5 Observational results

We review here several published and new observations, that are related to the problem of solar magnetic field measurements according the observations of microwave emission with high angular resolution.

5.1 RATAN and Westerbork observations between 1.5 and 11 GHz

(a) From RATAN observations of a bipolar spotless plage area (Bogod & Gelfreikh 1980) at $\lambda$ = 2.7 and 3.2 cm (the most accurate measurements with RATAN on August 1, 1978) we obtained $\delta I$ = 5400 and 7000 K, V = 340 and 380 K respectively. These values lead to estimates of B = 57 and 42 G through Eq. (27) and parameter check $<tT_{\rm e}>_{\rm PL}$ = 850 and 790.

(b) From Westerbork observations (May 9-10, 1974) of isolated plages at $% \lambda$ = 6 cm, (see Fig. 10 of Kundu et al. [1977], the 3 plages labelled a, b and c), we have $\delta I$ = 30, 40 and 30 (10³ K) with V = 8, 12 and 12 (10³ K). These values lead to estimates of B = 60, 100 and 80 G, and $<tT_{\rm e}>_{\rm PL}$ = 800, 900 and 800.

(c) RATAN observations of the great prominence-coronal arcade structure above the W-limb on October 4, 1996 reveal the presence of a circularly polarized signal from the prominence location with V/I=0.03 at $\lambda$= 1.83 cm (Bogod et al. [1998]) which leads to a magnetic field strength ofabout B = 50 G.

(d) With RATAN multi-wavelength observations of a coronal hole area near the disc center (on October 12 and November 9, 1996) a weak polarization signal was registered with $V/\delta I=0.015$ at $\lambda$ = 18.3 cm (Borovik et al. [1999]). At this wavelength the coronal hole was observed as a 30% brightness depression. This leads to a coronal parameter $<tT_{\rm e}>_{\rm CH}~\approx 0.7<tT_{\rm e}>_{\rm QS}$ and gives a magnetic field strength of about B = 2.5 G.

5.2 Nobeyama polarization observations at 17 GHz

Routine Nobeyama imaging techniques give polarization levels near the noise threshold for an active region with weak magnetic fields (see Shibasaki et al. [1991]).

By averaging hundreds of separate images (Gelfreikh & Shibasaki [1998]) a statistical polarization noise $V_{\rm rms}$ may be reduced to about 50 K with intensity signal up to 400 K for plage area(s) (at full brightness level 12000-14000 K). Here we present results of such observations during June 1995 for the active region AR 7877 in intensity and polarization at 17 GHz (see Fig.  2).

Figure 2: Summary of the observations of AR 7877 on June 9, 1995. a) Yohkoh SXT image (600 $\times$ 600 arcsec), b) Kitt Peak magnetogram (fragment of 600 $\times$ 600 arcsec), c) microwave scan at $\lambda = 4.4$ cm with Great Pulkovo Radiotelescope of intensity (solid line) and circular polarization (dotted line) with HPBW = 60 arcsec, d), e) Nobeyama Radioheliograph images in intensity and circular polarization (50 images have been averaged), and f) its superposition. HPBW is about 15 arcsec, spatial scale unit for both axes is 4.9 arcsec per tickwork. For all images, solar north is on top and west is right. Nobeyama and Kitt Peak images are aligned to an accuracy about 10''. The microwave peak brightness position coincides with the X-ray brightest loop top and is located on the neutral line both on the magnetogram and on microwave polarization

AR 7877 (N09 E49) arose near the East limb on 5 June and it is developed to a bipolar sunspot group near central meridian time. From Kitt Peak photospheric magnetograms we saw a large scale background activity complex, with two adjacent bipolar regions, which we label as a, b, c, d from leading to trailing plage areas (see Fig. 2b). At the coronal heights, this complex of activity has an unusual magnetic structure, as delineated at soft X-rays with Yohkoh SXT observations (see Fig. 2a). From coaligning X-ray and magnetogram images we conclude, that X-rays "butterfly'' originates from the region b, with two arcades (A, B) of magnetic flux lines, which emanate westword (labelled A) to region a and eastword (labelled B) to the region c. Nobeyama observations at polarization display four polarized regions labelled a-d (see Fig. 2e) which closely match photospheric magnetogram (see Fig. 2b), by position and signs (extraordinary emission mode) of polarization peaks. Peaks of polarized emission at locations a, b coincide with position of spots in the bipolar AR 7877, with strong magnetic fields up to 2500 G at photospheric level.

Nobeyma observations at intensity (see Fig. 2d) display, that the main brightness peaks do not overlay sunspot positions, but overlay the neutral line positions (see Fig. 2f for overlay of the I and V images). The same pattern is observed with Pulkovo 1D scan at longer wavelength $\lambda$ = 4.4 cm (see Fig. 2c). We do not discuss here the nature of these weakly polarized emission peaks, which are possibly originate at the apex of X-ray arcades.

In context of microwave estimates of magnetic fields we conclude, taking into account the low brightness of plage areas observed at microwaves, that chromospheric magnetic fields at location of polarization peaks are below of cyclotron emission estimate for a third (s = 3) harmonic at $\lambda$ = 1.76 cm (threshold of $B=10700/s\lambda =2000$ G).

Thus, we accept a thermal bremsstrahlung as emission mechanism at 17 GHz at the locations of all polarization peaks a-d, including sunspot plage areas. For two plage areas a-b the measured values of intensity are I = 16500 K and 22000 K, with $I_{\rm QS}=10^{4}$ K, and the polarization signals are V = +430 K and -400 K for, so that $\triangle I\$ = 6500 and 12000 K for areas a and b accordingly. For these plage areas we estimate a coronal parameter values $<tT_{\rm e}>_{\rm PL}$ = 1200 and 1800. These values exceed the expected coronal values about $<tT_{\rm e}>_{\rm PL}$ = 800, which suggest the hot X-ray loop contribution. If we adopt these values as coronal ones, then $% <tT_{\rm e}>_{\rm PL}\lambda ^{2}$ = 3800 K and 5600 K and the coronal contribution dominates the polarization. Thus, for a coronal magnetic field above these plage area(s) we would have B = 93 G and B = 47 G (Eq. 27). For plage c we obtain $\delta I$ = 4000 K, V = -145 and an estimated field value B= 51 G.

For the same data, but by using the more strict estimate by Eq. (25), we would have a twofold increase of the magnetic fields at a-c namely B = 150, 110 and 60 G respectively. These values are in good agreement with the photospheric values obtained from magnetograms made at Huiarou Station of BAO (courtesy of Prof. Hongqi Zhang), with peak values of approximately B= 320, 320 and 160 G for the same plage area(s).