18.7 fitting two and three together g10
Fiso3(E,→P,L)=−1/Kisodf(E∗)
Kisodf=P[0]M20E2−M2r+P[2]
Fiso3(E,→P=0,L)=1L[˜Fs3−˜Fs11/(2ωKs2)+˜Fs+˜Gs˜Fs]kp[12ωKs2]kp=δkp[(kcotδ)+|q∗2,k(1−H(→k))|]132πωkE∗2,k and kmcotδ=1a0m+r0m2k2m2P[3]=amP[4]=r0m The best fit:
χ2/d.o.f.=1.70598P[0]=74.0805±(11)P[1]=9.12899±(0.00017)P[2]=−706.844±(6e+02)P[3]=−0.159792±(0.0024)P[4]=−31.5321±(7.5) {C=(1300.581−0.449−0.119−0.170.5812.87e−08−0.572−0.481−0.414−0.449−0.5723.57e+050.1940.715−0.119−0.4810.1945.8e−060.666−0.17−0.4140.7150.66656.7)}
18.7.1 kcot 1 par
Kisodf=P[0]M20E2−M2r+P[2]
kmcotδ=1a0m The best fit:
χ2/d.o.f.=1.46817P[0]=81.5648±(12)P[1]=9.12952±(0.00059)P[2]=1455.65±(8.9e+02)P[3]=−0.155717±(0.0021) {C=(137−0.356−0.168−0.00778−0.3563.5e−070.374−0.255−0.1680.3747.96e+05−0.788−0.00778−0.255−0.7884.48e−06)}
18.7.2 kcot 1 kis0 3 par covariance
We minimise the correlated χ2 χ2=∑i,j(Epredicted−Elatt)iC−1i,j(Epredicted−Elatt)j
Kisodf=P[0]M20E2+M2r+P[2]
kmcotδ=1a0m The best fit:
χ2/d.o.f.=1.44987P[0]=96.629±(16)P[1]=9.12853±(0.0017)P[2]=1773.18±(9.8e+02)P[3]=−0.15631±(0.0027) {C=(1−0.4630.06740.0289−0.46310.105−0.1230.06740.1051−0.90.0289−0.123−0.91)}