How to understand this negative index array variable declaration?

Question

I am trying to understand (and reproduce) the following array declaration in a fortran code segment I am editing

real, dimension (-posint:) :: test_array
<