1 Introduction

The origin of cosmological Gamma-Ray Bursts (GRBs) remains one of the great mysteries of modern astronomy (van Paradijs et al. 2000). Over the past half decade many advances have been made in understanding the nature of the bursts and their afterglows throughout the electromagnetic spectrum. There are at present mainly two sets of models for GRBs. One set of models predicts that GRBs occur when two collapsed objects (such as black holes or neutron stars) merge (Eichler et al. 1989; Mochkovitch et al. 1993). The time-scale for binary compact objects to merge is large ( $\gtrapprox$1 Gyr), so GRBs can occur after massive star formation has ended in a galaxy. The other major set of models predicts that GRBs are associated with the death of massive stars (supernovae or hypernovae) (Woosley 1993; Paczynski 1998; MacFadyen & Woosley 1999). In this case GRBs will coincide with the epoch of star formation in the host. By determining the Spectral Energy Distribution (SED) and star formation rate (SFR) of a sample of GRB host galaxies we can distinguish between these two families of GRB progenitor models (see also Belczynski et al. 2002). Substantial insight has already been gained about the galaxies that the bursts occur in. Radio, optical and/or infrared afterglows have been observed for $\sim$40 GRBs, and the majority of these coincide with starforming galaxies.

As GRB host galaxies tend to be faint (R > 23) spectroscopic studies of the SED are only reachable with 8-10 m class telescopes. A cheaper and elegant alternative to spectroscopy is to extract information on the properties of the host galaxies based on multicolour broad band imaging. By determining the colours of GRB host galaxies we can derive or constrain the age of the predominant stellar population as well as the extinction. As part of the global fit, the photometric redshift of the host galaxies can be derived if the redshift is not known in advance from spectroscopic observations of the afterglow and/or the host galaxy. Additional advantages of the multicolour photometric studies compared to spectroscopic techniques are their simplicity and their multi-object feasibility. The photometric technique allows the determination of the colours of all objects in the field down to the imaging flux limit, thereby in principle permitting the study of the host galaxy environment. The precision of the photometric redshift estimate (which depends on the photometric accuracy, the spectral coverage and the number of bands) is evidently not as accurate as the spectroscopic one, but it is sufficient for a first order study of host galaxy environments.

So far it has been possible to detect optical afterglows for only about 30% of localised GRBs (Fynbo et al. 2001; Lazzati et al. 2002). It is important to understand the nature of the remaining (rather ill-termed) so-called dark GRBs if we wish to get a complete understanding on GRB selected galaxies and thereby constrain the GRB progenitors as well as the distribution of cosmic star formation over different modes (e.g. Ramirez-Ruiz et al. 2002; Venemans & Blain 2001). GRB 000210 is currently one of only few systems that allow a detailed study of a galaxy hosting a dark GRB. The burst exhibited the highest $\gamma$-ray peak flux among the 54 GRBs localized during the entire BeppoSAX operation, from 1996 to 2002 (Piro et al. 2002). However, no optical afterglow (OA) was detected in spite of a deep search (R> 23.5) carried out $\sim$16 hrs after the gamma-ray event (Gorosabel et al. 2000a). X-ray observations performed with the Chandra X-ray telescope 21 hrs after the GRB localised the X-ray afterglow of the burst to an accuracy of $2^{\prime \prime}$, later improved by Piro et al. (2002) to a $0\hbox{$.\!\!^{\prime\prime}$ }6$ radius error circle. The optical search revealed an extended constant source coincident with the X-ray afterglow which was proposed as the GRB host galaxy (Gorosabel et al. 2000b). In addition, Piro et al. (2002) have reported the detection of a radio transient at 8.5 GHz spatially coincident with the X-ray afterglow. Based on the detection of a single host galaxy spectral line, interpreted to be due to [O II], Piro et al. (2002) proposed a redshift of $z=0.8463\pm 0.0002$. Recently Berger et al. (2003) and Barnard et al. (2003) have reported $\sim$2.5 $\sigma$detections of sub-mm emission towards the position of GRB 000210 interpreted as emission from the host galaxy and hence suggesting a SFR of several hundred $M_{\odot}$ yr^-1.

 

 

Table 1: Chronologically ordered optical and NIR observations carried out for the GRB 000210 host galaxy.

Telescope	Filter	Date UT	$T_{\exp}$	Seeing	Limiting magnitude
(+Instrument)			(s)		($4\sigma$)
8.2VLT (+FORS1)	R	25.237-25.240/10/00	300	0.70	25.4 $^{\star\star}$
3.58NTT (+SOFI)	H	02.251-02.410/09/01	182$\times$60	0.90	22.8
3.6ESO (+EFOSC2)	V	13.219-13.253/09/01	4$\times$600	1.75	25.4
3.6ESO (+EFOSC2)	I	13.256-13.278/09/01	3$\times$600	1.45	23.1
3.6ESO (+EFOSC2)	B	13.280-13.302/09/01	3$\times$600	1.70	25.6
3.6ESO (+EFOSC2)	U	13.304-13.348/09/01	6$\times$600	1.55	24.7
8.2VLT (+ISAAC)	Ks	21.159-21.193/09/01	30$\times$60	0.45	22.2
8.2VLT (+ISAAC)	Js	21.194-21.218/09/01	15$\times$120	0.60	24.1$^{\dag }$
8.2VLT (+ISAAC)	Js	23.193-23.218/09/01	15$\times$120	0.75	24.1$^{\dag }$
1.54D (+DFOSC)	Z	19.090-19.254/12/01	14$\times$600	1.10	22.9$^{\star}$
1.54D (+DFOSC)	Z	20.042-20.394/12/01	21$\times$600	1.15	22.9$^{\star}$

$\star\star$ Published in Piro et al. (2002).

$\dag$ The images were coadded resulting in just a single Js-band magnitude.

$\star$ The images were coadded resulting in just a single Z-band magnitude.

In this paper we present the most intensive multi-colour host galaxy imaging performed to date. The host galaxies SED studies to date had a limited number of bands (Sokolov et al. 2001; Chary et al. 2002) and no photometric redshift determinations. Throughout, the assumed cosmology will be $\Omega_{\Lambda} = 0.7$, $\Omega_{M} = 0.3$ and H₀= 65 km s^-1 Mpc^-1 (except in Sect. 5.3 where the host galaxy luminosity is rescaled to the cosmology used by Lilly et al. 1995). At the proposed spectroscopic redshift (z=0.8463), the look back time is 7.59 Gyr (52.4% of the present age) and the luminosity distance is $1.79 \times 10^{28}$ cm. The physical transverse size of one arcsec at z=0.8463 corresponds to 8.24 kpc.