OT Gem has been observed from January to April 1996 with the 57 cm telescope at the La Luz Observatory (University of Guanajuato, Mexico, thereafter LLO). The technique used was differential photometry carried out with an Optec SSP-5 A photometer equipped with a Hamamatsu R1414 side-on photomultiplier and the standard B and V Johnson's filters. HR 2780 (HD 57049, A2Vn, V = 6.45) was the comparison star C1, and HR 2858 (HD 59059, B9IV, V = 6.22) was the comparison star C2.
Figure 1: 1996 observations of OT Gem carried out at La Luz Observatory.
The top panel displays OT Gem (P)-HR 2780
(C1) light variations in Johnson's V filter. The medium panel
shows the same in B-V. The bottom panel
shows HR 2780 
(C2) in V filter
In order to show the night to night variations, the nightly measurements
have been averaged. These averages and their standard deviation of the mean
are plotted in Fig. 1 (click here). The top panel shows, in the intrumental system,
the differences 
in V light, while the medium
panel shows the same in B-V. The bottom panel displays the
differences through the V filter; all three panels are
plotted at the same scale. The 
error bars have an average size
mag They are due to internal uncertainties that are driven mainly by improper centering or
telescope track imperfections. Typically 20-40 independent measurements were
used to calculate the nightly means. The non-variable nature of both C1
and C2 is evident while well developed variations in OT Gem are observed.
The medium panel of Fig. 1 (click here) shows no variations in the color. This is of most importance since it implies no temperature variations in the stellar photosphere during the light variations. This strongly suggests that the light variations are not due to classical radial or low order non-radial pulsations.
The night-to-night observations were explored in search of
short-term variations. No variations above our internal
accuracy of mag were observed in any of the nights.
High order g-modes could account for periods in the range of days (Dziembowski 1994) that we do not find here. An alternative scenario is of course that of stellar "spots" irregularly formed in time scales of the order of the light variations.
Poretti (1982) and Bozic et al. (1982), reported that no significant light or color variations during one night were observed. Likewise, we do not observe any significant period within the 2-5 hours range, which was the typical length of the nights. Even after subtracting the mean of each night to each night data, no positive detection of a period was accomplished. The period searches were done using both Fourier analysis and the PDM method (Stellingwerf 1978).
In order to look for eventual periodicities in the night to
night variations, we have performed a period analysis on the OT
differences in V light data in the following
manner: first the full set of individual observations were included and two frequencies were isolated: 0.012 cycles/day
(period P = 83 days) and 
.
These can be interpreted as characteristic times of the long-term variation, and not as strict periodicities. Secondly, the same analysis was carried out on the
nightly means of the top panel of Fig. 1 (click here). Essentially the same frequencies
were found: 
and 
.
| JD (2500000.+) | OT- C1 (V) | rms |
| (mag) | (mag) | |
| 93.736 | -0.307 | 0.009 |
| 94.762 | -0.331 | 0.005 |
| 96.768 | -0.303 | 0.008 |
| 97.760 | -0.292 | 0.006 |
| 100.709 | -0.304 | 0.011 |
| 102.765 | -0.299 | 0.007 |
| 103.755 | -0.326 | 0.010 |
| 104.760 | -0.312 | 0.009 |
| 105.783 | -0.307 | 0.011 |
| 106.771 | -0.321 | 0.007 |
| 107.763 | -0.334 | 0.011 |
| 108.767 | -0.328 | 0.005 |
| 109.715 | -0.348 | 0.002 |
| 123.687 | -0.280 | 0.008 |
| 132.719 | -0.262 | 0.010 |
| 134.705 | -0.303 | 0.010 |
| 136.732 | -0.253 | 0.007 |
| 137.710 | -0.255 | 0.007 |
| 139.702 | -0.212 | 0.009 |
| 146.696 | -0.313 | 0.008 |
| 147.697 | -0.337 | 0.009 |
| 148.682 | -0.354 | 0.008 |
| 149.691 | -0.366 | 0.006 |
| 150.719 | -0.354 | 0.017 |
| 153.676 | -0.369 | 0.007 |
| 156.669 | -0.359 | 0.008 |
| 157.675 | -0.404 | 0.011 |
| 158.660 | -0.397 | 0.009 |
| 159.668 | -0.395 | 0.009 |
| 162.665 | -0.360 | 0.008 |
| 167.663 | -0.339 | 0.011 |
| 172.675 | -0.340 | 0.007 |
| 173.651 | -0.337 | 0.011 |
| 175.660 | -0.337 | 0.010 |
| 176.633 | -0.289 | 0.008 |
| 186.627 | -0.272 | 0.012 |
| 187.607 | -0.288 | 0.006 |
| 195.605 | -0.227 | 0.010 |
|
|
While photometry was carried out in Mexico at LLO (38 nights, from JD 2450093 to 2450195), the GEOS continued observing from Europe, twelve observers providing 290 magnitude estimates obtained during more than 250 days (JD 2449971 to 2450226, i.e. from september 11th 1995 to may 22nd 1996). As it is usually done with such observations, the original data for each observer has been corrected to a common zero, in order to account for the differences in chromatic sensitivity of each observer, related to the spectral type differences between comparison stars. The GEOS experience is that with such a procedure, errors in one observers' data reaching 0.2 on the visual magnitude are quite rare. In the case of OT Gem, and averaging severals observers' data as we did, the internal precision on the resulting measurement is about 0.1 mag, partly due to the differences in spectral type of the close comparisons that one has to use with the Argelander's method. This precision is of course inferior to that obtained by photoelectric photometry, but the scatter of the data can be further reduced by averaging the available measurements over 5 days (with the exception of the data where significant "events" appear). This seemed to us a reasonable compromise between the data precision and time resolution. These averaged data are reported in Table 2 (click here).
JD  | mv | N | JD | mv | N |
JD | mv | N |
| 971.644 | 6.48 | 1 | 1065.500 | 6.29 | 2 | 1152.385 | 6.15 | 3 |
| 974.649 | 6.57 | 1 | 1071.685 | 6.37 | 1 | 1153.352 | 6.17 | 5 |
| 980.654 | 6.32 | 1 | 1077.500 | 6.30 | 10 | 1154.339 | 6.21 | 3 |
| 984.663 | 6.42 | 1 | 1080.488 | 6.23 | 3 | 1157.500 | 6.28 | 8 |
| 988.162 | 6.42 | 2 | 1085.702 | 6.32 | 1 | 1162.500 | 6.30 | 11 |
| 994.682 | 6.53 | 6 | 1090.497 | 6.21 | 3 | 1167.500 | 6.30 | 8 |
| 996.553 | 6.57 | 9 | 1092.798 | 6.20 | 2 | 1172.500 | 6.26 | 9 |
| 1007.682 | 6.32 | 1 | 1097.500 | 6.33 | 20 | 1175.365 | 6.35 | 2 |
| 1012.500 | 6.32 | 5 | 1102.500 | 6.28 | 14 | 1177.370 | 6.12 | 4 |
| 1015.694 | 6.32 | 1 | 1106.416 | 6.28 | 1 | 1178.985 | 6.34 | 3 |
| 1017.696 | 6.32 | 1 | 1112.500 | 6.18 | 9 | 1182.500 | 6.28 | 12 |
| 1027.500 | 6.31 | 5 | 1119.429 | 6.36 | 1 | 1187.500 | 6.32 | 11 |
| 1036.708 | 6.17 | 1 | 1122.500 | 6.32 | 20 | 1192.500 | 6.40 | 9 |
| 1039.536 | 6.27 | 6 | 1127.500 | 6.38 | 5 | 1197.500 | 6.34 | 3 |
| 1042.500 | 6.24 | 3 | 1132.500 | 6.39 | 8 | 1200.349 | 6.00 | 1 |
| 1046.591 | 6.26 | 4 | 1137.500 | 6.36 | 9 | 1207.500 | 6.29 | 4 |
| 1047.712 | 6.17 | 1 | 1141.438 | 6.36 | 1 | 1211.845 | 6.30 | 2 |
| 1049.474 | 6.04 | 2 | 1145.364 | 6.32 | 7 | 1218.330 | 6.42 | 1 |
| 1050.485 | 6.00 | 4 | 1147.272 | 6.32 | 1 | 1220.844 | 6.34 | 2 |
| 1052.713 | 6.32 | 1 | 1148.274 | 6.19 | 1 | 1223.363 | 6.22 | 1 |
| 1056.703 | 6.32 | 1 | 1149.310 | 6.23 | 5 | 1226.362 | 6.22 | 1 |
| 1064.480 | 6.24 | 2 | 1151.342 | 6.25 | 4 | |||
The resulting GEOS light curve (Fig. 2 (click here)) shows an internal dispersion lower than 0.05 magnitude, and it coincides quite well with the La Luz Observatory measurements, during the common observations of the winter (1996). This good correlation confirms that the GEOS data can be completely trusted, and particularly the observed light extrema.
Figure 2: 1995-96 GEOS visual observations of OT Gem. Crosses are
single observations while the solid circles are averages of nearby
multiple observations as described in the text
A mean behaviour can be clearly detected in this GEOS data and can be described as follows:
During the last 210 days of the GEOS observations, the
star had a mean magnitude of 
. However its
regular mean magnitude has been for many years around
(Dumont 1996), and it was this unexpected increase
which called our attention to the star.
Fluctuations of about 0.1 magnitude do appear around this mean brightness:
Two well-marked minima are distinguished, i.e. a "wide" minimum at
JD 
2450130, and a sharper one at 2450192, i.e. at a 60 days
distance, with a duration of 25 and 15-20 days respectively.
There are two clear maxima, at JD 2450050 and 2450152, and possibly a third one at JD 2450112 suggested by both sets of data but unfortunately
not fully covered by the photoelectric data, i.e. these maxima are separated by
60 and 40 day intervals. The first maximum reaches a magnitude of
6.0, while the second and third reach 6.2 and 6.15 respectively. The above
intervals between maxima are not in contradiction with the times suggested
by the analysis of the La Luz data in Sect. 2, specially considering the
rather short time span of this data set (102 days).
To help establishing the activity level of the star, rough estimates of the slopes of the light increases and decreases can be given.
The rising branches observed in the light curve have a mean slope of
mmag/day. The simultaneous LLO and GEOS observations
(JD 2450140-150) show
the same -11 mmag/day, another confirmation that one can rely on the
visual data.
The very unusual light increase at JD 2450048 has a slope of about -70 mmag/day and a steep decrease of about +150 mmag/day. This seems to be one of the most spectacular flare-like events in the star.
The light curve descending branches have a mean slope of 
mmag/day.