In this paper, we investigate the possibility of significant production of thermal bremsstrahlung radiation at radio continuum frequencies that could be linked to some Galactic supernova remnants (SNRs). The main targets for this investigation are SN
Rs expanding in high density environments. There are several indicators of radio thermal bremsstrahlung radiation from SNRs, such as a flattening at higher frequencies and thermal absorption at lower frequencies intrinsic to an SNR. In this work we discuss the radio continuum properties of 3 SNRs that are the best candidates for testing our hypothesis of significant thermal emission. In the case of SNRs IC443 and 3C391, thermal absorption has been previously detected. For IC443, the contribution of thermal emission at 1 GHz, from our model fit is 3-57%. It is similar to the estimate obtained from the thermal absorption properties (10-40% at 1 GHz). In the case of the 3C391 the conclusions are not so clear. The results from our model fit (thermal emission contribution of 10-25% at 1 GHz) and results obtained from the low frequency absorption (thermal contribution of 0.15-7% at 1 GHz) do not overlap. For the SNR 3C396 we suggest that if previously detected thermal absorption could be intrinsic to the SNR then the thermal emission (<47% at 1 GHz from our model fit) could be significant enough to shape the radio continuum spectrum at high frequencies. Polarization observations for these SNRs can constrain the strength of a thermal component. Reliable observations at low frequencies (<100 MHz) are needed as well as more data at high radio frequencies (>1 GHz), in order to make stronger conclusions about the existence of radio thermally active SNRs.
It has recently been reported that some of the flux density values for an evolved supernova remnant (SNR) HB 3 (G132.7 + 1.3) are not accurate. In this work we revised an analysis of the possible thermal emission at radio frequencies from the SNR HB
3 using the recently published, corrected, flux density values. A model including a sum of non-thermal (purely synchrotron) and thermal (bremsstrahlung) component is applied for fitting integrated radio spectrum of the SNR. The contribution of thermal component in total volume emissivity at $1 mathrm{GHz}$ was estimated to be $approx37 %$. The ambient density was also estimated to be $napprox 9 mathrm{cm}^{-3}$ for the $mathrm{T}=10^{4} mathrm{K}$. Again, we obtained the relatively significant presence of thermal emission at radio frequencies from the SNR so we could support interaction between SNR HB 3 and adjacent molecular cloud associated with the H {sc ii} region W3. Our model estimates for thermal component contribution to total volume emissivity at $1 mathrm{GHz}$ and ambient density are similar to those obtained earlier ($approx40 %$, $approx10 mathrm{cm^{-3}}$). It is clear that the corrected flux density values do not change the basic conclusions.
