A dedicated algorithm has been developed by Willie (blog 1) to select electrons (with some pi+,pi- contamination) which deposit significant energy in BSDM.

Among other characteristics Willie compared distance between high pT track and center of the BSMD cluster in both BSMD 'planes'. The left column of plots below is for data, the right for M-C of single electrons. In general the agreement is very good, on average residua are below 1cm in either direction, the rms is of ~1.5 cm (below the strip width, plots in blog2 at the bottom).

However the precision of this comparison allows us to draw some conclusions about relative TPC vs. BSMD alignment.

 1) the Z-direction real data (fig A):  there is a step for eta<-0.5 and eta>+0.5. One sees a jump in delZ of ~3mm at the place where BSMD strip changing its pitch. It looks like BSMD geometry problem.

2) Z-direction,  comparison of real data (figA) with M-C (fig B) shows for M-C we have additional systematic trend resulting with mismatch at eta~+/- 1 of +/- 8 mm. Furthermore there is a jump of ~3mm at the central membrane (eta=0). This smells like mismatched TPC corrections applied in M-C to smear and correct TPC points. The jump at eta=0 would lead to lower track reco efficiency for track crossing central membrane.

3) Phi-direction, real data (fig C). Looks like the East & West TPC are rotated by +/- 6 or 7 mm with respect to BSMD. From this study one can't tell which of the 2 BSMD or TPC is misaligned in geometry or if apparent East/West twist is a net result of applied TPC corrections.

4) Phi direction  comparison of real data (figC) with M-C (fig D) shows for M-C we have smaller East/West twist, suggesting large effect in fig C) comes from both : geometry misalignment of  +/- 4mm (as seen in fig D) and the hit correction effect of +3 and -1 mm.

 More plots from Willie's analysis of Run 9 data and M-C single electron are available at this (blog2)