1 / 13

ECAL Calibration with energy flow

ECAL Calibration with energy flow. Aurélien Martens. Purpose of this talk. Remember the previous talk on this subject (based on DC06 data) http://indico.cern.ch/materialDisplay.py?contribId=10&sessionId=3&materialId=slides&confId=51076 Today’s results : Sensitivity  cuts interdependence

deo
Download Presentation

ECAL Calibration with energy flow

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. ECAL Calibrationwith energyflow Aurélien Martens

  2. Purpose of this talk • Remember the previous talk on this subject (based on DC06 data) • http://indico.cern.ch/materialDisplay.py?contribId=10&sessionId=3&materialId=slides&confId=51076 • Today’s results : • Sensitivity  cuts interdependence • MC09 update • Dependence on the starting miscalibration • What could we have w/ Hcal or Prs ? • Results on the blind test (see Olivier’s talk) with a more appropriate miscalibration Aurélien Martens - ECAL Energy FLow Calibration

  3. Modus operandi • Miscalibrate (gaussian distributed µ=1,σ=0.1) • With the DeCaloCalib stuff located in Det/CaloDet • It miscalibrates the digits not the ADC. • To redo with a real miscalibration procedure (at the ADC level)… • Cut to remove noise (1.2 ADC A>10ADC i.e. Et> 25MeV for the digits) • Accumulate the flow over typ. 10Mevts (CaloEFAlg was written for this purpose) • Try to estimate the calibration coefficients (with CaloEFlowAnalysis) assuming : • some local smoothness • Symmetric energy flow (up/down & left/right) • One needs an intrinsically smooth energy flow • Et & occupancy (hits) are correlated : Et/Hits is a good candidate (almost flat one) • Et/evts is much less flat…but… Aurélien Martens - ECAL Energy FLow Calibration

  4. The sensitivity problem (1/3) • Comes from interaction between cuts and mean value • How to understand that ? • see previous talk • basic understanding : • ~ exponential spectrum (1rst order approximation in our case) : mean contains the exponential slope value (modified by the miscalibration) + Threshold energy term (depends on stat. fluctuations)  smearing of the miscalibration by the statistical fluctuations due to the cut Aurélien Martens - ECAL Energy FLow Calibration

  5. The sensitivity problem (2/3) • Non gaussian sensitivity loss (due to the fact that miscalibration is put at the digits level…) in ECAL with RMS ~ 7% if sum(Et)/Nhits is used • To do with a realistic miscalibration… • Limits too much the procedure • RMS of sensitivity loss is smaller with sum(Et)/Nevt (RMS ~3%) • This set the intrinsic limitation due to this effect • To be investigated in details • Conclusion : sum(Et)/Nevt energy flow is the best way to calibrate in ECAL Aurélien Martens - ECAL Energy FLow Calibration

  6. The sensitivity problem (3/3) • RMS of sensitivity loss is smaller with sum(Et)/Nevt (RMS ~3%) • 5% calibration is reached • Expect 3% by the method + 3% by the sensitivity + additional limitation due to the borders and non bi-linearity  3%sqrt(2)~4.2% • one gets ~5% • Conclusion : sum(Et)/Nevt energy flow is the best way to calibrate in ECAL Et/Evt To be confirmed and investigated in details with the miscalibration procedure at the level of ADC Aurélien Martens - ECAL Energy FLow Calibration

  7. summary of the methods • Local smoothness • Mean over the 8 neighbours. • Do not go across areas (do not use inner or outer cells for calibration coefficient estimate of the middle cells) • The limitation of the number of neighbours at the frontiers is compensated by an extrapolation across the area (in 8 directions) • Reach typically 5% calibration starting from 10% • Symmetric (left/right & up/down) assumption • Should be used as a first step (in order to avoid pb with intrinsic asymmetries) • Could reduce the miscalibration by a factor 2 in practice a bit less • Combination of methods : • 4% to 5% of calibration reached for ECAL, depending on the area Aurélien Martens - ECAL Energy FLow Calibration

  8. Results with different configurations Aurélien Martens - ECAL Energy FLow Calibration

  9. Intrinsic limitations • Sensitivity : • Already mentioned • Limits the calibration at ~3% • Depends on the normalization • To study in details w/ better misc. procedure • Local smoothness • Borders, mag. field • Depends on the normalization • Known Intrinsic asymmetries • hp Cross sections for K+, p and K-, pbar are very different (next slide) • Material before the calorimeters (Next to next slide) • And : • Tilt of the ECAL wrt x axis • Beam crossing angle Aurélien Martens - ECAL Energy FLow Calibration

  10. Hadronic cross sections • left right asymmetry from hadronic cross sections • With Down field, K+ goes to x>0 and then symmetric mean is bigger than the true value there in the ECAL • Flip of the asymmetry in HCAL : K+ are less energetic than K- after ECAL (1.1 int. length) • Very few % effect…in the ECAL at least ,more important in HCAL example : 3 GeV Kaons ~ outer 25 mb 17.5 mb Et/hit HCAL Et/hit ECAL needs confirmation with different MC09 mag. field config (DC06, mag up, inverted asym.) Aurélien Martens - ECAL Energy FLow Calibration

  11. Material effects • Et/evt : more sensible to material effects before ECAL : asymmetry in rad. Length is not negligible with this normalization • Observed asymmetry = Had. Cross sections + material before ECAL of the same order of magnitude • Other possible effects : • Tilt of ECAL wrt x axis • Beam crossing angle Et/evt ECAL Asym in rad. Length before ECAL Aurélien Martens - ECAL Energy FLow Calibration

  12. Hcal & Prs To be investigated in details with the good miscalibration procedure Aurélien Martens - ECAL Energy FLow Calibration

  13. Conclusions & prospects • Code is now commited in the CVS repository, if you want to play with it, feel free • New smoothing method could be introduced easily  (any idea ?) • 1D Histogramming is very fast  (and then, calibration computation) • 5% level reached for ECAL and HCAL • Usefulness for pi0 calibration ? • Usefulness for HCAL relative calibration cross check ? • HCAL possibly ameliorated by using 2 different polarities data samples and then a procedure normalized by events • Asymmetries are almost understood • Need confirmation for the field effect • Sensitivity to be continued • With a better miscalibration procedure (sorry for the noisy talk) • Different field configurations to investigate • Ask for Up field or No field generation ? • 10M minbias no-L0 is enough • Would be great to have a look to L0 data sample • 10Mevts to be stripped… • Check if thing are changed • Define the EFlow strategy for the first beam collisions Aurélien Martens - ECAL Energy FLow Calibration

More Related