The two variable sources (B0637+6313 and B0635+6318, hereafter referred to as the East and West source), are located at the edge of the synthesized field at 92 cm (which measures about at 325 MHz at half-power). They are only apart. Both sources are unresolved in the 92 cm ( beam) and 21 cm ( beam) observations. From a full 12 hour run at 21 cm we obtained accurate positions of both sources at , (East source) and , (West source; B1950). They are located at approximately the 2.3 and 2.6 confidence contour of the COMPTEL position of GRB 940301, respectively. They fall outside the IPN triangulation annulus for this GRB (the East source by about 9 and the West source by about 18 , taking half the width of the IPN annulus, , as 3 ; see Fig. 1).
Two contour plots of the area containing the two sources are shown in Fig. 2. The plot on the left-hand side is obtained from the WENSS data, the right-hand plot from our 325 MHz observations, made from data taken on April 2, 11, and June 5, 11, 20 (1994).
The East source was barely detected in the WENSS survey (see Fig. 2). The source continued to increase at 325 MHz until January 15/16, 1996 (see Fig. 3). The total increase between the beginning of 1992 and January 1996 is a factor of (1 ; henceforth all quoted errors are 1 ). A summary of all flux densities is given in Table 2.
Table 2. Flux densities for the East source (B0637+6313).
The source was detected as a radio source well before the time of the GRB. It appears in the GB6 6cm (4850 MHz) survey with the Green Bank telescope (Gregory et al. 1996) and in the NRAO NVSS survey at 21 cm (Condon et al. 1996; in progress). There is no source visible in the images from the Green Bank 21 cm survey (Condon and Broderick 1986) down to about 50 mJy. The Green Bank 21 cm survey is, however, confusion limited (1 25 mJy).
The broad band spectrum of the East source, from 0.3 to 5 GHz, is shown in Fig. 4. Although the data are obtained at different dates it is clear that the overall spectrum is fairly flat as is typical for compact extragalactic radio sources. The flux density in the July 7th, 1995, 1400 MHz observation was of sufficient strength to allow us to derive a spectral index within the 65 MHz wide band: = +0.25 0.16, consistent with the overal spectrum. In the 1994 92 cm data the spectral index of the East source was highly inverted, . In January 1996, when the source had brightened, the spectrum had flattened.
The West-variable source first increased in flux at 325 MHz, compared to the WENSS survey, by about a factor 2.1 0.3 and then declined again by almost the same factor (see Fig. 5 and Table 3). The July 7, 1995 data yielded a flux density at 1400 MHz, nearly identical to the value obtained from the NRAO NVSS survey. From the 21 cm survey with the Green Bank telescope an upper limit of 50 mJy, October 1983, was obtained. Also, this source was detected well before the time of the GRB in the 6 cm Green Bank survey (GB6).
Table 3. Flux densities for the West source B0635+6318.
The spectrum of the West source is shown in Fig. 6. The plot reveals that its spectrum is fairly flat (), though again we should note that the data are obtained at different dates. The July 7th, 1995, 1400 MHz observation yielded . In the 1994 92 cm data the spectral index of the West source was fairly flat . Also the spectrum obtained from the 92 cm January 15/16 1996 observation is flat to inverted, with a spectral index .
It is remarkable that both the East and West source have inverted spectra at low frequencies. We have therefore paid special attention to the accuracy with which we can determine spectral indices, by analysing the spectra of many, faint and bright, sources in the field. The brighest ten sources in the field have a spectral index from 325-380 MHz of about -0.9, typical for bright extragalactic sources selected at low frequencies. The 62 mJy source only west of the East source (within the dotted circle in Fig. 3) in fact has an average spectral index between 325 and 380 MHz of -1.9 in the 1994 and 1996 data, which is identical to within the errors to the value between 325 and 1400 MHz, which is obviously determined with much higher accuracy. We have therefore no doubt that these two variable sources have indeed inverted low-frequency spectra.
Neither the East nor the West source has an optical counterpart in the Palomar Optical Sky Survey. They must therefore be fainter than 20th magnitude (O and E). Observations of the East- variable source with the 1 m. JKT telescope at La Palma on 20 November, 1995, revealed a V = 21.0 non-extended object (seeing ), with a = 0.9, at the exact position of the radio source. Additional imaging of the West variable source with the 1 m. JKT at la Palma during service time observations revealed no optical counterpart down to V = 20.0.
© European Southern Observatory (ESO) 1997
Online publication: June 30, 1998