2 The background sources

2.1 Supernovae

The supernovae, especially those of Ia type, may represent an important source of confusion due their occurrence rates, rise and decay timescales, and magnitudes. The expected rate of occurrence for faint events (down to $R\,23$ mag) is roughly 2 deg^-2 or 0.0015 arcmin^-2 (Evans et al. 1989; Pain 1996; Brainerd 1998). But, at least some, SNe may be probably related to GRBs (SN 1998bw and GRB 980425).

2.2 AGN flares

The spread of particular AGN flare amplitudes is large (0.1... 6.7 mag). Recently, there is growing evidence for large amplitude (more than 10 mag) flares on AGNs (Hudec et al. 1996). Before, the largest known QSO amplitude for 3C 279 was more than 6.7 mag (Eachus & Liller 1975) but with undersampled light curve.

The expected AGNs/QSOs surface densities are $\sim$(10-37) deg^-2 with lim mag $B/V \sim 20.5$ (Iovino et al. 1996; Hartwick & Scade 1990) and 111 deg^-2 with lim mag B 22.6 (Trevese et al. 1989). Most of the QSOs are variable: the typical QSO variability in a sample of 149 opt. selected QSOs is 0.26-0.33 in B 0.22-0.30 in R (Cristiani et al. 1997). There are $\sim 100$ variable QSOs deg^-2 (by more than 0.1 mag) i.e. 1 variable QSO in $6 \times 6$ arcmin² brighter than B 22.6 (Trevese et al. 1989). 97% QSOs below B 22.5 are variable (Trevese & Kron 1994). The QSO variability seems to increase with decreasing luminosity (Cristiani et al. 1997).

2.3 Stellar flares

There is growing evidence for large amplitude (5 mag and more) stellar flares. Flare stars detected by extended plate surveys can serve as an example: e.g. the two pairs of OTs-flare stars detected by Greiner & Motch (1995) and Hudec et al. (1997) which seem to represent large amplitude flares (5-9 mag) from otherwise typical dMe flare stars. The exact statistics of such events is however unknown.

2.4 Variable stars

Also variable stars may exhibit light behavior similar to OAs and OTs. The Y Dra Mira type variable star located inside the error box of the GRB 910709 detected and positioned by COMPTEL can serve as an example. The star exhibits light variations between 6 and 15 mag, with gradual light decrease after maximum (period 322 d). Although the physical relation would be difficult to explain, the object shows a maximum almost exactly coincident with the GRB date. The estimated rates are available only for variable stars brighter than 20 mag: 80 deg^-2 for /bII/ < 20 deg and 4 deg^-2 for /bII/ > 40 deg (Hudec & Wenzel 1996), however the discovery probability is $\sim$ 0.1 (blinkmicroscope use). No statistics for variable stars below 20 mag is available (no systematic surveys). Variable stars are observed more commonly in decline than in increase since the declines are typically more slowly such as delta Cep stars, U Gem stars, flare stars, novae etc.

2.5 OTs of unknown nature and origin

There are real OTs of unknown nature but of astrophysical origin detected both on emulsions and CCDs. Examples (real CCD detections): (1) OT 970215. A real CCD detection with guiding error (trailing), V 13 mag, not on any of other 800 CCD observations, nothing down 20 mag on the position, amplitude more than 7 mag (Vidal-Saiz et al. 1996). (2) OT 950806. Confirmed real object: detected on 20 CCD frames, peak magnitude I 7.5, amplitude more than 10 mag, nothing down mag 21 48 hrs after detection (Toth et al. 1996).