CFBG with temperature change in the middle

I created a model of a CFBG in optigrating. I want to add a temperature difference in the middle of the spectrum so that there is a dip in the middle of spectrum. You can see in the picture that it now has a temperature gradient from high temperature to low, when it should actually be the other way around. It looks like the scale on the x-axis is the opposite as the grating length, is that correct?