CFBG with temperature change in the middle

Dear sir/madam,

I tried to make a simulation of a cfbg in optigrating. It doesn’t matter much what the specifications of this cfbg are. I want to apply a temperature difference in the middle of the grating length so that the spectrum shifts on the left side and there is a dip at the temperature difference, this all works fine when I apply a temperature difference with the user function at temperature. However, I am now using a temperature difference where the beginning of the fibre has a high temperature and the end has a low temperature (see picture). Actually, it should be a low temperature at the beginning and then to a high temperature. Then I get a peak instead of a dip. It looks like the grating length is the opposite as the length in the temperature function. Is that true?

Yours sincerely,

Quinten