This is an update on what I've done so far to look into the darkening of the lead glass by looking at LED signals.

1) To make the graphs more readable I have asigned one point for every run, indicating the average LED signal over that run.
2) The data now goes from day 78 to day 85, covering 107 runs total.
3) Many of the graphs appeared linear over this time scale, so I fit all the graphs to a pol1 function.

I've included pdf's containing all the graphs for each nstb at the bottom for those who are interested in looking them over.  The numbers after r and c are the row and column number respectively, with the last being the nstb.

The x-intercept in the graphs represenets day 78 run 1, with each integrer being a new day. Within a particular day each run (including those not neccessarily used in the data) is given a 0.01 + (small number) spacing. I'm not sure if there is a better way to space these, since beam dumps and such mean spacing between runs is not uniform.

To get a better feel for the data I've tried to characterize the amount of decay per day using the fit paramters. To do so I took -1*slope/yintercept*100, the reasoning being that the slope times 1 day should give the average change over one day and the y-intercept should be a good approximation of the starting value (so long as it is well behaved, see cuts below). I'm open to other suggestions if people think using the average over the entire span or some other process would be preferable.

I've plotted the % daily decay for the large cells and small cells below with the following cuts
  1) percent change within {-1,4} to remove outliers
  2) chi^2 < 10^5 to prevent ones where a few data points ruin the fit and make sure that the intercept is close to the starting value (see cutexamples.pdf)
  3) intercept > 10 to remove points with very low signal (see badexamples.pdf)

Observations:
  The large cells seem to have a decrease of 0.5-1% per day
  The small cells have a slightly larger decrease of 1-2% per day, with a wider distriubtion
  Peak at 0 may be radiation hardened cells

Futher Work
  Find dependence of decrease on position of cells