Fig 2: Stellar atmospheres: Colour-colour diagram  of stars in the CFHTLS W3 field (see below) compared to the synthetic colour-colour diagram from Basel 3.1 stellar libraries(left panel) and NextGen (right panel) for different metallicities.  The red lines are for [Fe/H]=0.0,  the green lines  [Fe/H]=-1.0 and the blue lines [Fe/H]=-2.0. We use only stars with a photometric error smaller than 0.01 mag in each filter for a more accurate stellar locus. Approximate temperatures are also indicated.

Figure 2 shows g'-r' versus r'-i' diagrams observed with Megacam, superimposed with the synthetic colours of dwarf stars using the Basel3.1 stellar library for solar metallicity, [Fe/H]=-1.0  and [Fe/H]=-2.0 (left panel) and the NextGen library for [Fe/H]=0.0 and [Fe/H]=-1.0  (right panel). We used here only stars which have a photometric error smaller than 0.01mag in each filter. The diagram illustrates the sensitivity of the  colours in the MEGACAM system  to metallicity and also the differences in the stellar libraries. Especially at temperatures below 3500 K which corresponds to K/M dwarfs, the Basel3.1 library does not give realistic colours. The r'-i' colour is sensitive to effective temperature while g'-r' is only sensitive to it at T_eff > 4000 K.  Note the striking difference in the atmosphere models in g'-r' at low temperatures. For late type dwarfs the most sensitive metallicity indicator seems to be the g'-r'colour. For objects with g'-r' > 1.0 one can distinguish, assuming a very accurate photometry, between stars with solar metallicity and [Fe/H]=-1.0. The comparison between both libraries and MEGACAM data lead us to adopt  Basel3.1 colours as a base and complementary NextGen colours for T_eff < 4000 K.

The Besançon model with the MEGACAM filter system

In the case of the MEGACAM photometric system, we have used the ccd+filter definition of the passbands, and applied these bandpasses to the spectral libraries. The model is otherwise the same as the standard version (Robin et  al, 2003).

Comparison between CFHTLS data and the Besançon model simulations

The CFHT Legacy Survey project is a large observationnal project with the CFH Telescope equiped with the new primary focus Megaprime and the Megacam instrument. It consists of three different surveys. The Deep is a survey of 4 patches  (D1,D2,D3,D4)  of 1 square degree in the filters u*,g',r',i'  and z', complete to r' ~ 28 mag; the Wide survey covers 3 patchs (W1,W2,W3) of ~ 7 deg², each observed to r' ~ 25 mag, and the Very Wide is dedicated to the ecliptic and covers 1300 deg² to r'~23. We refer for a detailed description to the corresponding web pages (CFHTLS). All the observations are reduced by the TERAPIX pipeline at the Institut d'Astrophysique de Paris.

As a test for the implementation of the Megacam filter set, we here present a simulation from the Besançon model in the MEGACAM filters to be compared with preliminary data of a subset of 4 deg² of the W3 field located at l = 98.85 and b = 58.4. The cut-off limit in i' is 20.5 imposed by the star/galaxy separation.

•  The colour-colour diagram:  Figure 3 shows the colour-colour diagrams of the CFHTLS (bottom panel) and the
  model predictions (top panel) for the W3 field. We indicate the three different galactic components  by different colours : thin disc (red), thick disc (green), spheroid (magenta). Note the excellent overall agreement between observed and predicted colours for the three populations.
  

Fig 3: g'-r' vs. r'-i' diagram. The bottom panel shows the data of the W3 field. The top panel shows the synthetic colours predicted by the Besançon model. The red points are thin disc stars, the green points are for the thick disc and magenta for the spheroid.

The blue part of the observed colour-colour diagrams is populated by a population which has a larger spread than the predicted stellar locus. We suspect these objects to be compact faint galaxies misclassified as stars from the morphological criterium used here. Additional proper motion studies and/or spectroscopy would help to make a cleaner separation between stars and galaxies.

•  The colour-magnitude diagram: Figure 4 shows the colour-magnitude (CMD) diagrams of a subset of the W3 field. As for the colour-colour diagram, the model predictions reproduce well the observations.The overpopulation of blue objects in the observations is due to compact galaxies and quasars, as already seen in the colour-colour diagram. We indicate the location of white dwarfs (open circles)  which are characterized by their blue r'-i' colour.

Fig 4: Colour-magnitude diagram of a subset of data in the W3 field. The left panel shows the W3 data and the right panel the synthetic colour-magnitude diagram derived from the Besançon model. The symbols are the same as in Fig.3. The open circles denote the location of the white dwarf populations.

•  The colour histogram:  Figure 5 shows the colour-distribution of the W3 field. Note the good agreement between the numbers expected from the Galaxy model (dashed line) and the observation (black solid line) for the three galactic components, although some colour shift at high temperature and/or low metallicity (blue side of the histogram) might be present.
   
  

Fig. 5: Histograms of the g'-r' colours. The black solid line is the distribution in colours of the stars in a field of 4 sqaure degrees, a subset of the W3 field of CFHTLS. The dashed black line the colour distribution predicted by the Besançon model. The red line is the thin disc population, the green line the thick disc and in magenta the spheroid population.

e-mail: mathias at obs-besancon.fr, annie at obs-besancon.fr