I tried making a fit of the space charge using: [0]*pow(x,[1])+[2]  Yuri suggested this at the TPC meeting since a uniform background of photons in the experiment hall could cause this shape.  I ran this fit on data where we subtracted the 3 degrees from edges to try and decrease acceptance edge effects.

Following is a comparison for Run 13078014 with All Acceptances:

Original Power Fit (Without Constant)                                                                                                     Power Fit With Constant

We can see that there is a noticeable change in the radial chi_squared/ndf (1.053 without constant versus 0.687 with constant).  As for the Data minus Fit for the azimuthal dependence, we can see the bottom left has a yellow bump extend a little further when we include a constant, but overall it seems to have small effects scattered throughout.

I ran over all the runs, and found that:

Without the constant, the magnitude of N varied from    1.635 to 1.673 (0.038 range)
With the constant, the magnitude of N varied from         1.214 to 1.237 (0.023 range)

Thus the order of magnitude of N seems to vary less with luminosity, though the constant we introduced varies from 5.334 to 3.404 (decreasing with luminosity).  It should be noted that the error placed on the constant is usually just as large as the value of the constant itself.

Below is a plot showing the results for each run.

I am working towards making appropriate cuts to remove the main bumps in our data so that we can see if the order of N still appears to be dependent on luminosity.