7 Discussion

7.1 Components of extended Galactic emission

Beuermann et al. (1985) have used the 408 MHz all-sky survey (Haslam et al. 1982) to produce a three-dimensional model of the Galactic radio emission using an unfolding procedure. In this model the Galaxy consists of a thick non-thermal radio disk in which a thin disk is embedded. The thick disk exhibits spiral structure, has an equivalent width of $\sim$3.6 kpc at the solar radius and accounts for $\sim$90% of the diffuse 408 MHz emission. Emission extends to at least 15 kpc from the Galactic centre, at which radius the thick disk has an equivalent width near 6 kpc. The thin disk, by comparison, appears in the model as a mixture of thermal and non-thermal emission also with spiral structure, but with an equivalent width of $\sim$370 pc, similar to that of the H I disk and of the distribution of H II regions in the inner Galaxy.

Our comparison of the 22 MHz and 408 MHz maps shows a remarkable constancy of spectral index in the extended emission corresponding to the thick disk component over our full range of longitudes from $\sim$0$^\circ$ to $\sim$240$^\circ$. The principal departures from this general tendency are (i) the slightly flatter spectral index in a broad area in the region of minimum Galactic emission at high latitudes toward the longitude of the Galactic anticentre and (ii) the somewhat steeper indices near Loop III and the outer rim of the NPS. In a similar comparison of the 408 MHz map with a map of 1420 MHz emission, Reich & Reich (1988) also note these general features. However, we see no indication in the lower frequency range for the steeper spectra seen by Reich & Reich in regions on the plane both near the Galactic centre and near longitude 130$^\circ$. This suggests that any steepening of the spectra in these regions must be a higher frequency phenomenon with spectral curvature above 408 MHz.

Details of the spectral index variations associated with the loops of emission also differ in the two frequency ranges. Figure 5 shows a slightly steeper index (by $\sim$0.03) for a substantial part of the arc forming the outer edge of the NPS. This contrasts with the 408-1420 MHz comparison (Reich & Reich 1988) which shows a steeper index in a relatively broad arc on the part of the NPS closest to the Galactic plane. Neither study indicates a difference between the spectral index of emission within the loop of the NPS and that outside the loop. The NPS has variously been considered as a nearby, very old supernova remnant (e.g. Salter 1983) and as a local magnetic "bubble'' (Heiles 1998).

7.2 Absorption at lower longitudes

We have noted that at longitudes less than $\sim$40$^\circ$  there exists a continuous trough of absorption along the Galactic plane. We illustrate this in Fig. 6 which shows a map of the "quasi optical depth'' at 22 MHz calculated from a comparison with the 408 MHz map on the assumption that the absorption is due entirely to cool ionized gas on the near side of the emission. (We define quasi optical depth, ${\tau}'$, by the relation ${\tau}' = \ln(T_{408} (408/22)^{{\beta}'}/{T_{22}})$, where ${\beta}'$ is the mean spectral index of the emission off the Galactic plane). This represents an underestimate of the true optical depth of the absorbing gas since (i) a proportion of the non-thermal emission will be on the near side of some absorption and, (ii) the kinetic temperature of the thermal gas will lessen the apparent depth of the absorption. A more accurate estimate of the true optical depth would require a modelling of the intermixed emission and absorption components which is beyond the scope of this paper. Nonetheless, it is obvious from Fig. 6 that the full angular width of the absorbing region is less than $3^\circ$ which, at an assumed mean distance of 4 kpc, corresponds to a thickness of less than 250 pc. Thus, it is apparent that the absorption corresponds to the "thin disk component'' of emission identified by Beuermann et al. (1985) as comprising the known disks of H II regions, diffuse thermal continuum emission, diffuse recombination line emission and the distribution of atomic hydrogen.

  

Figure 6: The "quasi optical depth'' at 22 MHz along the Galactic plane in the first quadrant, from a comparison of the 408 MHz and 22 MHz emissions, assuming all absorbing (thermal) gas is on the near side of the background synchrotron emission. Contours are at optical depths of 0.4, 0.8, 1.2, 1.6 and 2.0

The extended absorption in the plane in the region of Cygnus between longitudes 70$^\circ$ and 90$^\circ$ is also shown in Fig. 6. Note that the region appears at least twice as extensive in latitude as the continuous trough, probably because much of the absorbing gas is at distances of 1 kpc or less.

7.3 Non-thermal emissivities in the plane

Several of the discrete H II regions which appear in absorption at 22 MHz and which are listed in Table 2 can be used to estimate the emissivity of local synchrotron emission. We have calculated the emissivities for eight H II regions at well-determined distances, which are sufficiently extended compared to the observing beam to ensure that only thermal radiation from the ionized gas and foreground non-thermal radiation contribute to the measured emission. An assumed contribution from the opaque ionized gas of 6000 K was subtracted from the brightness temperature in the depression and the result divided by the distance to the H II region. The values of emissivity are presented in Table 3.

  

Table 3: Synchrotron emissivities in the directions of H II regions

• [$^{\mathrm{a}}$] Georgelin et al. (1973)
• [$^{\mathrm{b}}$] Georgelin (1975)
• [$^{\mathrm{c}}$] Humphreys (1978)
• [$^{\mathrm{d}}$] Turner (1976).

Note: The electron temperature in the H II regions is assumed to be 6000 K.

In the longitude range 85$^\circ$ to 205$^\circ$, six H II regions are at distances from 400-900 pc and the values of 22 MHz emissivity range from 21 Kpc^-1 to 60 Kpc^-1 with a mean of 40.1 Kpc^-1. Excluding two H II regions, Sh220 and Sh264, which are more than 10$^\circ$  off the plane, but including IC 1805, at a distance of 2.2 kpc, we find a mean emissivity of 30.2 Kpc ^-1 with an rms of 9.6  Kpc^-1. These emissivities are comparable with similarly derived emissivities tabulated (at 10 MHz) by Rockstroh & Webber (1978). In addition, our value in the direction of IC 1805, 20.9 Kpc^-1, is close to the value of 18 Kpc^-1 obtained by Roger (1969) using a detailed modelling of 22 and 38 MHz data for the IC 1805-IC 1848 complex.

However, there is a problem reconciling a mean value of local emissivity of 30 Kpc^-1 with the model of Galactic emission of Beuermann et al. (1985) which assumes a lesser value of 15 Kpc^-1 (11 Kkpc^-1 at 408 MHz) at the solar radius. If we take the value of the brightness temperature at the Galactic poles (27 kK), subtract an extragalactic component of 6 kK (Lawson et al. 1987) and divide by the model's half-equivalent-width of 1.8 kpc, we derive a mid-plane emissivity of only 11.7 Kpc^-1, almost a factor of 3 less than our measured mean value. To reconcile our measurement with the model, one or more of the following must apply: (i) our measured local mean emissivity is greater than the typical value at the solar radius; (ii) the equivalent width of the "thick-disk'' component is locally less than the model predicts; (iii) the extragalactic component of the polar emission is less than is estimated from extrapolations of extragalactic source counts at higher frequencies; and/or (iv) a zero-level correction should be added to the 22 MHz brightness temperatures. With regard to the extragalactic component of emission, we note that estimates are usually derived from source count ("log N - log S'') analyses at frequencies above 150 MHz (e.g. Lawson 1987), extrapolated with an assumed spectral index $\beta$ $\approx$ 2.75. Analyses of source counts at substantially lower frequencies are needed for accurate estimates of the extragalactic component. We noted in Sect. 6.1 the possibility of a zero-level correction as indicated by T-T plot comparisons with 408 MHz data. In this regard, it is interesting to note that very low resolution measurements with scaled antennas at several low frequencies (Bridle 1967) predicted a brightness temperature at 22 MHz in the area of the North Galactic Pole 4 kK higher than our value. This is of the same magnitude and sense as the offset suggested by the T-T plot analysis.

We note the unusually high emissivity derived for the direction toward $\zeta$-Oph (Sh27), a relatively nearby complex some 23$^\circ$  above the plane at the longitude of $\sim$6$^\circ$. Emission from this direction may include components from the North Polar Spur and from a minor spur that is most prominent near l=6$^\circ$, b=14$^\circ$, both of which may be foreground features. Also, it is possible that this somewhat diffuse region is not completely opaque at 22 MHz, in which case an unknown amount of background emission may contribute a spurious component to the emissivity.

Acknowledgements

We are indebted to several colleagues for their assistance in collecting and processing the observational data, and we particularly thank J.D. Lacey, J.H. Dawson and D.I. Stewart. We are also grateful to Dr. J.A. Galt for his encouragement at various stages of this project.