We have used the NRAO Very Long Baseline Array (VLBA) to image OH(1720 MHz) masers in the supernova remnants W28 and W44 at a resolution of 40 mas. We also used MERLIN to observe the same OH(1720 MHz) masers in W44 at a resolution of 290 × 165 mas. All the masers are resolved by these VLBA and MERLIN observations. The measured sizes range from 50 to 180 mas and yield brightness temperature estimates from 0.3-20×10 8 K.We investigate whether these measured angular sizes are intrinsic and hence originate as a result of the physical conditions in the supernova remnant shock, or whether they are scatter broadened sizes produced by the turbulent ionized gas along the line of sight. While the current data on the temporal and angular broadening of pulsars, masers and extragalactic sources toward W44 and W28 can be understood in terms of scattering, we cannot rule out that these large sizes are intrinsic. Recent theoretical modeling by Lockett et al. suggests that the physical parameters in the shocked region are indicative of densities and OH abundances which lead to estimates of sizes as large as what we measure.If the sizes and structure are intrinsic, then the OH(1720 MHz) masers may be more like the OH(1612 MHz) masers in circumstellar shells than OH masers associated with H II regions. At two locations in W28 we observe the classical S-shapes in the Stokes V profiles caused by Zeeman splitting and use it to infer magnetic fields of order 2 milliGauss.

We were unable to reliably determine scattering disk parameters for 2021]317 at frequencies from 8.5 GHz down to 1.67 GHz because of its complex intrinsic source structure. The scattering disks are elliptical with axial ratios of about 0.75 : 1 with little measurable variation between these sources. We interpret our measurements as due to the e †ects of anisotropic interstellar turbulence. The anisotropy parameters, axial ratio and position angle, for those sources for which we have data at multiple frequencies appear to be frequency-independent making refractive distortions of isotropic turbulence an unlikely explanation. Our estimates of b, the power-law index of the power spectrum of electron density Ñuctuations, based upon direct model Ðts to the data on a source-by-source and frequency-by-frequency basis, are consistent with b \ 4 for most of the data but are strongly a †ected by our estimate of the intrinsic structure of the sources. Two of these sources, 2005]403 and 2021]317, exhibit signiÐcant intrinsic source structure at frequencies as low as 1.67 GHz, while the other two, 2008]332 and 2048]312, lack discernible source structure at frequencies lower then 2.3 GHz. The frequency scalings of the scattered angular sizes are also consistent with b \ 4. We discuss the implications of our measurements for the inner and outer scales of the turbulence.

No abstract

No abstract