Extrapolation Neutrino Flux measured at Near Detector to the Far Detector Near Detector Workshop, CERN, 30 July 2011 Paul Soler, Andrew Laing

2 Near Detector workshop, CERN, 30 July 2011 Outline o Near detector flux systematics at neutrino factory o Near detector location and spectra o Neutrino oscillation fits and sensitivities o Near to Far extrapolation method o Near to Far extrapolation simulations o Near to Far extrapolation results

3 Near Detector workshop, CERN, 30 July 2011 Near Detector Flux systematics o Importance of Near Detector for systematics o 2.5% error on flux makes big difference in CP coverage o Better to reduce error in flux below 1% if possible Winter, Tang

4 Near Detector workshop, CERN, 30 July 2011 Near Detector Location o Assume one ND per straight per ring (ie 4 detectors)  The idea is to have ND m from end of each straight to measure flux and possibly also measure divergence ~0.1/ 

5 Near Detector workshop, CERN, 30 July 2011 Spectra at Near Detector oNear Detector sees a line source (600 m long decay straight) oFar Detector sees a point source 1 km 100 m 2500 km ND FD Need to take into account these differences for flux measurement 100 m 1 km

6 Near Detector workshop, CERN, 30 July 2011 Fitting Far Detector Sensitivity * J. Burguet-Castell et al. Nucl. Phys., B608:301, 2001; Nucl.Phys., B646:301, 2002; Nucl. Phys., B725:306, o For Andrew’s thesis, carried out sensitivity plots by fitting spectra with NuTS framework developed by Valencia group* where M i is response matrix and N i,j is interaction matrix i= channel; j=baseline; k=background channel  Fitting for  13 and  CP simultaneously, minimise  2 : n e i,j =Data sim, N e i,j = predicted spectrum, e=energy bin; A j =rate factor (fiducial mass), x j =ratio cross sections;  A =0.05;  x =0.01.

7 Near Detector workshop, CERN, 30 July 2011 Far detector sensitivity  Set up grid of points in  13 and  CP and fit sensitivity contours : NH contours fitted with wrong mass hierarchy:  2 values much worse Normal mass hierarchy fitted with NH assumption This was all done using migration matrices from MIND analysis

8 Near Detector workshop, CERN, 30 July 2011 Far detector sensitivity  Sensitivity to  13 : Inverted mass hierarchyNormal mass hierarchy This was all done using migration matrices from MIND analysis

9 Near Detector workshop, CERN, 30 July 2011 Far detector sensitivity  Sensitivity to  CP : Inverted mass hierarchyNormal mass hierarchy This was all done using migration matrices from MIND analysis

10 Near Detector workshop, CERN, 30 July 2011 Far detector sensitivity o Sensitivity to the mass hierarchy: Inverted mass hierarchyNormal mass hierarchy This was all done using migration matrices from MIND analysis

11 Near Detector workshop, CERN, 30 July 2011 Flux extrapolation method o Extrapolation near-to-far at Neutrino Factory : –Our first idea was to directly use a matrix method similar to MINOS: –Where M FD =matrix of x-section plus response for numu at FD –M ND =matrix of x-section plus response for nue at ND –M nOsc =matrix of FD nue flux extrapolated from ND nue flux –N FD =number of numu events in FD –N ND =number of nue events in ND –P osc is the probability of oscillation and depends on  13 and  CP o However, there are problems with this direct method due to the finite resolution of response matrices: inverting three matrices means that the fit did not converge in many cases

12 Near Detector workshop, CERN, 30 July 2011 Flux extrapolation method o Extrapolation near-to-far at Neutrino Factory : –We now use indirect method, we extract P osc by fitting this formula: –Where M FD =matrix of x-section plus response for numu at FD –M ND =matrix of x-section plus response for nue at ND –M nOsc =matrix of FD nue flux extrapolated from ND nue flux –N FD =number of numu events in FD –N ND =number of nue events in ND –P osc is the probability of oscillation and depends on  13 and  CP  There is only one ND matrix that we need to invert and because the resolution on this matrix should be better than at the FD, then the fits converge for all values of  13 and  CP

13 Near Detector workshop, CERN, 30 July 2011 o Extrapolation near-to-far at Neutrino Factory : –Simulate a near detector nue response by assuming: Flux extrapolation simulation Prediction of flux from response matrices

14 Near Detector workshop, CERN, 30 July 2011 Flux extrapolation results o Extrapolation near-to-far at Neutrino Factory : –Using the FD spectrum formula: –Fit FD spectrum to predicted spectrum from ND: Fits assuming standard flux error 1% Fits using near-far projection method

15 Near Detector workshop, CERN, 30 July 2011 Flux extrapolation results o Extrapolation near-to-far at Neutrino Factory : –Using the FD spectrum formula: –Fit FD spectrum to predicted spectrum from ND: Fit improves at 3  level Comparison fitted  13 and  with true values

16 Near Detector workshop, CERN, 30 July 2011 Flux extrapolation results  Fitted vs true values of  13 and  CP : no observed biases Fitted  13 vs true value Fitted  CP vs true value

17 Near Detector workshop, CERN, 30 July 2011 Conclusions o Developed extrapolation method for near to far detector  While fluxes can be calculated accurately at a neutrino factory, fitting the spectrum from near to far detector improves performance especially at the 3  level o In my view, the near detector flux measurement and extrapolation to far detector is necessary to go to sub-1% level flux error