I recently proposed that GRB afterglows can be found in single epoch
observations using optical color-color diagrams (Rhoads 2000,
astro-ph/0008461). This is possible because afterglows generally have
power law spectra, which can be distinguished from curved stellar
spectra in color-color plots.
GRB 000830 may be a good test case to try this method,
given its medium-sized error box and the nearly moonless sky.
The required observations would include I band, either U or B band,
and either V or R band. The approximate colors expected for an
afterglow are U-B=-0.75, B-V=+0.35, V-R=+0.4, R-I=+0.5 (for standard
Johnson-Cousins magnitudes and an f_nu ~ nu^-1 power law spectrum).
The relative depth required in different filters can be estimated from
these colors. Absolute photometric calibration is not strictly needed
so long as enough area is covered to get a clearly defined stellar
locus in the color-color plane.
The required photometric accuracy depends on the spectral coverage.
With U, V and I filters, the afterglows are expected to lie about 0.7
magnitudes from the stellar locus. Using instead B, V, R, I filters
gives a smaller offset of about 0.2 magnitudes. The large color-plane
offset with the U filter approximately balances the extra difficulties
of obtaining U band data.
The middle filter (i.e., V or R) should be observed last, so that
fading behavior will serve to increase the apparent offset of the
afterglow from the stellar locus rather than reducing it.
Given the 38 square arcminute error box of GRB 000830, one might expect
about one quasar of magnitude R ~ 22 in the error box; any brighter point
source with the expected colors of an afterglow is unlikely to be a quasar,
while at fainter magnitudes multiple candidates (most of them quasars) would
More details of the method are available
from http://xxx.lanl.gov/abs/astro-ph/0008461 .