AAVSO Maxima and Minima of Long Period Variables, 1900-2008

Elizabeth O. Waagen, Janet A. Mattei, and Matthew R. Templeton

Click here for data access


Long-term observations of long period variable stars are essential to the determination of period, change in period, change in behavior resulting from mass loss and/or evolution, mode of pulsation (fundamental or overtone), internal structure, and phase at specific times for these stars.

Members of the American Association of Variable Star Observers (AAVSO) have been observing large-amplitude long period variable stars since the founding of the Association in 1911, and the AAVSO has been compiling the observations on these stars in the AAVSO International Database. The AAVSO published maxima and minima dates and corresponding magnitudes for 272 long period variables north of declination -25˚ from 1900 through 1920 in Harvard Annals, Volume 79, Part 2. The same type of information on 390 long period variables over the whole sky from 1921 to 1949 was published in the Studies of Long Period Variables. Maxima and Minima of Long Period Variables, 1949-1975 contained the maxima and minima dates and corresponding magnitudes on 384 stars from 1945 through 1975 and was a continuation of the information in the previous publications.

The current work is presented as a MySQL database (available here) and includes for 394 long period variables the maxima and minima dates, magnitudes, cycle numbers, weights, and other information from 1900 through 1975 essentially as published, and extends the work through 2017 with dates of maxima and minima as determined from the AAVSO International Database. Some typographical and other errors from the 1900-1975 data have been corrected and the AAVSO unique identifier (AUID) has been added for each star. Pertinent comments on the data for Y Cap and SZ And are given below. This database does not include all of the information included in the published data tables. 

For a given star, each record includes the AUID, star name, cycle number, Julian Date, magnitude, type of magnitude (P=positive, F=fainter-than, X=no magnitude and published, N=no magnitude and not published), whether the date is a maximum or minimum, whether the date is uncertain (1=uncertain), the weight (1-5, with 1 indicating very little or no data at the time of extremum and thus low confidence and 5 indicating a large amount of data and thus high confidence), and whether the extremum is primary (0) or secondary (1).

Future releases of this database will include more information, including magnitudes for the data from 1976 through 2017 and additional dates and magnitudes now available as a result of the incorporation into the AAVSO International Database of databases such as those from the Royal Astronomical Society of New Zealand, the British Astronomical Association, and the historical observations of Alexander Roberts. Magnitudes for data in earlier years will also be added where possible.

Observed maxima and minima dates 1975-2003 were determined by Janet Mattei and from 2004 to the present by Elizabeth Waagen using the AAVSO mean curves obtained by Leon Campbell and published in Studies of Long Period Variables.  Thus, the method of determining the observed maxima and minima dates has remained consistent since 1920. The dates for each cycle were obtained by superimposing the mean curve of the star on its individual cycles. The use of some mean curves, notably those for T UMi, R Aql, and Y Cap, was discontinued in recent years due to substantial period changes in T UMi and R Aql and to a major revision of Y Cap’s period.

Also for Y Cap, data for cycles 0-33 reflect an approximate period of 205 days; these data were published in Studies of Long Period Variables with these cycle numbers. In 1980 the period was doubled to approximately 410 days. To preserve the nature of the published data, in this set cycles 0-33 reflect a period of 205 days, and cycles 34-88 a period of 410 days. The next release of this database will include revised cycle numbers from cycle 0 to reflect the true period of 410 days. (If the data for Y Cap are needed before the revised data are released, please contact the AAVSO.)

When the first data for SZ And were published in Studies of Long Period Variables, only 6 cycles were included. Subsequently, extrema for an additional 2000 days before cycle 0 were measured by Margaret Mayall. In the current database, the cycle numbers have been revised to include the earlier 2000 days; the original cycle 0 is now cycle 6.


Many, many people are represented by this 117-year long database: the thousands of observers around the world who contributed the observations used to determine the extrema and who are acknowledged with profound gratitude; numerous AAVSO Headquarters assistants, data entry technicians, and programmers who have helped compile the data for publication or have written crucial software, including Katherine Hazen, Sara J. Beck, and Katherine L. Davis, who worked with the post-1975 data; and most importantly, the AAVSO Recorders and Directors Leon Campbell, Margaret W. Mayall, and Janet A. Mattei, who determined the great majority of the extrema information and ensured continuity of the database and the homogeneity of its contents.

Elizabeth O. Waagen, Senior Technical Assistant
Matthew R. Templeton, Science Director 2004-2015