[vsnet-alert 25942] CSS_J231909.5+082832 extremely unusual light curve

[Denis Denisenko d.v.denisenko@gmail.com via vsnet-alert]

For some reason I have returned to the unusual variable in Pegasus
mentioned on this list 5 years ago, see [vsnet-alert 19859]. It is listed
as a EW-type eclipsing variable with P=0.215519 d in AAVSO VSX and as WDMS
binary in Rebassa-Mansergas et al., 2013. It is also related to the faint
ROSAT source 1RXS J231909.9+082832 = 2RXS J231909.8+082831.

Since January of this year the data from ATLAS Forced Photometry Server are
publicly available (Tonry et al., 2018). The phased light curve accumulated
in 2015-2021 is absolutely weird:
http://scan.sai.msu.ru/~denis/J2319+0828-ATLAS-phased_LC.gif

While it looks like a slightly tilted eclipsing binary in an orange filter,
there are three maxima and three minima in cyan. The type should be changed
from EW (E-double-U) to EVVV (E-triple-U)!

What's even more amazing, the eclipse was turning into this VVV shape
gradually over the last 15 years. Here is the animation of the phased plots
from CRTS data (Drake et al., 2009) separately for 2005-2008 and 2009-2013:
http://scan.sai.msu.ru/~denis/J2319+0828-anim-CSS-2005-2013.gif

One can see the main peak dropping, while the shoulder is rising on the
secondary minimum. Now the transition is complete.

AAVSO VSX entry:
https://www.aavso.org/vsx/index.php?view=detail.top&oid=397574

Additional names:
GALEX J231909.6+082832
SDSS J231909.65+082832.5
USNO-A2.0 0975-21245624
USNO-B1.0 0984-0656158
WISE J231909.65+082832.6

Gaia EDR3 distance is 226+/-2 pc and the absolute magnitude M_G=9.0.

Denis Denisenko