Lastly I find new fits for Qsca and Qabs. The original fix to scattering was a non-linear correction (corr_sca) which is a function of xm and m0, multiplies the csca to fit the model scattering to T-matrix. I first updated the parameters in the corr_sca in the original model. Although this produced a good fit the residuals were not normal (with p value for normality = xxx where the null hypothesis is that the residuals are not normal. )