arXiv:astro-ph/9912402v1 18 Dec 1999
LOTIS Upper Limits and the Prompt
OT from GRB 990123
G. G. Williams∗, D. H. Hartmann∗, H. S. Park†, R. A. Porrata†,
E. Ables†, R. Bionta†, D. L. Band¶, S. D. Barthelmy§, T. Cline§,
N. Gehrels§, D. H. Ferguson‡, G. Fishman⋆, R. M. Kippen⋆,
C. Kouveliotou⋆, K. Hurley♯, R. Nemiroff♭, and T. Sasseen||
∗Dept. of Physics and Astronomy, Clemson University, Clemson, SC 29634
†Lawrence Livermore National Laboratory, Livermore, CA 94550
¶Los Alamos National Laboratory, Los Alamos, NM 87545
§NASA Goddard Space Flight Center, Greenbelt, MD 20771
‡Dept. of Physics, California State University, Hayward, CA 94542
⋆NASA Marshall Space Flight Center, Huntsville, AL 35812
♯Space Sciences Laboratory, University of California, Berkeley, CA 94720
♭Dept. of Physics, Michigan Technological University, Houghton, MI 49931
||Dept. of Physics, University of California, Santa Barbara, CA 93106
Abstract. GRB 990123 established the existence of prompt optical emission from
gamma-ray bursts (GRBs). The Livermore Optical Transient Imaging System (LOTIS)
has been conducting a fully automated search for this kind of simultaneous low energy
emission from GRBs since October 1996. Although LOTIS has obtained simultaneous,
or near simultaneous, coverage of the error boxes obtained with BATSE, IPN, XTE,
and BeppoSAX for several GRBs, image analysis resulted in only upper limits. The
unique gamma-ray properties of GRB 990123, such as very large fluence (top 0.4%)
and hard spectrum, complicate comparisons with more typical bursts. We scale and
compare gamma-ray properties, and in some cases afterglow properties, from the best
LOTIS events to those of GRB 990123 in an attempt to determine whether the prompt
optical emission of this event is representative of all GRBs. Furthermore, using LOTIS
upper limits in conjunction with the relativistic blast wave model, we weakly constrain
the GRB and afterglow parameters such as density of the circumburster medium and
bulk Lorentz factor of the ejecta.
The ultimate reward for the Gamma-Ray Burst Coordinates Network (GCN) 
came when the Robotic Optical Transient Search Experiment (ROTSE) detected
prompt optical emission from GRB 990123 . Although this discovery marks
another milestone in comprehending the physics of GRBs, bright optical tran-
sients (OTs) may be the exception rather than the rule. Both LOTIS and ROTSE
have unsuccessfully attempted to detect these predicted flashes on many occa-
sions [3–8]. Although some of the non-detections may be attributed to large ex-
tinction, GRB 990123 demonstrated that the progenitor is not always obscured.
OBSERVATIONS & ANALYSIS
During more than 1100 nights of possible observations (since October 1996), LO-
TIS has responded to 127 GCN triggers. Of these, 68 triggers were unique GRB
events; a rate of approximately one unique GRB event every 16.5 days. The quality
of the LOTIS “coverage” for a given event depends on five factors: observing con-
ditions, LOTIS response time, difference between the initial and final coordinates,
size of the final error box, and the duration of the GRB. Table 1 lists 13 events for
which LOTIS achieved good coverage.
First we compare GRB 990123 with the LOTIS upper limits to test whether the
flux of the prompt optical emission scales with some gamma-ray property. Here
and throughout the analysis we neglect extinction effects. The first row in Table 1
lists the properties of GRB 990123 [9,2]. The columns display the UTC date of
the burst, the BATSE trigger number, the 64 ms and 1024 ms peak fluxes (50 -
300 keV), and the gamma-ray fluence (>20 kev) of each event. The last three
columns are the scaled magnitudes,
mGRB= mGRB990123− 2.5log
TABLE 1. LOTIS GRB events with good coverage and predictions for the scaled magnitudes
of the prompt optical emission.
Fp(64 ms)Fp(1024 ms)S/10−7
where mGRB990123 = 9.0, the peak magnitude of GRB 990123, and XGRB and
XGRB990123are the peak flux or fluence values for those events.
The LOTIS sensitivity varies depending on observing conditions but in general
a conservative limiting magnitude is m ≈ 11.5 prior to March 1998 (upgrade to
cooled CCD) and m ≈ 14.0 following that date. Table 1 shows that the scaled
prompt optical emission for both peak flux and fluence is often brighter than the
LOTIS upper limits which suggests that these simple relationships are not valid.
Briggs et al.  show that the optical flux measured during GRB 990123 is not
consistent with an extrapolation of the burst spectrum to low energies. However
Liang et al.  point out that the extrapolated tails rise and fall with the optical
flux. A low energy enhancement would produce an upward break which might
account for the measured optical flux during GRB 990123. It is important to
determine if there is a low energy upturn in the spectrum since it would establish
whether or not the optical and gamma-ray photons are produced by the same
electron distribution. The LOTIS upper limits can be used to constrain a low
energy enhancement assuming it is common to all GRBs.
For the events listed in Table 1 we fit the gamma-ray spectra during the LOTIS
observations to the Band functional form .
extrapolation is near the LOTIS upper limit. The solid line in Figure 1 shows
the Band fit to GRB 971006 and its extrapolation to low energies. Fits to the
spectra of GRB 990123 during the first (short dash), second (dash-dot), and third
In a few cases the low energy
FIGURE 1. Extrapolated spectrum of GRB 971006 during the LOTIS observation and
GRB 990123 during ROTSE detections.
FIGURE 2. Predicted magnitude of the prompt optical flash for E52 = 3.5, ǫe = 0.12, and
ǫB= 0.089 (left panel) and E52= 0.53, ǫe= 0.57, and ǫB= 0.0082 (right panel).
(long dash) ROTSE observations and the corresponding ROTSE detections (filled
circles) are also shown. The extrapolation of GRB 971006, predicts an m ≈ 12.4
optical flash. Even a slight upward break in the spectrum would have produced a
detectable OT. We conclude that the LOTIS upper limits support the hypothesis
that the low energy emission is produced by a different electron distribution than
the high energy emission.
Finally we attempt to use the LOTIS upper limits and the external reverse
shock model to constrain the physical properties of the GRB blast wave. Sari
and Piran  show that the fraction of the energy which gets emitted in the
optical band depends on the values of the cooling frequency and the characteristic
synchrotron frequency. For the external reverse shock these frequencies are given
νc= 8.8 × 1015Hz
νm= 1.2 × 1014Hz
where ǫeand ǫBare the fraction of equipartition energy in the electrons and mag-
netic field, E52is the total energy in units of 1052erg, n1is the density of circum-
burster medium in cm−3, γ0is the initial Lorentz factor, and tAis the duration of
the emission in seconds.
Sari and Piran assume the frequency dependencies modify the fluence of a mod-
erately strong GRB, i.e. 10−5erg cm−2. In this analysis we compare the afterglow
properties of GRB 970508 found by Wijers and Galama  to those found by Gra- Download full-text
not et al. . Therefore we use a fluence of 3.1 × 10−6erg cm−2emitted over the
entire LOTIS integration time of tA= 10.0 s. The index of the electron power-law
distribution is set to p = 2.2.
Figure 2 shows contour plots of the predicted magnitude of the prompt OT for
GRB 970508 as a function of n1and γ0. GRB 970508 could not be observed by
LOTIS or ROTSE since it occurred during the day. Values of E52= 3.5, ǫe= 0.12,
and ǫB= 0.089 from Wijers and Galama are used in the left panel and values of
E52= 0.53, ǫe= 0.57, and ǫB= 0.0082 from Granot et al. are used in the right
panel. The right panel demonstrates the effect of altering the total energy and the
distribution of energy to the electrons and the magnetic field. The smaller values
of E52and ǫBshift the contours to the upper left while the larger ǫesteepens the
breaks in the contours. The increased shading corresponds to a decreasing detection
probability. However for nearly all values of n1and γ0shown the predicted OT
could have been detected by the upgraded LOTIS system.
Wijers and Galama find a circumburster medium density of n1= 0.030 which
predicts an m = 9.0 − 9.5 optical flash nearly independent of the initial Lorentz
factor. Granot et al. find a considerably higher vlaue of n1= 5.3, which predicts
an m = 8.7 − 12.4 OT which is very dependent on the initial Lorentz factor. The
LOTIS upper limits mildly favor the GRB blast wave values determined by Granot
et al. since dim OTs are predicted over a larger range of initial Lorentz factors.
1. Barthelmy, S. D., et al., ”The GRB Coordinates Network (GCN): A Status Report”,
in Gamma-Ray Bursts 4th Huntsville Symposium, edited by C. A. Meegan, R. D.
Preece, & T. M. Koshut, AIP Conf. Proc. 428, Woodbury, 1998, pp. 99-103.
2. Akerlof, C. W., et al., Nature 398, 400-402, (1999).
3. Park, H. S., et al., ApJ 490, 99-108 (1997).
4. Park, H. S., et al., ApJ 490, L21-L24 (1997).
5. Williams, G. G., et al., ”First Year Results from LOTIS”, in Gamma-Ray Bursts
4th Huntsville Symposium, edited by C. A. Meegan, R. D. Preece, & T. M. Koshut,
AIP Conf. Proc. 428, Woodbury, 1998, pp. 837-841.
6. Williams, G. G., et al., ApJ 519, L25-L29 (1999).
7. Schaefer, B. E., et al., ApJ 524, L103-L106 (1999).
8. Kehoe, R., et al., submitted to Proceedings of the 1999 May Symposium of the Space
Telescope Institute, (1999); (astro-ph/9909219).
9. Briggs, M. S., et al., ApJ 524, 82-91 (1999).
10. Liang, E. P., et al., ApJ 519, L21-L24 (1999).
11. Band, D. L., et al., ApJ 413, 281-292 (1993).
12. Sari, R., and Piran, T., ApJ 520, 641-649 (1999).
13. Wijers, R. A. M. J., and Galama, T. J., ApJ 523, 177-186 (1999).
14. Granot, J., et al., ApJ, submitted (1999); (astro-ph/9808007).