Reichart, D. (U. North Carolina) comments:
Assuming that the new, bright source that Zerbi et al. (GCN 2466) have
identified in H and Ks in the very small (<6") S1 XMM-Newton error circle
(Santos-Lleo & Calderon, GCN 2464) is indeed the afterglow of GRB 031203
(Gotz et al., GCN 2459), the deep limits that Bailyn et al. (GCN 2468)
measure at the location of this source in I and J do indeed imply a sharp
spectral break between the H and J bands: J - H > 5 mag and H - Ks ~ 1 mag
when corrected for Galactic extinction (Bailyn et al., GCN 2468).
Although one explanation for this dropout signature might be that GRB
031203 is at redshift z ~ 10, this is difficult to reconcile with the
brightness of the afterglow (Price, private communication), which is
probably in the top 10% of all afterglows.
Nor can so sharp a spectral break be explained by extinction by regular
dust, whether Galactic or in the source frame.
However, the dust along the line of sight, at least within hundreds of
parsecs of the GRB, should *not* be regular. Although dust within parsecs
to tens of parsecs of the GRB will be returned to the gas phase by
sublimation, dust at greater distances -- out to hundreds of parsecs --
will be repeatedly fragmented by grain charging and Coulomb explosions
(Waxman & Draine 2000, Fruchter, Krolik & Rhoads 2001, Reichart 2001).
If most of the dust fragments down to PAH sizes, its visual absorption edge
-- which can be quite sharp -- will shift to the source-frame R or I band
(the exact wavelength depends on the size of the grains; Li & Draine 2001).
Longer-wavelength light will pass through this dust as if were not even
For a GRB at a typical redshift of z ~ 1, this visual absorption edge will
be redshifted to between the J and H bands. Such a redshift would also be
easier to reconcile with the brightness of the GRB 031203 afterglow in the
H and Ks bands.
So actually, every highly extinguished, low- to moderate-redshift GRB
afterglow might look like this in the NIR.
For GRB 031203, this hypothesis can be tested by (1) pursuing NIR
spectroscopy tonight while the afterglow is still sufficiently bright, but
be warned that if at z ~ 1 H-alpha might be blueward of the spectral break,
and/or (2) pursuing deep imaging blueward of the spectral break in hopes of
identifying a typical-redshift host galaxy, but again be warned: since the
Galactic A_V ~ 3 mag and possibly more along this line of sight, J, z, and
I bands are preferred.
The Follow-Up Network for Gamma-Ray Bursts (a.k.a. the FUN GRB
Collaboration), the resources of which are currently concentrated more in
the northern hemisphere, will not be able to pursue either of these
observing strategies tonight.