Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 321, 229-235 (1997)

Previous Section Next Section Title Page Table of Contents

3. Observations and data reduction

Observations with the WSRT of the combined COMPTEL and IPN error region (from now on referred to as COMPTEL+IPN) of GRB 940301 began, three days after the first of March event, at 1400 MHz (21 cm). Data were recorded in 8 spectral bands with a total bandwidth of 65 MHz (5 bands with a width of 10 MHz at 1375, 1385, 1395, 1405 and 1415 MHz, and 3 bands with a width of 5 MHz at 1367.5, 1397.5 and 1423.5 MHz). Full polarisation information was recorded. Observations were obtained at several pointings (fields 1 to 5, A and B in Fig. 1). The COMPTEL+IPN error region is not fully covered at 21 cm (see Fig. 1). Coverage of the whole COMPTEL+IPN error region was not obtained until 32 days after the occurrence of the event, by which time the operating wavelength had changed to 92 cm.

The 92 cm broad-band data cover a total continuum bandwidth of 40 MHz, recorded in 8 bands each 5 MHz wide, centered at 319.3, 325, 333, 341, 355, 360, 375 and 380 MHz (these bands correspond to wavelengths 94 [FORMULA] 79 cm). Henceforth we will refer to these data as the 92 cm data. A total of 48.8 hours of data were taken in 9 observations, covering the period from April 2 until June 25, 1994; additional observations were obtained January 15 and 16, 1996 at slightly different band frequencies. Table 1 provides a log of the data. Full polarisation information was measured, but no polarisation was detected from any of the interesting sources. Due to interference about 20 % of the data were unusable. This interference was usually concentrated in a few bands. The data were reduced using the Netherlands East-West Synthesis Telescope Array Reduction package (NEWSTAR) 1.


Table 1. Broad-band 92 cm WSRT observations.

Because of incompleteness of the [FORMULA] coverage of the 92 cm observations that last less then 12 hours, we combined data from April 2 1994 with data from June 25 1994, which, together, yielded almost a complete [FORMULA] synthesis. From these data we constructed an accurate model of the field at 92 cm in a number of steps. We began with calibrating the complex gains for each observation and each band, using the calibrator sources 3C48, 3C147 and 3C286. For these primary WSRT flux calibrators we adopt flux densities of 46.1, 56.7 and 26.9 Jy at 325 MHz and spectral indices [FORMULA] = -0.65, -0.62 and -0.35 in the range 300-400 MHz (where spectral index [FORMULA] is defined by [FORMULA] = d log S /d log [FORMULA]). The data were then self-calibrated in phase using a model of the field obtained from the initial data. This process was iterated a few times and thereby the model refined until nearly 400 discrete background sources above a flux density of about 6 mJy were included in the model. The excellent [FORMULA] coverage resulting from the wide range in frequencies resulted in a very low sidelobe level in the synthesized beam, a property essential to imaging of a wide field with many hundreds of sources. The total flux density in the model is 19 Jy with 5 sources in the range from 0.5 to 1 Jy. The relatively large spectral baseline in the 92 cm data allowed us to also solve for the spectral index of each source in the field.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: June 30, 1998