Samstag, 26. November 2016

Timing of next big Dip

When will the next Dip appear at Boyajians Star?

In the last posts (d792, d1519), I described a model of the different large dips. This model is based on a beam that lifts matter from the star surface into an orbit. Due to the geometric situation, it should be possible to calculate the rotation speed of the beam. Actual the rotation speed is part of the parameter set in the shape calculation of the beam.
In this post, I will suggest a possible period.

The parameter rotation speed

The timing of the beam calculation is based on the equation

α(t) = ((BJD-2454833) – t) ω

BJD-2454833 is the time for all the Kepler data [1], based on the BJD time,
t0 describes the dip at day 792 and has the exact value 792.7216d
ω is the rotation speed and has to be fitted to the data.
If we know the exact value, we caught the rotation period of the beam and can calculate the reappearance of more dips.

In the first paper, I suggest ω has the value of 1.00E-02[1/d]. This correlates to a period of circulation of 628 days (2pi*100). The loosely guess of the 100 was done with no special precision.
In a review of the situation, I changed the value from 100 to 115,67 (Distance from d792 to 1519) and checked the effect to the shape of the calculated dips, with a special concentration on the lower part of the absorption. As shown in figure 1.

Figure 1: Dip 792 optimized right part
This fits very nicely in the in the area where the error was considered, shown as red error bars. A calculation when I change the period by one-day results already in a higher error in this region.
The other part of the plot is not that perfect, but other factors are not optimally implemented in the model, like the beam cross section.
To remember, in the right part under investigation, only the beam generates a shadow, the shape of the beam is not as influential as in the area of the deep dip.

Dips in the context

If we use the flux calculation for the whole Kepler period, we can see the repetition of the calculated flux as shown in figure 2:

Figure 2: Flux over the whole period (Right scale "Calculated Flux" lowered for good readability)

We see of course the tick at d792, which is the base of calculation. But we see also a dip at day 429, and day 1156, both are in some way an artifact, because the equation works with the absolute sin and contains in this sense a beam that would be on the other side of the star.
But there is an interesting detail, very near to this dips, we see in each case a small dip in the measure of Kepler. Let's have a look at the detail of dip d429 in figure 3:

Figure 3: Timing of d426 and the timing of the dip artifact.

It might be, that the technology behind star lifting generates some small beam at the opposite side of the star, this might be due to a magnetic field. But this can also be a pure coincidence. Due to the fact, that at day 1150 another similar dip is visible, should focus some resources to understand this coincidence.

The second appearance

At day 1519, we see the second appearance of our dip d792. The shape has changed but an interesting detail is still there, look into figure 4:

Figure 4: The dips around day 1519

In this plot, the shape of the d792 reappears due to the period of 726.78 days. Other elements add to that dip as discussed in the post 1519, but there is one amazing element, at day 1518, the flux recovers just to the level of the remaining effect of d792, but not more. This might be interpreted in a way, that this is, in fact, the separate structure of dip 792 appearing together with other elements of beams which reduce the flux at other times in the interval.

When will the next Dip appear

The most interesting question is, when will the next Dip appear?
I make the prediction as following in a table:

Time of dips in UTC time, Kepler starts at t = 0 (BJD-2454833): 2009-Jan-1 11:29:59, I used the converter from Ohio State University

  dip      BJD              UTC
   -   2454833      2009-Jan-1  11:29:59 (Start of Kepler time)
  792  2455625.722  2011-Mar-5   4:49:40
 1519  2456352.500  2013-Mar-1   0:35:59
 2246  2457079.278  2015-Feb-25 18: 4:19
 2973  2457806.057  2017-Feb-21 13:58: 4
 3700  2458532.835  2019-Feb-18  7:26:24
 4427  2459259.614  2021-Feb-14  3:20: 9 

(I am not completely sure about the time conversion, if someone is here an expert, he may check the Kepler start time) I have now revisited the timing conversion, using Eastman, Jason tool [2] and U.S. Naval Observatory Astronomical Applications Department the result is slightly different but very well within the error margin. Tuesday, A.D. 2017 Feb 21, 13:18:25.1 (JD 2457806.054457)

On February 21st should see a dip!
 I am very curious to see if the result matches the prediction. 

I am very curious to see the result.

[2] Eastman, JasonSiverd, RobertGaudi, B. Scott, 
Achieving Better Than 1 Minute Accuracy in the Heliocentric and Barycentric Julian Dates
DOI 10.1086/655938


  1. Lets wait for the 21th of Feb then! :o)

  2. The star is dipping right now, so the prediction was little bit off, but still :D !