The observations were carried out jointly at radio wavelengths (Spectrograph of the Tremsdorf Solar Radio Observatory - OSRA, Nançay Radioheliograph - NRH). The OSRA instrument consists of swept-frequency spectrographs in the ranges 40-90, 100-170, 200-400 and 400-800 MHz, with a sweep rate of 10 (Mann et al. 1992). The dynamic spectrograms were recorded on film; since September 1993 digitally recorded spectra have been available. The NRH provides two one-dimensional scans of the corona per 0.25 s (1990), 1 s (1991-Jan. 1993) and 0.1 s (thereafter), respectively. In 1990 its east-west branch was operated at 164 MHz, the north-south branch at five frequencies (most often 164, 236.6, 327, 408 and 435 MHz). Since summer 1991 both branches observed at five frequencies (The Radioheliograph Group 1993).
The Meudon spectroheliograph takes one image of the Sun per day. Images in the centre of the H line are used in this paper. Yohkoh soft X-ray images and the necessary software were kindly provided by the Yohkoh community. Details are described e.g. in The Yohkoh Analysis Guide (1994).
By inspection of the OSRA dynamic spectra we selected a sample of type U bursts with branches extending to at least one of the NRH observing frequencies. Those bursts for which NRH data were available were retained for further study. Tables 1 (click here) to 3 (click here) summarize the observed properties of these bursts. In Figs. 4 (click here) to 27 (click here) the events are presented using a unified format as follows (from top to bottom):
In order to give the full spectral and spatial information, we usually had to choose different time scales for the plots of the spectrum and the brightness distribution. The NRH positions and, whenever possible, the flux densities of the different radio sources given in Tables 2 (click here) and 3 (click here) have been evaluated by a Gaussian (or Gaussian + straight line) fit to the NRH scans. Whenever a significant polarization signature is observed, it is described in the text. The radio positions are plotted as crosses whose extent gives the half-power diameter of the Gaussian sources measured with the two antenna arrays; the source of the ascending branch is plotted by a solid cross, that of the descending branch as a dashed cross (same line style as the corresponding brightness curves). Corrections for ionospheric refraction were applied using a static standard model or measurements of the ionospheric electron density (correction routines courtesy C. Mercier). The influence of time-dependent density fluctuations was corrected by using simultaneously observed noise storm source locations for reference. Clearly the uncertainty of radio heliographic positions increases with decreasing elevation of the Sun (i.e. in the winter months and far from local noon which is about 11:50 UT at Nançay). Even under worst conditions this uncertainty was smaller than the indicated source size.