5 Results

5.1 Map of the northern sky at 22 MHz

Figure 2 shows in equatorial coordinates a contoured gray-scale map of the 22 MHz emission from the sky between declinations -28$^\circ$ and 80$^\circ$ in five segments. Figure 3 depicts the data between Galactic latitudes -40$^\circ$ and +40$^\circ$ with the same contours and grayscale in Galactic coordinates, and with positions of extended Galactic sources indicated. Figure 4 is a grayscale representation of the full data set in Aitoff projection of Galactic coordinates.

  

Figure 2: A map of the emission at 22 MHz. Contours of brightness temperature are at the following levels in kilokelvins: 14, 19, 24, 29, 34, 41, 48, 55, 65, 75, 85, 100, 115, 130, 150, 170, 190, 220, 250. Levels in bold have thick contour lines and are labelled. Regions affected by sidelobes of four strong sources are blanked out. The superimposed grid shows Galactic co-ordinates in steps of 30$^\circ$ in l and b, labelled only where grid lines intersect the right-hand side and the top of the figure

 

Figure 2: continued

 

Figure 2: continued

  

Figure 3: The 22 MHz emission in Galactic coordinates in two segments with positions of prominent Galactic sources indicated. Contours, at the same levels shown in Fig. 2, are indicated with a bar scale. The superimposed grid shows equatorial co-ordinates (J2000) in steps of 2$^{\rm h}$ in right ascension and 30$^\circ$ in declination

  

Figure 4: An Aitoff projection of the 22 MHz emission in Galactic coordinates. The Galactic centre is at the map centre and grids are at 30$^\circ$ intervals in longitude and latitude, positive to the left and upwards respectively. Note the arc of the North Polar Spur rising from the plane near longitude +30$^\circ$

The brightness temperature of the 22 MHz emission varies from $\sim$17 kK towards a broad minimum about 50$^\circ$ off the Galactic plane at the longitude of the anticentre to over 250 kK on the plane near the Galactic centre. The brightness temperature near both north and south Galactic poles is approximately 27 kK. The Galactic plane itself is apparent over the full range of longitude from +1$^\circ$ to +244$^\circ$.At various points along the plane, particularly at lesser longitudes, depressions are apparent in the emission. These represent thermal free-free absorption of bright synchrotron background emission by relatively nearby, opaque regions of dense ionized gas.

Two other large-scale features apparent in the maps are Loop I, the North Polar Spur (NPS), rising from the plane near longitude 30$^\circ$,and Loop III, centred near longitude 87$^\circ$.

We emphasize that the main value of the data lies in the representation of structure larger than the beam. The strongest point sources (Cas A, Cyg A, Tau A and Vir A) have been removed from the map. While other point sources remain in the maps, these data cannot be used to determine their flux densities. First, the ionospheric effects mentioned above cause the point sources to be very poorly represented in these maps. Second, the scaling applied after comparison with the 408 MHz data will have further affected the flux densities of point sources at declinations away from the zenith. Reliable point source flux densities are already available in the published lists referred to in Sect. 1.

5.2 Extended Galactic sources - supernova remnants

A number of extended supernova remnants are apparent in the data and the positions of these are indicated with labels in Fig. 3. The flux densities of most of the SNRs have been previously measured from the original observations and published in various papers. We have collected these and listed them in Table 1 together with new flux densities for two additional remnants not previously reported. One other SNR, HB21, is indicated in Fig. 3 but not listed in Table 1 because of difficulties in separating its emission from that of nearby confusing sources.

  

Table 1: Flux densities of supernova remnants at 22 MHz

5.3 Absorption of the background emission - H II regions

Depressions in the background emission near the Galactic plane are identified with a number of extended H II regions, which at frequencies near 20 MHz will largely obscure background emission. We list the properties of 21 of these discrete absorption regions in Table 2, with positions plotted in Fig. 3.

  

Table 2: H II regions in absorption

Figures 2, 3 and particularly 4 show the extended trough of absorption between l=10$^\circ$  and l=40$^\circ$.This trough undoubtedly extends to and past the Galactic centre but the increasingly extended N-S width of the telescope beam at large zenith angles was unable to fully resolve the feature below l=10$^\circ$.

5.4 The spectral index of emission, 22 to 408 MHz

Figure 5 shows a map of spectral index calculated from the final 22 MHz map and the 408 MHz map (Haslam et al. 1982), the latter convolved to the declination-dependent beamwidth of the 22 MHz telescope. The spectral index, $\beta$, as displayed, is related to the brightness temperatures at each frequency, T₂₂ and T₄₀₈, by the expression

$$\beta = \log (T_{22} / T_{408}) / \log (408/22).$$

Because the 408 MHz map has been used to establish the variation of the 22 MHz temperature scale with declination, great care is needed in interpreting this map. The process of revising the 22 MHz scale could eliminate or reduce spectral index features between 8 and 16 hours right ascension with structure in the declination direction if they extend over a large range in this dimension. On the other hand, features in the spectral index map which have structure in the right ascension dimension are likely to be largely unaffected by the correction process. Similarly, spectral index features with structure in various directions, including most features which have counterparts in the individual maps, would be suspect only if distortions appeared in the declination dimension. No such artefacts are apparent. However, some "banding'' in declination, particularly below -3$^\circ$, can be seen in specific right ascension ranges. This effect is easily recognized as spurious and is probably related to zero level errors in the 22 MHz or, possibly, in the 408 MHz map. Such effects may be aggravated by the large zenith angles at which these low-declination regions were observed at 22 MHz.

  

Figure 5: A map of the spectral index from 22 MHz to 408 MHz in shaded grey levels as indicated by the bar scale. Regions near four strong sources are blanked out. The superimposed grid shows Galactic co-ordinates in steps of 30$^\circ$ in l and b

Errors in the spectral index map can arise from zero-level or temperature-scale errors at either frequency. We deal with zero-level errors first. Taking 5 kK as a possible error in the 22 MHz data (see the detailed discussion in Sect. 6.1) we estimate the effects on the spectral index map. The large frequency separation between 22.25 and 408 MHz means that zero-level errors have relatively little effect: the $\pm$ 5 kK error will change spectral index by $\pm$ 0.1 at the sky minimum and by $\pm$ 0.01 on the brightest part of the Galactic plane. We tested the effect of an error of $\pm$ 5 kK on the map of Fig. 5 by computing maps with this error applied to the 22 MHz data in both senses. All the main features visible in Fig. 5 remain in both new maps. When we discuss the spectral features which we see in Fig. 5, we discuss only those which survive this test.

The effects of temperature scale errors are more difficult to assess. Once again, the large frequency separation is an asset. At the zenith (declination 48.8$^\circ$) we have an independent determination of spectral index since we have not changed our data in any way at that declination. We estimate that the probable error there is 0.05, taking into account both systematic and random errors (amounting to a 16% difference in the temperature ratio T₂₂/T₄₀₈), and we assign this error to the whole map. Further errors at other declinations depend on the validity of the assumption on which our calibration of the 22 MHz temperature scale is based, the constancy of the spectral index of the Galactic emission over the region 8$^{\rm h}$ to 16$^{\rm h}$, -28$^\circ$ to 80$^\circ$. This can only be tested with an extensive study of spectral index using data at a number of frequencies, a study beyond the scope of this paper. Note that we have assumed the constancy of the differential spectrum: we are assuming that the Galactic component of the emission has constant spectral index over this region, which includes the Galactic pole and lies mostly at latitudes higher than $\sim$20$^\circ$. The total spectrum includes an extragalactic (presumably isotropic) component of emission, and the total spectrum may still vary across this region, as it appears from Fig. 5 to do.

In the final analysis, the main value of our spectral index work is in the assessment of differences in spectral index between regions rather than in a precise determination of the spectral index of a given region. In this spirit, we make the following observations.

• [1.] In general, the map in Fig. 5 shows few large scale variations. The mean index away from the Galactic plane is 2.47. Within this area (latitude greater than $\sim$10$^\circ$) the mean index of extended regions varies from values of 2.41 in the area of synchrotron minimum (10$^{\rm h}$, 40$^\circ$) to 2.54, near the position of Loop III, both with an rms variation of $\sim$0.03. On the Galactic plane, the flatter spectrum region coincident with the absorption trough at low longitudes is narrow in the latitude dimension. At all points on the Galactic plane it is notable that the spectral index is relatively constant over most of the broader synchrotron emission along the plane with no gradual decrease in spectral index as latitude decreases.
• [2.] The outline of the North Polar Spur at higher latitudes can be seen in the spectral index map. An integration over the area of this arc shows its emission to have an index of 2.51, slightly steeper than its surroundings by $\sim$0.03.