We investigate whether flavor changing neutrino interactions (FCNIs) can be sufficiently large to provide a viable solution to the atmospheric neutrino problem. Effective operators induced by heavy boson exchange that allow for flavor changing neutrino sca

C.ConstrainingSU(2)Lbreakinge ects

fIfSU(2)Lbreakinge ectsarenegligiblethenGfνανβisequaltoGαβ.Comparingthe

experimentalbounds(2.10),(2.15),(2.16)and(2.20)with(1.3),we ndthatintheSU(2)Lsymmetriclimitthenewneutrinointeractionsthatweconsideredcannothaveasigni cantcontributiontotheANanomaly.

TheexcellentagreementbetweentheSMpredictionsandtheelectroweakprecisiondataimpliesthatSU(2)Lbreakinge ectscannotbelarge.Toshowthattheycannotsu ciently

′fweakentheupperboundson fνand νtobeconsistentwith(1.3),werecallfromEq.(2.5)

f22thatingeneraltheratioofthecouplings,Gfνανβ/Gαβ,isgivenbyratioM1/M2.HereM1

andM2arethemassesoftheparticlesbelongingtotheSU(2)Lmultipletthatmediate

ftheprocessesdescribedbyGfαβandGνανβ,respectively.IfM1=M2thismultipletwill

contributetotheobliqueparameters[28]S,Uand,mostimportantly,T.Thenwecanusea ttotheprecisiondatatodeterminethemaximallyallowedratio(M1/M2)2max.

WeusetheprogramGAPPbyJ.Erler[29]tocalculatetheSMpredictions.ForthelatestprecisiondatafromtheZ-pole[30],theW-bosonandtopquarkmasses[31,32],deepinelasticscattering[33],neutrino-electronscattering[34]andatomicparityviolation[35],weobtainthebest tvaluesoftheobliqueparameters:

S= 0.07±0.11,T= 0.10±0.14,U=0.11±0.15.(2.21)

WecalculatethecontributionstoS,T,andUfromthevariousscalarrepresentationsinTab.1andTab.2anddeterminethebest ttothedataateachvalueofthemasssplitting.Thebest tHiggsmassMHvarieswiththesplitting,andwelimittheHiggsmasstotherange95GeV<MH<1TeV.Constructingaχ2functionwedeterminetheupperboundonthemasssplittingbetweenthedi erentmembersofamultipletatagivencon dencelevel(CL).Theindividual90%CLboundson(M1/M2)2aregiveninthelastcolumnofTab.1andTab.2.(Notethatthelimiton(M1/M2)2isstrongerifthelightestmassisheavier.FromtheZ-widthmeasurement,thelightestmassmustbeheavierthanMZ/2,andtheboundspresentedinthetablesarederivedforthecasewherethelightermassis50GeV.)

Wedidnotcalculatetheboundsforthevectormultiplets.Sincevectorbosonsgiveingenerallargercontributions,weexpecttheboundsinthevectorcasestobeasgoodorbetterthanthecorrespondingboundsforthescalarmultiplets.Thus,forthevectormultipletsaratherconservativeupperboundis(M1/M2)2<7.

Hence,eventhemaximalpossibleSU(2)Lbreakinge ectscouldweakentheboundswe

′fderivedonlybyafactorofafewand fν, νcannotexceedthefew-percentlevel.Welearn

thatalsothe avorchangingneutrinoscatteringνµf→ντfinducedbyheavybosons,thataredoubletsortripletsofSU(2)L,cannotsigni cantlycontributetotheANanomaly.