual Doppler images obtained with the MDI instrument on the SOHO spacecraft and process them to remove the p-mode oscillation signal, the axisymmetric flows, the convective blueshift signal, and instrumental artifacts. The remaining Doppler signal contains only non-axisymmetric flow structures. The Doppler signal from the horizontal flows in these structures varies like sin, where is the heliocentric angle from disk center. The Doppler signal from radial flows varies like cos. We fit the center-to-limb variation of the mean squared velocity signal to a straight line in sin² over the central portion of the disk. The intercept of this line at disk center gives the amplitude of the radial component of the flow. The slope of the line gives the amplitude of the horizontal component. We find that the radial flows for typical supergranules have speeds about 10% that of their associated horizontal flows or about 30 m s^–1. The ratio of the radial to horizontal flow speed increases from 9% to about 18% as the size of the structure decreases from >60 Mm to 5 Mm. We use data simulations to check these results and find a ratio that increases from 5% to only about 12% over the same range of sizes. These smaller ratios are attributed to an underestimation of the horizontal flow speeds due to the fact that the transverse component of the horizontal flow is not detected by Doppler measurements.