I have a nephew who got his PhD in Astrophysics a few
years back. He says
FORTRAN is about all they use because they have vast libraries of FORTRAN
code that they can slap together to do big projects quickly. It has been
developed over quite a few years and is well debugged.
I was told that my Ph.D. thesis in particle physics is almost unique
because it doesn't contain (or make use of) one single line of Fortran.
There's plenty of Occam, VAX Pascal, and even a custom-designed
microcode, but no Fortran...
-tony