Coronal loops: Isothermal OR multithermal?

Are coronal loops isothermal? A controversy over this question has arisen recently because different investigators using different techniques have obtained very different answers. Analysis of data using narrowband filter ratios to obtain temperature maps has produced several key publications that suggest that coronal loops may be isothermal. We have constructed a multi-thermal distribution for several pixels along a relatively isolated coronal loop on the southwest limb of the solar disk using spectral line data from the Coronal Diagnostics Spectrometer (CDS), on SOHO taken on 1998 April 20. These distributions are clearly inconsistent with isothermal plasma along either the line of sight or the length of the loop, and suggested rather that the temperature increases from the footpoints to the loop top. We convolved these Differential Emission Measure curves with two of the Soft X-ray Telescope (SXT) response functions. This gives us the intensity (in Data Numbers/sec) of what the instrument would “see” in these filters if it were observing the same loop. We then took a ratio of these values, and used the regular Yohkoh software to calculate a temperature at each pixel. The instrument “sees” a loop that is marginally consistent with an almost uniform temperature, but a linear or quadratic model is a much better fit to the data. These results are different from those of a similar analysis with narrow-passband instruments that produced data consistent with a uniform temperature loop, even though the actual temperature input was multi-thermal both along the line of sight and along the length of the loop. We suspect that these apparent uniform-temperature loops may be an unfortunate byproduct of the simplistic filter-ratio method that is used for temperature analysis. Our results are consistent with earlier analysis of Yohkoh data, where there were strong indications that the SXT temperatures were a kind of Differential Emission Measure-weighted temperature. There is a problem, however, when we compare these temperatures with those calculated with the actual SXT data, which are much higher than even the hottest plasma observed by CDS.