[vsnet-chat 5372] Re: V4641 Sgr: July 8-9 data from Monard

[Taichi Kato <tkato@kusastro.kyoto-u.ac.jp>]

Re: [vsnet-chat 5367] Re: V4641 Sgr: July 8-9 data from Monard

   Upon closer examination on the CCD data,

> 
> 20020708.916   119  (Maciej Reszelski)  !!!
> 

   this "flare" (?) was not recorded in Monard's data.  The light modulations
covering +/-minutes around this epoch only showed <0.1 mag variation.
The only viable possibility is that the flare was short enough to fall
on the short gap between CCD exposures.

   However, Toni's experience is harder to explain.  Statistically speaking,
the chance to miss all reported flares on CCD exposures is negligiblly
small.  More serious concern is that Toni apparently observed the star
every second (i.e. time resolution of 1s or better).  Using the similar
equipment and a CCD, we obtained a standard error of 0.075 mag at 5s
exposure.  Assuming that best human eyes have 1/10 quantum efficientcy,
and assuming the integration time of 1s, the expected error becomes
0.5 mag (1-sigma error) in ideal conditions (no scintillation, no cloud
interference etc.).  In some cases (about 10% of observations),
the object can look 2 sigma (1.0 mag) brighter than usual.  This seems to
explain the occurence of "flares" reported by Toni.

   1 min visual integration, as proposed by Maciej, sounds unrealistic.
The durations of flares (record with CCDs) is an order of several tens
of seconds, so they can be missed at longer time resolution.  Continious
watch (without blinking?) and recording (using a tape recorder or something
like that) an average of 10s impression would be an acceptable solution.
At this time-average, the expected photon noise becomes comparable to
the accuracy of human estimates.

Regards,
Taichi Kato

VSNET Home Page

Return to Daisaku Nogami

---

vsnet-adm@kusastro.kyoto-u.ac.jp