COMMISSIONS 27 AND 42 OF THE IAU INFORMATION BULLETIN ON VARIABLE STARS Number 3763 Konkoly Observatory Budapest 2 September 1992 HU ISSN 0374 - 0676 ON HOUR-SCALE PHOTOMETRIC VARIATIONS OF TT ARIETIS The cataclysmic variable TT Ari was observed photometrically during an international campaign in 1988. Tremko et al. (1990) and Hudec et al. (1989) published the results of sudden decreases in the brightness. From the whole data set we chose 15 relatively long runs to test reality of a 4.68-hour secondary photometric period (Wenzel et al.,1986; Andronov et al.,1992). These runs were obtained mainly by the authors (4 in Sonneberg, 4 in Skalnate Pleso, 4 in Cracow and 1 in Piszkesteto), and 2 were published by Andronov et al. (1992). The original observations were averaged over bins of duration tipically less than 0.06d. Their number n_alpha_j in j^th bin in alpha^th run was used as "weight" for the least squares solution m_alpha_j = a_alpha + s_1 sin(2 pi f_1 t_alpha_j) + c_1 cos(2 pi f_1 t_alpha_j) + s_2 sin(2 pi f_2 t_alpha_j) (1) + c_2 cos(2 pi f_2 t_alpha_j) = = a_alpha - r_1 cos[2 pi f_1 (t_alpha_j - t_1)] - - r_2 cos[2 pi f_2 (t_alpha_j - t_2)] where m_alpha_j and t_alpha_j are averaged values of the magnitude and time in j_th bin, and a_alpha, s_1, c_1, s_2, c_2 are the parameters being determined for trial frequencies f_1 and f_2. The moments t_1 and t_2 correspond to a maximum brightness of waves with amplitudes r_1 and r_2. The function S(f_1,f_2) = sigma^2_O-C (f_1,f_2)/sigma^2_0 was used as a test function where sigma^2_O-C and sigma^2_0 are weighted variances of residuals from the fit (1) and from nightly mean. A total number of averaged observations is n = 478, a r.m.s. deviation from nightly means is sigma_0 = 0.048m. The shifts a_alpha for each night were computed as best fit parameters, thus taking into account their possible night-to-night changes. The first reason for this is that there may exist systematic differences among brightness values obtained at different telescopes in instrumental systems. Secondly, there may exist physical variations in the mean brightness from night to night. By using non-linear least squares procedure in the vicinity of the previously found values f_1 = 7.5346 cycles/day and f_2 = 5.123 c/d (Wenzel et al., 1986), we found that f_1 = 7.5211 c/d and f_2 = 5.1074 c/d. The corresponding best fit parameters are listed in Table 1. For a wide-range period search, we also computed a periodogram S(f_1,0) corresponding to one-frequency model with 17 unknown parameters (a_alpha_j, j = 1, ..., 15; s_1, s_2). It is shown in Fig. 1b. The most prominent feature in this periodogram occurs at f_1 = 7.5215 c/d. Several features at daily bias frequencies f_1 +/- 1, f_2 +/- 2 are seen as well, corresponding to a spectral window of observations. A linear ephemeris for the moments of maximum brightness obtained for our observations in a time interval J.P. 2447411-2447471 is as follows: Max. HJD = 2444739.898 + 0.132953E (2) +/- 2 +/- 13 This value of the photometric period differs significantly from P_1 = 0.132771d published by Wenzel et al. (1986) for observations obtained in 1986, as well as from P_1= 0.13277082d (Roessiger, 1988). However, it is close to P_1= 0.132957 published by Udalski (1988) for observations obtained in 1987-1988. Udalski (1988) also suggested that this photometric period may undergo real long-term changes. It may be noted that it differs by a few per cent from the period P_SP = 0.13755114d (Thorstensen et al., 1985) of spectral variations. Such discrepancy allows to classify TT Ari as an intermediate polar, despite the rotational period of the white dwarf is still unknown (Schwarzenberg-Czerny, 1990). To search for a secondary period in a wide frequency range we used two methods. At first, a one-frequency model was applied to the prewhitened observations m'_alpha_j=m_alpha_j-a_alpha+r_1 cos[2 pi f_1 (t_ alpha_j-t_1)] with r.m.s. deviation sigma_1 =0.041m. A corresponding test function S(f_2) = sigma^2_O-C/sigma^2_1 is shown in Fig. 1a. It shows a variety of dip features, the most prominent of which does not coincide in frequency with the previously obtained value of f_2 (Wenzel et al.,1986). Secondly, we computed a test function S(f_1,f_2) (Eq. (1), Fig. 1c), which exhibits similar behaviour. The relatively large noise at low frequencies corresponds to larger value of estimated parameters (l=19) as compared with l=3 (Fig. 1a). The relative depths of dips in Fig. 1a differ from those in Fig. 1c, but their frequencies coincide within the error estimates. Best fit parameters for four most prominent dips are listed in Table 1. Their frequencies exhibit well pronounced biases. However, the durations of our individual observational runs are not sufficient to cover the corresponding periods P_2 = 1/f_2, thus new observations longer than 5-6 hours are needed to solve the problem of a secondary photometric period. [FIGURE 1] Fig. 1 Periodograms for 478 mean points of TT Ari obtained in 1988. Vertical bars correspond to the spectroscopic frequency (7.27 c/d) and previously detected secondary oscillation (5.1 c/d). a) S(f_2) + 0.10 is a shifted periodogram for prewhitened observations; b) S(f_1,0) is a periodogram without taking into account the variations with a secondary period; c) S(f_1,f_2) is a periodogram for a two-frequency model (1), where f_1 was slightly corrected to obtain a minimum value of the test-function. Table 1 Parameter fits of TT Arietis f_1 f_2 r_1 r_2 t_1 t_2 S(f_1,f_2) 7.5217 3.2914 0.032 0.016 39.8975 39.8432 0.673 7.5216 3.4635 0.036 0.020 39.9007 39.7631 0.672 7.5217 4.2918 0.033 0.020 39.8975 39.9993 0.674 7.5211 4.4614 0.037 0.019 39.9007 39.9284 0.678 7.5211 5.1074 0.037 0.016 39.8984 39.9576 0.695 7.5215 - 0.035 - 39.8979 - 0.747 This research was partially supported by the Grant 2/61/92 of the Slovak Academy of Sciences. J. TREMKO ^1, I.L. ANDRONOV ^2, R. LUTHARDT ^3, G. PAJDOSZ ^4, L. PATKOS ^5, S. ROESSIGER ^3, S. ZOLA ^4,6 1 Astronomical Institute, Slovak Academy of Sciences 2 Odessa State University, Ukraine 3 Sonneberg Observatory, Germany 4 Mt. Suhora Observatory, Pedagogical University Cracow 5 Konkoly Observatory, Hungarian Academy of Sciences 6 Astronomical Observatory, Jagiellonian University Cracow References: Andronov I.L., Kolosov D.E., Movchan A.I., Rudenko A.N.: 1992, Commun. Spec. Astrophys. Obs. Russian Acad. Sci. 69, 79 Hudec R., Valnicek B., Tremko J., Wenzel W., Goetz W., Roessiger S., Kraicheva Z., Patkos L.: 1989, Comm. Konkoly Obs. 10, 244 [BIBCODE 1989CoKon..93..239H ] [CoKon No. 93] Roessiger S.: 1988, Mitt. Veranderl. Sterne 11,112 [BIBCODE 1988MitVS..11..112R ] Schwarzenberg-Czerny A.: 1990, Contrib. Astron. Inst. Skalnate Pleso 20, 9 [BIBCODE 1990CoSka..20....9S ] Thorstensen J.R., Smak J., Hessman F.V.: 1985, Publ. Astron. Soc. Pacif. 97, No 591, 437 [BIBCODE 1985PASP...97..437T ] Tremko J., Antov A., Genkov V., Kraicheva Z., Azimov A.A., Chernysev A.V., Zakirov M.M., Babaev M.B., Hudec R., Kolosov D.E., Kumsiashvili M.I., Luthardt R., Roessiger S., Movchan A.I., Rudenko A.N., Patkos L., Shpychka I.V.: 1990, Contrib. Astron. Inst. Skalnate Pleso 20, 69 [BIBCODE 1990CoSka..20...69T ] Udalski A.: 1988, Acta Astron. 38, 315 [BIBCODE 1988AcA....38..315U ] Wenzel W., Bojack W., Cristescu C., Dumitrescu A., Fuhrmann B., Goetz W., Grzelczyk H., Hacke G., Hudec R., Huth H., Kozhevnikov V.P., Kumsiashvili M.I., Mrkos A., Olah K., Oprescu G., Patkos L., Peresty R., Pfau W., Reimann H.-G., Richter G., Roessiger S., Shpychka I.V., Schult R., Stecklum B., Toth I., Tremko J., Valnicek B., Verdenet M.: 1986, Preprint Astron. Inst. Czechoslovak Acad. Sci. No. 38, 44