Seon-Hee Seo: Results from RENO
RENO (see picture below) announced a new result using roughly doubled data : 402 live days of data for this new result. However, the new result is not about simply twice more data but also reduction in their systematic errors. The reduced systematic error was achieved by improving better estimation and reduction of backgrounds: Li/He, accidental, and fast neutron backgrounds.
RENO estimated Li/He background using “scaling method”. The idea of the”scaling method” is based on the fact that there should be no signal IBD events beyond 8 MeV.
If there is anything beyond 8 MeV then it should be Li/He background
because Li/HE spectrum extends to beyond 8 MeV.
There is flat fast neutron background as well as Li/He beyond 8 MeV.
Thus RENO counted events beyond 8 MeV after subtraction of fast neutron background and took them as total number of Li/He background events beyond 8 MeV. Using this number they scaled the Li/He background below 8 MeV
using already measured Li/He spectrum which was obtained by selecting events within 500 msec since last muon.
This method reduced their systematic error on Li/He background.
However, Their Li/He systematic error announced for this workshop is rather conservative and it is expected that the error would be reduced further down soon.
RENO also adopted a more effective flasher removal cut because they started seeing increased flasher PMT activity in their data since October 2012.
The new flasher removal cut is based on the topology of such events,
which have a PMT with a very big charge surrounded by neighboring PMTs
with much smaller amount of charge.
While for normal events the charge distribution for a cluster of PMTs was more homogeneous.
From these facts, they derived a variable called “anti-isolation”, R,
which is defined as mean charge of a cluster of PMTs divided by max charge in the cluster. For flasher events, this “anti-isolation”, R, value is smaller.
By combining this variable cut with q_max_q_tot cut, RENO removed flasher events very effectively, which also reduced accidental backgrounds.
RENO has also put a major effort to the energy scale calibration for this new result.
RENO obtained pe to MeV conversion function by fitting three source data points.
The fitting accuracy is about 0.1 %.
Seon-Hee, in her talk, did not describe their efforts improving energy scale calibration
but they plan to describe it in detail in a paper which will come out later.
The number of IBD events RENO has collected for this analysis is
about 280,000 events for Near detector and about 30,000 events for Far detector.
Their background level is 2.7 % (6.5 %) for Near (Far) detector.
Total background rate (per day) is 20.48 ± 2.13 for Near and 4.89 ± 0.60 for Far detectors.
RENO’s systematic error is ruled by Li/He systematic error.
Using this data, the observed deficit in Far detector is about 7%,
which remains stable beyond 7 MeV where there should be no oscillation.
The new result of sin^2(2theta_13) RENO obtained is 0.100 ± 0.010 (stat.) ± 0.015 (sys.) and this is 5.6 sigma level significance against no oscillation hypothesis.
The statistical error RENO has now is the same as Daya Bay.
The systematic error in this new preliminary result is rather conservative, and this is expected to be lowered down soon.
RENO Far detector was contaminated by very tiny fraction Cf252 source
because of loose O-ring of the source container. This occurred 13th of Oct. 2012 (this data was not used for the new result).
However, RENO has developed software cuts which can remove fake IBD events
caused by Cf252 source.
One of cuts they used for removing Cf252 source events helped reducing fast neutron background in general.
RENO now is in total control of the Cf contamination and the contaminated data
can even be used for the data analysis.
Seon-Hee advertised RENO-50 workshop which will be held in Seoul from June 13th to 14th, 2013. Registration if open now to anyone who is interested in future reactor neutrino physics: mass hierarchy, precise measurement of theta_12, SuperNova, Solar and Geo neutrinos etc..
The website of the RENO-50 workshop is linked below: