COMMISSION 27 OF THE I. A. U. INFORMATION BULLETIN ON VARIABLE STARS Number 2785 Konkoly Observatory Budapest 2 September 1985 HU ISSN 0374 - 0676 PHOTOELECTRIC OBSERVATIONS OF TWO MAJOR FLARES ON UV CETI IN OCTOBER 1983 In 1983 an international program of RS CVn and flare star observations was organized by M. Rodono of the Catania Astrophysical Observatory, Italy (Rodono 1983). This program involved ground-based radio and optical observations carried out in conjunction with IUE satellite observations. One of the stars initially included in this program was UV Ceti (L726-8), and we obtained telescope time at Mount Stromlo Observatory, Australia, for the period 30 September - 7 October. After our observing session was over and we returned to our home cities, we learnt that UV Ceti had been dropped from the international program. Nevertheless, we think there is value in documenting our observations, given the detection of two major flares. Continuous monitoring of UV Ceti took place on two nights only, 30 September and 1 October, for a total of 4 hours 42 minutes, with a single channel photon-counting EMI Gencom Starlight-1 photometer (see Wolpert 1982) and the 76 cm "Reynolds Reflector". Observations were carried out in the B band of the Johnson UBV system. The photometer was interfaced through UART circuitry to an Hitachi MB-6890 Personal Computer. All machine programs to monitor, store and analyse the photometer output were written in BASIC. A chart output of the photon counter's digital readings provided a real time visual means of flare detection. Standard stars HD 9228, 10254 and 10824 were selected from Cousins and Stoy (1963) and Cousins, Lake and Stoy (1966), and comparison stars B and C on the finder charts provided by Rodono were used. Continuous monitoring of the program star was undertaken, using a 10 second integration time; the monitoring intervals are indicated in Figure 1 [FIGURE 1] Figure 1 UV Ceti monitoring periods, September-October, 1983. The two flares detected are indicated by spikes. Problems associated with flare detection and identification during monitoring runs have been discussed by Kunkel (1973), Moffett (1974) and Gurzadyan (1980), amongst others. The detection criterion we adopted for a flare was the existence of several successive readings above the quiescent level of the star, where the maximum magnitude level exceeded three standard deviations. On this basis, no flares were detected during the short monitoring run on 30 September, but on 1 October two major events were recorded. Details of these are provided in Table I (following the reporting system recommended by Andrews et al. 1969), and light curves are given in Table I Characteristics of Observed Flares Date Flare U.T. of Max. tb ta Duration Delta mB sigma Air 1983 Number h. m. s. (sec) (sec) (sec) (mag) Mass Oct.1 1 13 13 57 18 450 468 2.46 0.06 1.16 Oct.1 2 (14 30 00) -- (720) (>740) >2.00 0.09 1.06 Figures 2 and 3. The gaps in the light curves indicate breaks in the continuous monitoring to check the centering of the star in the 30 arc second aperture. In Table 1, tb and ta give the pre-maximum and post-maximum flare duration, respectively, and "Duration" = tb + ta. The second last column gives the standard deviation of random noise during the pre-flare quiescent state. It should be noted that because the commencement of Flare No. 2 was missed and its peak intensity may not have been recorded, some values in Table 1 are in parentheses, and the tb column is blank. Another reason for viewing some of the data in Table I with some degree of caution is the problem of accurately defining the flare start and end points, but particularly the latter (see Kunkel 1973; Oskanian 1968). [FIGURE 2] Figure 2 Light Curve of Flare No. 1 [FIGURE 3] Figure 3 Light Curve of Flare No. 2 Of all known flare stars, UV Ceti is the most active in the visual wavelength band (Bateson 1971), which is to be expected given that there is an inverse relationship between flare frequency ("f", in flares per hour) and absolute magnitude (Mv). For the U band, this is given by the following equation (after Gurzadyan 1980:14): log fU = -1.78 + 0.148Mv The absolute magnitude of UV Ceti is 14.78. Flare frequency, meanwhile, is dependent on wavelength of observation: fU > fB > fV We obtained a value of 0.43 for fB as a result of our observations (but it is important to note that in order to obtain a more statistically reliable value for fB, a larger number of flares would have to have been observed). Gurzadyan (1980:15) has also shown that flare frequency is not constant, but can exhibit considerable variation on an annual basis. For example, Kunkel (1973) observed that fU appeared to increase by of the order of 40% over the two years he monitored the star, while Lacy et al. (1978) noted a marked difference in flare frequency between 1976 and 1978. Bateson (1971) has suggested that UV Ceti exhibits a 6-year cycle of flare activity, not in terms of flare frequency per se, but rather the total flare energy output per year. He has proposed 1963 and 1969 as "peak" years, in which case our 1983 observations lie between two maxima (1981 and 1987). Bateson's interesting hypothesis has yet to be subjected to critical examination. The two flares on 1 October 1983 occurred within 1 hour and 16 minutes of one another, and in this regard our results mimic those of other researchers who report two or more major flares in quick succession, at a rate that differs quite markedly from the flare frequency (e.g., Bateson and Kohler 1968; Higgins et al. 1968; Ichimura et al. 1970; Osawa et al. 1968). This is a feature noted of other flare stars also, and prompted Jarrett and Eksteen (1969:133) to report that flares "...tended to occur within a few hours of one another." In fact, this is not always the case: Lacy et al. (1976) and Petersen et al. (1984) have clearly demonstrated that the "separation interval" between successive flares is random, and that flare occurrence follows a Poisson distribution. Both of our flares can be classed as "major flares" (i.e., Delta mB>1.0), which are also comparatively common for UV Ceti given that an inverse linear relationship has been established between Delta m and Mv. Gurzadyan (1980:19) defined a parameter he terms the "distribution function of flare amplitude", denoted by F(Delta m), which is given by the equation F(Delta m) = n(Delta m)/(summa n(Delta m)) where n(Delta m) is the number of flares with amplitudes between Delta m and Delta m + 1, and summa(n(Delta m)) is the total number of flares recorded (in any one wavelength band per unit interval of time). Using his own data, derived from 142 UV Ceti flares observed in the B band, Gurzadyan (1980:20) provides the values for F(Delta m) listed in Table II. Table II Distribution function of B band flares, UV Ceti Delta m 0 - 1 1 - 2 2 - 3 3 - 4 4 - 5 F(Delta m) 0.76 0.17 0.09 0.03 0.00 If these figures are indicative of the 1983 situation, we can count ourselves exceedingly lucky in detecting two major flares of such magnitude, given that only 7% of all UV Ceti flares could be expected to display Delta mB values of 2.0 or greater. Although our light curve for Flare No. 2 is not complete, it would appear that both of the outbursts we detected were discrete flares rather than the more complex "flare events" defined by Moffett (1979). Of the four different types of flare light curves identified by Oskanian (1968), both of our outbursts belong to Type 1, which is characterised by a very rapid rise time, and comparatively rapid decline. Oskanian found this to be the most common type in the case of UV Ceti flares. In Figure 2 there is a critical short section of light curve missing immediately before the outburst, but the basic pattern suggests that the flare was preceded by a "dip" or "negative flare" of about three minutes duration and Delta mB ~ 0.15. For many years the existence of such phenomena was in doubt, but their occurrence is now well established, particularly after the remarkable example provided by EQ Pegasi in 1980 (Giampapa et al. 1982). Preflare dips have been noted for UV Ceti by Cristaldi et al. (1980), who produced documentation associating these rare events with 11% of all flares they detected between 1968 and 1976 (where n = 174). Finally, the post-flare baseline fluctuations that have been reported by Bateson and Kohler (1968) were not apparent during our series of observations. We would like to thank the Mount Stromlo and Siding Spring Observatories for the use of the Reynolds Reflector, and Victoria College for contributing towards the cost of this project. W. ORCHISTON* Mount Stromlo and Siding Spring Observatories, Canberra, Australia. C. BEMBRICK British Astronomical Association (New South Wales Branch), Sydney, Australia. J. PARK Astronomical Society of Victoria (Inc.), Melbourne, Australia. B. POPPLETON Astronomical Society of Victoria (Inc.), Melbourne, Australia. *Present Address: Faculty of Applied Science, Victoria College (Rusden Campus), Melbourne, Australia. References: Andrews, A.D., Chugainov, P.F., Gershberg, R.E., and Oskanian, V.S., 1969, Inf.Bull. Var. Stars, No. 326. Bateson, F.M., 1971, Southern Stars 24: 23-33. [BIBCODE 1971SouSt..24...23B ] Bateson, F.M., and Kohlez, D., 1968, Southern Stars 22: 118-127. Cousins, A.W.J., Lake, R., and Stoy, R.H., 1966, Photoelectric Magnitudes and Colours of Southern Stars, II. London, H.M.S.O. (Royal Obs. Bull. 121). [BIBCODE 1966RGOB..121....1C ] Cousins, A.W.J., and Stoy, R.H., 1963, Photoelectric Magnitudes and Colours of Southern Stars. London, H.M.S.O. (Royal Obs. Bull. 64). [BIBCODE 1963pmcs.book.....C ] Cristaldi, S., Gershberg, R.E., and Rodono, M., 1980, Astron. Astrophys. 89: 123-125. [BIBCODE 1980A&A....89..123C ] Giampapa, M.S., Africano, J.L., Klimke, A., Parks, J., Quigley, R.J., Robinson, R.D., and Worden, S.P., 1982, Ap. J. 252: L39-L42. [BIBCODE 1982ApJ...252L..39G ] Gurzadyan, G.A., 1980, Flare Stars. Oxford, Pergamon Press. [BIBCODE 1980flst.book.....G ] Higgins, C.S., Solomon, L.H., and Bateson, F.M., 1968, Austr. J. Phys. 21: 725-734. [BIBCODE 1968AuJPh..21..725H ] Ichimura, K., Noguchi, T., and Watanabe, E., 1970, Tokyo Astron. Bull. 198: 2299-2305. [BIBCODE 1970TokAB.198.2299I ] Jarrett, A.H., and Eksteen, J.P., 1969, Mon. Not. Astr. Soc. S. Afr. 28: 131-133. [BIBCODE 1969MNSSA..28..131J ] Kunkel, W.E., 1973, Ap. J. Supp. 25: 1-36. [BIBCODE 1973ApJS...25....1K ] Lacy, C.H., Evans, D.S., Quigley, R.J., and Sandmann, W.H., 1978, Ap. J. Supp. 37: 313-320. [BIBCODE 1978ApJS...37..313L ] Lacy, C.H., Moffett, T.J., and Evans, D.S., 1976, Ap. J. Supp. 30: 85-96. [BIBCODE 1976ApJS...30...85L ] Moffett, T.J., 1974, Ap. J. Supp. 29: 1-42. [BIBCODE 1974ApJS...29....1M ] Osawa, K., Ichimura, K., Noguchi, T., and Watanabe, E. 1968, Tokyo Astron. Bull. 188: 2205-2211. [BIBCODE 1968TokAB.188.2205O ] Oskanian, V.S., 1968, In Detre, L. (ed.). Non-Periodic Phenomena in Variable Stars. Budapest, Academic Press. Pp. 131-136. [BIBCODE 1969CoKon..65..131O ] [CoKon No. 65] Petersen, H.R., Coleman, L.A., and Evans, D.S. 1989, Ap. J. Supp. 54: 375-386. [BIBCODE 1984ApJS...54..375P ] Rodono, M., 1983, Inf. Bull. Var. Stars, No. 2322. Wolpert, R.C.; 1982, In Hall, D.S. and Genet, R.M. (eds.). Photoelectric Photometry of Variable Stars. I.A.P.P.P. Pp. 9.54-9.57, 5.32-5.34.