weighted least squares regression - height~diameter

Question

I am trying to fit linear regression models in a dataset of palms (log(height)~log(dbh)) (for 1200 observations). After performing shapirro and BPtest, both the assumptions