modprecision-dp.f90

!> \file modprecision.f90

!>
!! Declare all precisions used throughout DALES
!>
!! We will start with the precision on a per-file basis 
!!  \author Victor Azizi, Escience Center
!!  \par Revision list

!! Notes
!! Real(4) does not (necessarily) means a 4-byte single precision floating point
!! this is up to the compiler to decide.
!! please use the iso_fortran_env real32, real64 etc. or use the
!! selected_real_kind intrinsic to select a suitable kind
module modprecision
use iso_fortran_env, only : real32, real64, real128, int16, int32, int64

integer, parameter :: longint = int64

!< double precision kind for lacz_gamma
INTEGER, PARAMETER :: dp = SELECTED_REAL_KIND(12, 60)

!< Parameter kinds, for rrtmg radiation scheme
integer, parameter :: kind_rb  = real64     !selected_real_kind(12) ! 8 byte real
integer, parameter :: kind_im  = int32      !selected_int_kind(6)   ! 4 byte integer
integer, parameter :: SHR_KIND_R4 = real32  !selected_real_kind( 6) ! 4 byte real
integer, parameter :: SHR_KIND_IN = kind(1) ! native integer

!! Module specific precisions
integer, parameter :: field_r = real64  ! Precision for the most common fields u, v, w
                                        ! And all other fields that do not have
                                        ! their own kind and need to interoperate

integer, parameter :: pois_r  = real64  ! Precision for the poisson solver,
                                        ! Precision for p and Fp fields
                                        ! Precision for all internal fields in
                                        ! the old fft and fftw module
end module