Could the spectrum changes be used to predict 1st photometric contact?

This is the question I asked myself last night as I took another spectrum clearly showing increasing signs of the eclipsing object. Well this same spectroscopic feature was studied by Lambert and Sawyer last eclipse.  You can read their paper here.

The relevant graph published there is reproduced below.

 

 It is a plot of the intensity of the KI 7699 line throughout the eclipse and shows how the changes in the line are seen well before 1^st contact and after 4^th contact, which are based on the visual brightness changes. (Indeed the line did not return to normal until over 200 days after the end of the eclipse photometrically!)

 

My results from last night show that the KI absorption line has now doubled in intensity relative to the pre eclipse level. Unfortunately Lambert and Sawyer did not make many measurements prior to 1^st contact (Unlike amateurs, professionals have to compete for time on busy telescopes) but we can plot the point at which the intensity doubled on their best fit line. (The red star shown on the magnified part of the graph below) 

 

It turns out that this point is about 10 days before the photmetric 1^st contact point (t1)   Of course this is a very rough and ready estimate but it does suggest that perhaps we should not be surprised if we do not detect any drop in visual brightness for a week or two yet.
 

 

Robin