Search for a command to run...
We present a comprehensive investigation of the Balmar-dominated northeast limb of the Cygnus Loop supernova remnant. Data presented include H alpha (O III), and X-ray images, UV and visible spectrophotometry, and high-resolution spectroscopy. The two relatively bright Balmer-dominated filaments visible on the POSS prints are seen to be part of a very smooth and regular complex of filaments. These filaments mark the current location of the blast wave and are seen to bound the sharply limb-brightened X-ray emission, including the previously reported X-ray, 'halo.' The (O III)/h beta ratio throughout the region is approximately 0.1, except for regions in which the shock is undergoing a transition from nonradiative to incomplete radiative to incomplete radiative conditions. At these locations (O III) emission from the cooling region is quite strong, while collisionally excited Balmer-line emission can be weak because of photoionization of the preshock medium by UV from the nascent cooling region. As a result (O III)/H beta is greater than 100 in some locations. The nonradiative/radiative transition is best studied along the length of the northwestern of the two brightest filaments, where the shock velocity and swept-up column go from approximately 180 km/s and 10<SUP>17</SUP>/sq cm at one end to approximately 140 km/s and 8 x 10<SUP>17</SUP>/cm at the other. There are also a number of locations of such incomplete radiative emission where the shock has recently encountered denser regions with characteristic sizes of approximately 10<SUP>18</SUP> cm. There is a considerable amount of evidence that the shock has decelerated from approximately 400 km/s to less than 200 km/s in the last 1000 yr. We interpret this as the result of the blast wave hitting the wall of a cavity which surround the supernova precursor and succeed in matching a wide range of data with a reflected shock model in which the density ofthe cavity wall is approximately 1.2/cu cm and the density in the interior of the cavity is about 0.09/cu cm. This analysis supports previous suggestions that the Cygnus Loop was a cavity explosion. We opredict that the X-ray shell will continue to brighten for the next 1000 years or so, during which time the shock front will become radiative throughout the northeast region. The similarity of the time scales for post-shock cooling, shock deceleration, and the 1ressure evolution of the region violates assumptions of steady flow, and greatly complicated interpretation of shock spectra in the region. This may be a common and important effect in many supernova remnants. High-resolution spectroscopy of the H alpha line profile shows that the broad component is surprisingly narrow (approximately 130 km/s). Consistency between this line width and the shock velocity inferred from spectrophotometry requires either rapid postshock equilibration of T<SUB>i</SUB> and T<SUB>e</SUB>, or that the shock front has decelerated from 180 km/s to 150 km/s in the last approximately 200 yr. While we cannot unambiguously distinquish between these models, the required deceleration is consistent with extrapolation of the shock decleration rate inferred over the last 1000 years. The width of the narrow component of the h alpha line profile is approximately 3 km/s, which indicates that the preshock medium has been heated to approximately 25,000 K by a shock precursor. <The most promising types of precursor are heating either by fast neutrals which overtake the shock or by cosmic rays. &In the latter case the observed properties of the precursor constrain the cosmic-ray diffusion coefficient to be between about 10<SUP>23</SUP> sq cm/s. The characteristics of the precursor seem to be about right to explain the presence of weak (S II) and (N II) emission previously reported by Fesen & Itoh (1985).