The Metsähovi radio telescope was upgraded in 1992-1994, and at present the surface accuracy is 0.1 mm (rms). The 14-meter antenna gives 1.0 arcmin spatial resolution (i.e. HPBW beam size) at 87 GHz (3.5 mm). The maps are made by scanning the Sun in right ascension, by changing declination between the scans. The full solar disk is scanned in 9 minutes in the "fast map'' mode, which makes 29 scans, 91 samples each. The flux resolution for this study was estimated from nearby active region track files. Track files are made by pointing the beam to a selected region on the solar disk and by measuring the flux density (sampling rate 20 samples/s). By using active region track files we get a maximum amount of noise fluctuations, as fluctuations from the active region itself are included. Large noise levels can therefore be the result of very active regions on the disk. A large noise level can affect the quality of the map (i.e., the dynamic range). An example of a low-quality radio map is in Fig. 8, from May 20, 1997. The variation in this case is most probably instrumental, and depends on the set-up. Set-ups may be changed between the days or between the observing runs. In the analysed radio maps the flux resolution varied from 0.3% to 0.5% of the quiet Sun flux, depending on the observing run in question.
No absolute flux calibration is done during the Metsähovi solar observations, but the quiet Sun brightness temperature has previously been estimated to be around 7200 K at 87 GHz (see Pohjolainen & Urpo, 1997), giving a brightness temperature resolution of 22-36 K for this study. The observed flux density, which is directly related to brightness temperature, is expressed in the tables as a percentage of the quiet Sun level, which is determined from the A/D converter counts. For example, on April 12, 1996 the quiet Sun level was estimated to be at 7240.0 A/D count units, which is equivalent to the estimated 7200 K brightness temperature. A bright area showing a 1.009 relative intensity is therefore 0.9% above the quiet Sun level, i.e., is showing a brightness enhancement of 64.8 K.
The solar maps that are presented here are the "raw'' maps, without any deconvolution methods applied. Because the beam is not sharp but more like a broad Gaussian with wings (sidelobes), an artificial limb darkening and an artificial sky brightening are observed near the true edge of the solar disk (see Lindsey & Roellig [1991]). The effect can be seen as the dark ring in the 87 GHz radio maps inside the circle which marks the optical solar limb, and as additional emission outside the solar limb. We analyse here only those brightenings and depressions that were well outside this effect (i.e. well inside the solar disk). The artificial limb darkening makes it impossible to see the features as high as the polar caps, but because of the changing B0 angle, features up to 70 deg in latitude (at 0 longitude) can be discerned.
The Extreme Ultraviolet Imaging Telescope (EIT) onboard the SOHO spacecraft, images the corona in four EUV wavelengths (Delaboudiniere et al. [1996]). From EUV lines - which represent different ion transitions - the chromosphere (He II), the transition region (Fe IX/X) and the corona (Fe XII, Fe XV) are imaged, see Table 1. Temperature diagnostics are made by wavelength ratios (Neumark et al. [1997]), and 3-D coronal structure evolution is studied by comparing filter images (Portier-Fozzani et al. [1996]).
The selected four EIT images per day - observed as close as possible
to the radio observations - were taken with filter Clear or Al+1.
The EIT pixel resolution is 2.6 arcsec. The full Sun images were
either 1024
1024 (full resolution, denoted "F'' in Table 2)
or 512512 (half resolution, denoted "H'' in Table 2).
The EIT images were calibrated using the usual software methods described in Moses et al. ([1997]). To remove the shadow of the tiny grid that supports an aluminium filter standing in front of the CCD camera, included in all EIT images, smooth filtering using a local median filter was preferred (Portier-Fozzani et al. [1996]). More details concerning the EIT calibration and in flight use can be found in Moses et al. ([1997]), Newmark et al. ([1997]), Defise ([1999]), and Portier-Fozzani ([1999]).
| Wavelength | Ion | Temperature |
| 304 Å | HeII | 8.0 104 K |
| 171 Å | FeIX, X | 1.3 106 K |
| 195 Å | FeXII | 1.6 106 K |
| 284 Å | FeXV | 2.0 106 K |
The Soft X-ray Telescope (SXT) onboard the Yohkoh satellite observes the solar corona through several different filters (Tsuneta et al. [1991]). The broad-band instrument is sensitive over a range of energies between 0.25 and 4 keV. The corresponding coronal temperatures are in the range of 2-20 MK. In the case of the quiet solar corona, the temperatures are at the lower end of this scale.
We used SXT full frame images, taken with the AlMg and Al.1 filters near the times of the radio observations, with the longest available exposure times, in order to get the faint features visible. The full frame SXT images were observed either in Half resolution (denoted "H'' in Table 2) with 4.9 arcsec pixel size, or in Quarter resolution (denoted "Q'' in Table 2) with 9.8 arcsec pixel size. The exposure times varied between 0.6 s (August 13, 1996) and 30.2 s (April 12, 1996). The SXT images were corrected for dark noise and particle hits.
As it is well known than H
dark filaments can cause radio
depressions (Vrsnak et al. [1992]),
we checked from the H
images in the Web
(http://sohodb.nascom.nasa.gov/cgi-bin/synop_query_form)
that no filaments were present where we had radio depressions.
Therefore the radio depressions analysed in this study must be caused
by some other mechanism than an overlaying absorbing dark filament.
Coronal holes and their borders are sometimes difficult to define. We used also other wavelength data from the Web, when available, to check the locations under study. Especially He I (10830 Å) images were useful for this purpose.
| Date | 87 GHz | EIT map | SXT map | ||||
| radio map | UT | UT | |||||
| UT | He II | Fe IX/X | Fe XII | Fe XV | AlMg | Al.1 | |
| Apr. 12, 1996 | 13:57-14:05 | 14:50(F) | 14:28(F) | 15:12(F) | 14:06(F) | 13:52(H), 14:23(Q) | 13:55(H) |
| Apr. 15, 1996 | 12:05-12:13 | 13:12(F) | 12:14(F) | 12:20(F) | 13:01(F) | 11:18(H), 13:13(H) | |
| Aug. 09, 1996 | 07:23-07:32 | 07:40(H) | 07:24(H) | 07:34(H) | 07:30(H) | 07:00(H) | 06:48(H) |
| Aug. 13, 1996 | 12:31-12:39 | 12:35(H) | 12:30(H) | 12:34(H) | 12:32(H) | 12:39(Q), 12:48(Q) | |
| Aug. 14, 1996 | 12:55-13:03 | 12:35(H) | 12:30(H) | 12:34(H) | 12:32(H) | 12:53(H) | 12:39(H) |
| May 20, 1997 | 13:03-13:12 | 13:18(F) | 13:00(F) | 13:12(F) | 13:06(F) | 12:53(Q), 13:15(Q) | |
| Aug. 08, 1997 | 11:30-11:39 | 13:28(F) | 13:10(F) | 11:47(F) | 13:16(F) | 11:04(Q), 12:17(Q) | |
| Aug. 27, 1997 | 12:13-12:22 | 13:19(F) | 13:01(F) | 12:15(H) | 13:07(F) | 12:10(H), 12:55(H) | 12:53(H), 13:08(H) |
| Aug. 28, 1997 | 07:16-07:24 | 07:18(F) | 07:00(F) | 07:12(F) | 07:06(F) | 07:06(H), 07:23(H) | 07:04(H), 07:21(H) |
We went through the radio maps measured at Metsähovi at 87 GHz (3.5 mm) in 1996 and 1997, and selected the ones that had close-in-time observations at least with one of the EUV wavelengths, observed by the EIT. In this way 9 radio maps from 9 days were selected, see Table 2. A lot more radio maps are available if nearness in time is not a requirement.
All four EIT wavelengths were then analysed, to confirm the structural classification in EUV. In the analysis the maximum time difference between a radio and an EUV map was 1.5 hours, but in that case more close-in-time observations were available at the other three EUV wavelengths, and we checked that the observed features were consistent. SXT maps were selected on the basis of being taken as near as possible to the selected radio and EUV maps, and with the longest available exposure times.
The tops and the bottoms of the selected radio brightenings and depressions, respectively, were given heliographic coordinates. The positions (latitude, longitude) and intensities (relative to the quiet Sun level) for altogether 104 radio sources are given in Tables 3-11, for each of the days separately. In the tables, EIT and SXT structures are defined for the radio locations and their surrounding area.
In comparing the features in radio and in EUV one should be aware of the fact that the spatial resolution at 87 GHz is rather poor, about 1 arcmin (HPBW). Radio emission sources that are located within the 1 arcmin beam area will be convolved, and this means that the true peak brightness can be "off'' the given coordinates. The scanning technique poses problems as well: pointing errors are common when using heavy antennas at high speed. To overcome the "off'' pointing, all features within the 1 arcmin beam were classified and listed.
The Metsähovi scan map method was originally developed for 37 GHz, with a much larger beam size. At 87 GHz the beam size is smaller than the separation between scans, 1.2 arcmin. This can mean missing some flux, if a very small point source is located just in between the scan paths. However, the non-symmetrical main beam and strong sidelobes usually reduce this problem. Atmospheric changes between separate scans can also cause defects, which are usually seen as the sawed edge of the solar limb.
Another, a much larger error, comes from the determination of the center of the radio Sun. Due to the non-symmetrical beam and errors in scanning, the radio Sun is not a perfect circle, and the person making the data reduction is very much responsible for putting the center point to the "right'' position. Our data set was analysed separately by two different people, and the maximum difference in coordinate determination was 10 degrees in heliographic longitude, which was caused by moving the center point in an East-West direction. However, this kind of difference in coordinates was extremely rare.
Also, a small bright point in a cool dense region will be diluted in a large radio beam, and will not necessarily show up. If sources looking similar in EUV are not both detected in radio, one must look for differences in temperature and/or density. The basics of radio emission and antenna beams can be found in, e.g., Dulk ([1985]) and Pohjolainen ([1996]). To explore the effects of the different resolutions of the two instruments, a test was made by plotting EIT images degraded into radio resolution. It was found that structural mixing was present in some cases, but also good fit to the radio images was found (Pohjolainen et al. [1999a]).
Copyright The European Southern Observatory (ESO)