SpringerLink
Forum Springer Astron. Astrophys.
Forum Whats New Search Orders


Astron. Astrophys. 364, L54-L61 (2000)

Previous Section Next Section Title Page Table of Contents

4. Follow-up observations and data analysis

Additional images of the field were obtained on 2000 Feb. 8.1 UT. The source was detected in a deep FORS1 R-band exposure and was found to have faded by 1.9[FORMULA]0.2 mag relative to the first epoch confirming its transient nature. As the source, based on the V, R and I photometry, appeared to be very red, we also acquired near-infrared (JHKs) images with the ESO New Technology Telescope (NTT). A final set of R and I band images were acquired with FORS1 on 2000 Mar. 5.0 UT. A log of our photometric observations and the standard calibrated magnitudes of the afterglow is given in Table 2. All image reductions were performed in the IRAF 1 environment.

4.1. Image analysis and photometry

The afterglow was detected at low signal-to-noise in most of the images, which made it essential to use PSF photometry to derive the magnitudes. We used DAOPHOT-II (Stetson 1987) within IRAF to derive PSF magnitudes of all point sources in the images. In general, when deriving PSF magnitudes, the center and amplitude (magnitude) of the PSF is fitted. For objects close to the detection limit, fitting the centroid introduces a bias in the magnitude of up to one magnitude, because of a tendency to fit the PSF to a nearby noise spike. If instead DAOPHOT is used on objects with accurately known pixel coordinates, the magnitude can be derived without re-centering the object during the PSF fit. In addition to removing the magnitude bias, deriving the PSF magnitude without re-centering also has the advantage that magnitudes down to the 2-[FORMULA] level can be obtained. We could therefore use the accurate position of the afterglow, as determined from the first epoch images, to derive reliable magnitudes from the observations at later epochs. Limiting magnitudes were obtaind by adding a set of artificial point sources to the images at locations of apparently blank sky, setting the limiting magnitude at the level where DAOPHOT ALLSTAR could recover 95% of the point sources.

4.2. Optical images

The optical observations were all acquired on nights with photometric sky quality. The photometry was calibrated to standard star fields (Landolt 1992) through CCD aperture photometry relative to the comparison stars marked in Fig. 3 and listed in Table 3. The errors in the magnitudes of the reference stars are dominated by the determination of the photometric zero points. Extinction correction was applied using observatory standard extinction values. Because the standard star exposures were acquired at approximately the same airmass as the object exposures, corrections were in all cases smaller than the individual photometric error of the afterglow magnitude. From the errors in the color transformation, it is estimated that the photometric zero point errors are smaller than 0.02 mag. In the R and I images from Mar. 5, there is no detection of a source within 2" of the afterglow location. This implies an upper limit to the host galaxy magnitude of R = 25.7 and I = 24.8.

[FIGURE] Fig. 3. R-band image of the GRB 000131 afterglow from Feb. 4. The afterglow is marked with two bars and the four comparison stars are labeled according to Table 3. North is up and East is to the left.


[TABLE]

Table 3. Magnitudes of internal reference stars.


4.3. Near-infrared images

Near-infrared (IR) J, H, and Ks-band images were obtained at La Silla with the NTT and SOFI on 2000 Feb. 8.1 UT. SOFI is equipped with a Hawaii 1024[FORMULA]1024 pixel HgCdTe detector, and we used a plate scale of [FORMULA] which gives a field-of-view of roughly [FORMULA]. Each image comprises several tens of elementary co-added frames acquired by randomly dithering the telescope several arcsec once a minute.

The frames were reduced by first subtracting a mean sky, obtained from frames acquired just before and after the source frame. Before using the frames for sky subtraction, stars were eliminated with an automatic star finder coupled with a background interpolation algorithm. Then, a differential dome flat-field correction was applied, and the frames were registered to fractional pixels and combined. We calibrated the photometry with standard stars taken from Persson et al. Persson et al. (1998), acquired directly before and after the source observations. The standards were placed in five different positions on the detector, and reduced in the same way as the source frames. Standard star and source photometry was corrected for atmospheric extinction using the mean ESO La Silla extinction coefficients given in Engels et al. Engels et al. (1981). Formal photometric accuracy, as judged from the standard deviation of the standard star observations, is about 0.03 mag.

At the time of observations, about 8 days after the IPN detection, the afterglow was very faint and thus only marginally detected in H (3 [FORMULA]) and K (2.5 [FORMULA]). In J there is apparently a source, but it is displaced about 1 pixel from the location of the afterglow. A PSF magnitude of this source could only be derived if re-centering during the PSF fit was allowed. To maintain consistency in our magnitudes, we have therefore chosen to disregard this apparent source and only give the proper limiting magnitude. The H and K magnitudes are both very close to the detection limit. It was verified that magnitudes at this level can be reliably derived by performing photometry of a large number of artificially added stars of the same magnitude.

4.4. Spectroscopy

Starting 2000 Feb. 8.09 UT, we obtained spectroscopic observations, using FORS1 configured with a 300 lines/mm grism, a cut-on order sorter filter and a [FORMULA] slit. The wavelength range covered was 3800 Å - 8200 Å, with a resolution of 9 Å and with order overlap from about 7600 Å. The total integration time was 10800 s, divided into 6 exposures of 1800 s. During the spectroscopic observations, the seeing ranged from [FORMULA] to [FORMULA]. The combined spectrum was flux calibrated relative to the flux standard LTT3218. The continuum is not well detected at visible wavelengths but is weakly detected in the red part of the spectrum, with a signal-to-noise around 1 per spectral bin of 2.5Å in regions that are not dominated by sky lines. The spectral slope is consistent with the broad-band photometry. The spectrum does not show signs of statistically significant emission lines, as could be expected from a host galaxy.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: December 15, 2000
helpdesk.link@springer.de