One suggestion to come out of last week's Spin-PWG meeting was to examine the asymmetry for different detector locations. For this excercise I use the traditional single-arm method of calculation. For further details, see my blog here. This method is used extensively by Steve Heppelmann with recent FMS measurements.

One potential issue with this is that the current calculation is not really optimised for isolating different areas of the detector. For the cross-ratio I divide the detector into upper and lower hemispheres. Of course, the EEMC is physically divided into 30º sectors, two of which straddle the horizontal and vertical axes. Thus, I have made a tweak to my code in this calculation. I reset the angular range to -11π/12 < φ_S < 13π/12, i.e. -165º < φ_S < 195º. Now, the 12 bins are 30º each but with bin centers which straddle the horizontal and vertical axes. To first order, then, this should yield the sector-dependence of the asymmetry.

Figure 1: Asymmetry in x_F-bins (x_F > 0)

Of course, statistics are quite limited. If anything the χ²/ν's are too good, so it is diffuclt to say with certainty there is some systematic difference in a particular φ location.

Figure 2: Asymmetry in p_T-bins (x_F > 0)

The story is similar in terms of p_T.

Figure 3: Asymmetry in x_F-bins (x_F < 0)

Again, now, with x_F < 0, it is difficult to decipher phi-dependent systematic trends.

Figure 4: Asymmetry in p_T-bins (x_F < 0)

A very similar story exists in p_T for x_F < 0.

Figure 5: Asymmetry for 0.03 < x_F 0.5 and 4 < p_T < 16 GeV/c

Integrating over the relevant range of x_F and p_T can enhance our statistics and give us perhaps a better chance to observe systematic behavior in the phi-bins. Figure 5 shows the two asymmetry figures for positive and negative x_F. The blue beam asymmetries show some fluctuation, though, the χ²/ν is not absurd. The yellow beam asymmetries are slightly higher than 1σ, and the χ²/ν is too good. A constant fit returns -0.00247357±0.00778415 with χ²/ν = 7.01796/11.