2009 Polarizations at RHIC. ---------------------------------------- Final polarization along with absolute stat. and syst. uncertainties in each physics fill are provided in Pol2009_500gev_blue.dat - Blue sqrt(s)=500 GeV Pol2009_500gev_yell.dat - Yellow sqrt(s)=500 GeV Pol2009_200gev_blue.dat - Blue sqrt(s)=200 GeV Pol2009_200gev_yell.dat - Yellow sqrt(s)=200 GeV The supporting analysis note with all details of the analysis pC_2009.pdf Plots for polarization vs fill Pol2009_500gev.gif (eps) - for sqrt(s)=500 GeV Pol2009_200gev.gif (eps) - for sqrt(s)=200 GeV ---------------------------------------- All uncertainties below are relative (Delta P/P or Delta A/A). Throughout the 2009 run the hydrogen jet target polarimeter (HJet) was used running with both beams incident on the target. It provides absolute polarization measurements at RHIC. The statistical uncertainty (deltaP) in Run9 was ~0.05 per fill in HJet measurements (so deltaP/P~15% and ~10% in sqrt(s)=500 and 200 GeV, correspondingly). The whole collected Run9 HJet statistics allowed to provide pC to HJet normalization with stat. uncertainties ~2.5% in sqrt(s)=500 GeV data and ~1% in in sqrt(s)=200 GeV data. The systematic uncertainty of HJet measurements includes uncertainty on molecular hydrogen contamination in polarized atomic hydrogen jet (2% for deltaP/P) and an upper limit for the effect of "other" background (2-3% for deltaP/P), estimated from the asymmetry in "non-signal" strips, and variation in the results due to different cuts applied. In 2009, we had upgraded PC polarimeter system, with two independent polarimeters in each ring (sharing the same DAQ system). For the measurements, we used target continuous scan mode, with horizontal target in one polarimeter and vertical target in the other one, which allowed us to measure both horizontal and vertical polarization profiles in each fill separately (though with limited stat. precision). During each fill pC provided 3-4 measurements with each of the polarimeters in a ring. Polarization in a fill was defined from the weighted average over measurements in a fill (weighting with luminosity was also checked to account for polarization and luminosity decay vs time in a fill). The strategy is to obtain the normalization for pC measurements, for each pC polarimeter separately, using absolute polarization measurements with HJet in the fills for which HJet measurement is available, and after that use the properly normalized pC measurements to define the polarization in each physics fill. The fill polarization from two poarimeters in a ring were combined in the final result. The normalization correction for pC polarimeters in Run9 was 0.84-0.88 for 250 GeV beams and 0.96-0.99 for 100 GeV beams, with initial normalization from Run4 100 GeV beam measurements. The shift in the normalization (analyzing power) for 250 GeV beams is consistent with theoretical calculations. In Run9 we experienced systematic problems in pC measurements related to high event rates due to smaller transverse beam size (in sqrt(s)=500 GeV) and on the average thicker targets used for the measurements, compared to previous RHIC runs. The average rate effect is absorbed by the HJet/pC normalization; fill by fill variation of the effect is estimated from the discrepancy between two polarimeter measurements in a ring (using different target thickness, and so having different rate effects). The 12% and 6% fill-by-fill uncertainties were assigned due to rate effect, for sqrt(s)=500 GeV and 200 GeV data correspondingly; 5% (10%) and 3% (3%) global uncertainties for blue (yellow) beam were assigned for sqrt(s)=500 GeV and 200 GeV data, correspondingly, to account for that not all physics fills had HJet absolute polarization measurements (about a half at sqrt(s)=500 GeV and about a quoter at sqrt(s)=200 GeV). Due to non-flat beam polarization profile we observe at RHIC, average beam polarization at a collision is higher than the one seen by the polarimeters, because in latter case polarization is wighted with beam lum. profile, and in former case - with two beam lum. profiles (~ lum. profile squared). The correction due to pol. profile depends on the squared ratio of width of the beam intensity profile and beam polarization profile, which we denote R. The average R was found to be 0.36 and 0.08, which led 15% and 4% corrections for the average polarization in a collision, at sqrt(s)=500 GeV and 200 GeV, correspondingly. The fill by fill pol. profile measurements were affected by high rates, so pC polarimeters were able to provide reliable pol. profile measurements only in several fills. We assigned fill by fill uncertainties for pol. measurements of 15% and 4%, at sqrt(s)=500 GeV and 200 GeV, correspondingly, which also reflects the observed variation of the profile from fill to fill. The global uncertainty for pol. was defined to be 5% and 2%, which reflects the uncertainty in the average over all fills pol. profile. In two fills, 10528 (blue and yellow) and 11020 (yell), pC didn't provide reliable data, so HJet measurements corrected for the average pol. profile have been used for the final results in these fills. ****************** Summarizing, the fill-by-fill syst. uncertainties (included in tables) are 15% and 12% (4% and 6%) at sqrt(s)=500 GeV (200 GeV), accounting for polarization profile uncertainty and rate related systematics. A few fills also had additional syst. uncertainty due to polarization time dependence in a fill. Below is a summary of global systematic uncertainties, in % (considered as correlated from fill to fill), discussed above. blue-500 yell-500 blue-200 yell-200 Jet normalization, stat: 2.5 2.7 1.0 1.0 Jet normalization, syst (molecular): 2.0 2.0 2.0 2.0 Jet normalization, syst (other): 3.0 3.0 2.0 2.0 Time dependence (rate): 5.0 10.0 3.0 3.0 Pol. profile: 5.0 5.0 2.0 2.0 Total: 8.3 12.1 4.7 4.7 Considering that "Jet normalization, syst" as well as "Time dependence" uncertainties are mostly correlated between blue and yellow, the final global uncertainties for a product of two beams, delta(P_B*P_Y)/(P_B*P_Y): sqrt(s)=500 GeV: 18.5%, which is sqrt{2.5^2+2.7^2+(2.0+2.0)^2+(3.0+3.0)^2+(5.0+10.0)^2+5.0^2+5.0^2} sqrt(s)=200 GeV: 8.8%, which is sqrt{1.0^2+1.0^2+(2.0+2.0)^2+(2.0+2.0)^2+(3.0+3.0)^2+2.0^2+2.0^2} Similarly, the uncertainty for the average between two beams (P_B+P_Y)/2: sqrt(s)=500 GeV: 9.2% sqrt(s)=200 GeV: 4.4%