2 Experimental setup and observations

The MARLY telescope and its two cameras, the way we carry out our observations, as well as our data reduction sequence are described in Paper I and references therein. The two EROS passbands are non standard. The so-called EROS-red passband $R_{\rm E}$ is centred on $\bar\lambda = 762\ {\rm nm}$, close to $I_{\rm C}$ Cousins, with a full width at half maximum $\Delta\lambda \simeq 85\ {\rm nm}$, and EROS-visible passband $V_{\rm E}$ is centred on $\bar\lambda = 600\ {\rm nm}$, close to $V_{\rm J}$ Johnson, with $\Delta\lambda \simeq 78\ {\rm nm}$. The EROS II colour magnitude system is defined as follows: a zero colour star with $V_{\rm E} - R_{\rm E} = 0$ (a main sequence A0 star) will have its $R_{\rm E}$ magnitude numerically equal to its Cousins $R_{\rm C}$ magnitude and its $V_{\rm E}$ magnitude numerically equal to its Johnson $V_{\rm J}$ magnitude. The colour transformation between the EROS II system ($V_{\rm E}$,$R_{\rm E}$) and the standard Johnson-Cousins ($V_{\rm J}$,$I_{\rm C}$) system is then:

 

$$I_{\rm C} R_{\rm E} + 0.01 \times (V_{\rm E} - R_{\rm E})$$ = (1)

$$V_{\rm J} V_{\rm E} + 0.39 \times (V_{\rm E} - R_{\rm E}).$$ =

The colour coefficients are obtained from the study of our passbands based on Landolt standards and on one of the EROS II LMC fields observed simultaneously in B$V_{\rm J}$$R_{\rm C}$$I_{\rm C}$ with the Danish 1.54 m (at ESO-La Silla) and with the MARLY. The zero points are established with tertiary standards in BVRI taken with the Danish 1.54 m (Regnault 2000). We have cross-checked our $R_{\rm E}$ photometry with the $I_{\rm C}$ one of DENIS (Fouqué et al. 2000). Furthermore, using Eqs. (1), the mean magnitudes of the LMC red-giant clump stars agree within 0.1 $^{\rm mag}$ with determinations made by Harris & Zaritsky (1999) and Udalski et al. (1998). We thus estimate that the precision of the zero points of the MARLY calibration is $\sim$ $0.1^{\rm mag}$.

 

 

Table 1: Characteristics of the 6 one square degree fields monitored for this study. The equatorial (J2000) and galactic coordinates of the field centres, the number of photometric measurements per light curve per colour $N_{\rm m}$and the number of analysed light curves $N_{\rm obs}$ (in millions) are given for each field. The observation duration was $\sim$100 days.

Field	$\alpha$	$\delta$	$b^{\circ}$	$l^{\circ}$	$N_{\rm m}$	$N_{\rm obs}$
Norma ($\gamma$ Nor)					59	1.30
gn450	16:09:45	-53:07:03	-1.17	330.49	63	0.28
gn453	16:22:28	-52:06:20	-1.69	332.24	52	0.36
gn455	16:26:52	-52:21:02	-2.35	332.54	57	0.30
gn459	16:15:51	-54:48:45	-2.86	329.82	61	0.36
Musca ($\theta$ Mus)					60	0.61
tm550	13:27:04	-63:02:18	-0.47	306.98	61	0.27
tm551	13:31:18	-63:34:41	-1.07	307.37	60	0.34
Total						1.91

Among the 29 fields of the EROS GSA microlensing program, the six fields considered here were monitored about once per night between April and June 1998. They represent 2 square degrees towards $\theta$ Mus and 4 towards $\gamma$ Nor (named after the closest bright star). Table 1 gives their coordinates, the number of images $N_{\rm m}$ taken in each direction and the number of analysed light curves $N_{\rm obs}$.

Figure 1: Average time sampling for the 6 fields monitored (2 towards $\gamma$ Nor, upper panel, and 4 towards $\theta$ Mus, lower panel), in number of measurements per week and per field.

To avoid CCD saturation by the brightest stars ( $I_{\rm C} \sim 9$), by Cepheids or Miras in particular, we have reduced the exposure time to 15 s instead of the 120 s used in the microlensing survey. As a consequence the catalogue is incomplete as far as faint stars are concerned, but it could be updated later by using the total set of available GSA images. Figure 1 shows the average time sampling. Three gaps can be seen in our data: the first two (around weeks 119 and 123) were due to bad weather conditions while the third one (around week 127) corresponds to the annual maintenance of our setup.