Using portions of data from run 12178042 (FF, 280k triggers) and 12166028 (ZeroField = ZF, 300k triggers), I processed events using large inner TPC hit errors and looked at residuals for a proof-of-principal test on whether this data can be used for alignment. Cuts for track quality are included (at least 35 hits, no crossing sector boundaries nor the central membrane, and |η| < 1).

Residuals vs. local x (approximately radius) for FF (left) and ZF (right), east (red) and west (black) sectors combined:

I double checked the log files for both fields that the alignment picked up has beginTime 2010-12-20 00:00:01, which I believe is the Y2000 alignment.

Interesting that some misalignment-like features are seen, and that FF and ZF do not agree with each other fore this data. But I'm not sure what to make of that at this point. Yuri noted that ZF data will include notably more low pT particles which aren't swept out by the usual magnetic field.

-Gene

______________

Update [2011-07-01]:

More statistics, binned by sector (FF first [0.55M triggers], then ZF [1.67M triggers]), residuals [-0.05,0.09 cm] vs. local x:

(note that this uses my previously used method of resid = YFit-YHit, while in the plots I made above I accidentally used YHit-YFit, so there's a flip of sign)

RFF	ZF

The two field settings have many qualitative similarities but qualitative differences (open the images in a new window/tab of your browser if you want to see more detail). Again, the ZF data looks like it will be usable.

-Gene

____________

(NB: some of the image file names have "RFF" when they were actually "FF" as I at first mistakenly thought runs taken on day 178 were RFF)