next previous
Up: A three-year Strömgren photometric stars


Subsections

3 Data analysis

Using standard reduction programs, the uvby observations were reduced by the SAT group of the Copenhagen University Observatory (see Crawford & Barnes 1970; Olsen 1994). The data are available in files containing the Henry Draper (HD) number of the star, the Julian Date (JD) of the observation, the air mass factor of the observation, the Strömgren V magnitude, the b-y color index, the m1=(v-b)-(b-y) and the c1=(u-v)-(v-b) parameters.

   
3.1 Short term variations

From our set of data, it was possible to search for short term as well as long term variations. The data analysis were performed on differences between stellar measurements and the corresponding measurements made on a comparison star. Short term variations would be detected if two subsequent measurements differ by more than 3 sigma from the average value of the data set. An eventual detection can be confirmed when comparing the star to the other comparison star. Over our three campaigns, no fluctuation was observed in any of our program stars except two of them, HR 10 and HD 38392 which will be discussed below.

This result already shows the quality of the survey since we obtained on the average the following precisions on the different parameters evaluated: $\sigma_V \mathrel{\mathchoice {\vcenter{\offinterlineskip\halign{\hfil
$\displa...
...offinterlineskip\halign{\hfil$\scriptscriptstyle ... mag for V and $\sigma_{b-y} \mathrel{\mathchoice {\vcenter{\offinterlineskip\halign{\hfil
$\di...
...ffinterlineskip\halign{\hfil$\scriptscriptstyle ... mag for the b-y color index. Within this accuracy, most of our surveyed stars did not present any short term variations, including $\beta$ Pictoris.

   
3.2 Long term variations

Concerning long term variations, we will illustrate here our approach directly in the case of $\beta$ Pictoris, since it is also the only case where some long range variation were detected.

Figure 1 shows the Strömgren V magnitude evolution between October 1995 (JD 2450030) and March 1998 (JD 2450880) for $\beta$ Pictoris and its two comparison stars HD 35580 and HD 40200. It is clear that the three stars present similar variations which are thus not intrinsic variations but systematic residuals. However it is noteworthy that the statistical dispersion of individual data is at this stage of the order of only $\sigma \mathrel{\mathchoice {\vcenter{\offinterlineskip\halign{\hfil
$\displays...
...ffinterlineskip\halign{\hfil$\scriptscriptstyle ... mag.

To detect possible weak variations, as in the case of our search for short term variations, we evaluate the difference between the data of a given star and its comparison star, provided that the measurements are contiguous in time. Practically, we set a maximum time difference between two measurements and we consider only the couples of measurements separated by less than 0.004 day ($\sim$6 min). This process is done for the Strömgren Vmagnitude, the b - y color index, the m1 parameter and the c1 parameter of all the stars. Figure 2 shows the magnitude difference between $\beta$ Pictoris and its two comparison stars (HD 35580 and HD 40200), whereas Fig. 3 shows the difference for the b-y color index.


  \begin{figure}
\includegraphics[width=8cm,clip]{plot_bp_v.ps}\end{figure} Figure 1: Plot of the Strömgren V magnitude of $\beta\:$ Pictoris, HD 40200 and HD 35580. We see that the data of the three stars present the same apparent signature of variations. These variations are obviously not real; they are only due to a residual noise identical for each star


  \begin{figure}
\includegraphics[width=8cm,clip]{plot_bp_vdiff.ps}\end{figure} Figure 2: Plot of the magnitude differences of the same three stars. Here it is clear that the apparent variations visible in the previous figure have been removed. The difference between the two comparison stars is flat, and neither short term nor long term variations are detected over the three years of observations (bottom panel). This is exactly the case of all other stars of the program. In contrast, $\beta\:$ Pictoris shows a long term variation with a significant slope which furthermore and as expected, does not depend upon the comparison star used for the subtraction


  \begin{figure}
\includegraphics[width=8cm,clip]{plot_bp_bydiff.ps}\end{figure} Figure 3: Same as the previous figure but for the b-y color index. The slopes of the three differences (solid lines) are not significantly different from zero. The long term variations observed in the case of $\beta\:$ Pictoris are thus relatively achromatic

As already mentioned, no short-term variations are detected in the case of $\beta$ Pictoris. To detect long-term variations, we used the same method as described in Lecavelier et al. (1995a). We fitted the data of each star (s) by a linear function of the time (t): ( $a+b\cdot t$) and evaluate the probability that the calculated slope $b_{\rm s}$ for the star s is far from 0 assuming that b=0 is true. For a Gaussian distribution of the statistical noise in the data, the distribution of $b_{\rm s}/\sigma_{b_{\rm s}}$ must follow a Student law. Here we find that the Student law is well followed and consequently that the probabilities $P(b>b_{\rm s})$ are uniformly distributed between 0 and 1 (see an example in Lecavelier et al. 1995a). For all the stars (s) except $\beta\:$ Pictoris, we do not find a $b_{\rm s}$ value significantly different from 0. This shows that our program stars as well as the comparison stars had no significant light variations during the survey. Again this underlines the quality of the survey and the strength of the very few real detections that are presented now.


 

 
Table 2: List of the UBVB1 B2 V1G Geneva photometric system program stars. PT is for our main targets or "Primary Targets''. CMP is for the comparison stars
  star * $\alpha_{2000}$ $\delta_{2000}$ mV Spectral type $v \sin i$
HR 10 HD 256 PT 00h 07m 18.3s -17$^\circ$23' 13'' 6.23 A2IV/V 220
  HD 693 CMP 00h 11m 15.8s -15$^\circ$28' 04'' 4.89 F5V 8
$\beta$ Pictoris HD 39060 PT 05h 47m 17.1s -51$^\circ$04' 00'' 3.80 A5V 104
  HD 35580 CMP 05h 22m 22.1s -56$^\circ$08' 04'' 6.08 B8.5V  
HR 2174 HD 42111A PT 06h 08m 57.8s +02$^\circ$29' 58'' 5.73 A3Vn 120
  HD 41692 CMP 06h 06m 38.7s -04$^\circ$11' 38'' 5.39 B5IV 20
51 Ophiuchi HD 158643 PT 17h 31m 24.8s -23$^\circ$57' 46'' 4.81 A0V 210
  HD 155450 CMP 17h 12m 58.6s -32$^\circ$26' 19'' 6.01 B1II  
TY CrA CD -37$^\circ$13024B PT 19h 01m 40.1s -36$^\circ$52' 33'' 9.54 Ae  
  HD 180885 CMP 19h 19m 39.9s -35$^\circ$25' 17'' 5.59 B3V 158
68 Ophiuchi HD 164577 PT 18h 01m 45.1s +01$^\circ$18' 19'' 4.45 A2Vn 232
  HD 157089 CMP 17h 21m 07.0s +01$^\circ$26' 35'' 6.95 F9V  



next previous
Up: A three-year Strömgren photometric stars

Copyright The European Southern Observatory (ESO)