Added option to use double precision for calculating the deposition fields

src/concoutput.f90

@@ -62,6 +62,7 @@ subroutine concoutput(itime,outnum,gridtotalunc,wetgridtotalunc, &
   use outg_mod
   use par_mod
   use com_mod
+  use mean_mod
 
   implicit none
 
@@ -89,9 +90,10 @@ subroutine concoutput(itime,outnum,gridtotalunc,wetgridtotalunc, &
 !integer sparse_dump_i(numxgrid*numygrid*numzgrid)
 
 !real sparse_dump_u(numxgrid*numygrid*numzgrid)
-  real :: auxgrid(nclassunc),gridtotal,gridsigmatotal,gridtotalunc
-  real :: wetgridtotal,wetgridsigmatotal,wetgridtotalunc
-  real :: drygridtotal,drygridsigmatotal,drygridtotalunc
+  real(dep_prec) :: auxgrid(nclassunc)
+  real(sp) :: gridtotal,gridsigmatotal,gridtotalunc
+  real(dep_prec) :: wetgridtotal,wetgridsigmatotal,wetgridtotalunc
+  real(dep_prec) :: drygridtotal,drygridsigmatotal,drygridtotalunc
   real :: halfheight,dz,dz1,dz2,tot_mu(maxspec,maxpointspec_act)
   real,parameter :: smallnum = tiny(0.0) ! smallest number that can be handled
   real,parameter :: weightair=28.97

src/concoutput_mpi.f90

@@ -67,6 +67,7 @@ subroutine concoutput(itime,outnum,gridtotalunc,wetgridtotalunc, &
   use par_mod
   use com_mod
   use mpi_mod
+  use mean_mod
 
   implicit none
 
@@ -94,9 +95,10 @@ subroutine concoutput(itime,outnum,gridtotalunc,wetgridtotalunc, &
 !integer sparse_dump_i(numxgrid*numygrid*numzgrid)
 
 !real sparse_dump_u(numxgrid*numygrid*numzgrid)
-  real :: auxgrid(nclassunc),gridtotal,gridsigmatotal,gridtotalunc
-  real :: wetgridtotal,wetgridsigmatotal,wetgridtotalunc
-  real :: drygridtotal,drygridsigmatotal,drygridtotalunc
+  real(dep_prec) :: auxgrid(nclassunc)
+  real(sp) :: gridtotal,gridsigmatotal,gridtotalunc
+  real(dep_prec) :: wetgridtotal,wetgridsigmatotal,wetgridtotalunc
+  real(dep_prec) :: drygridtotal,drygridsigmatotal,drygridtotalunc
   real :: halfheight,dz,dz1,dz2,tot_mu(maxspec,maxpointspec_act)
   real,parameter :: smallnum = tiny(0.0) ! smallest number that can be handled
   real,parameter :: weightair=28.97

src/concoutput_nest.f90

@@ -60,6 +60,7 @@ subroutine concoutput_nest(itime,outnum)
   use outg_mod
   use par_mod
   use com_mod
+  use mean_mod
 
   implicit none
 
@@ -87,7 +88,7 @@ subroutine concoutput_nest(itime,outnum)
   !integer sparse_dump_i(numxgrid*numygrid*numzgrid)
 
   !real sparse_dump_u(numxgrid*numygrid*numzgrid)
-  real :: auxgrid(nclassunc)
+  real(dep_prec) :: auxgrid(nclassunc)
   real :: halfheight,dz,dz1,dz2,tot_mu(maxspec,maxpointspec_act)
   real,parameter :: smallnum = tiny(0.0) ! smallest number that can be handled
   real,parameter :: weightair=28.97

src/concoutput_nest_mpi.f90

@@ -64,6 +64,7 @@ subroutine concoutput_nest(itime,outnum)
   use par_mod
   use com_mod
   use mpi_mod
+  use mean_mod
 
   implicit none
 
@@ -91,7 +92,7 @@ subroutine concoutput_nest(itime,outnum)
   !integer sparse_dump_i(numxgrid*numygrid*numzgrid)
 
   !real sparse_dump_u(numxgrid*numygrid*numzgrid)
-  real :: auxgrid(nclassunc)
+  real(dep_prec) :: auxgrid(nclassunc)
   real :: halfheight,dz,dz1,dz2,tot_mu(maxspec,maxpointspec_act)
   real,parameter :: smallnum = tiny(0.0) ! smallest number that can be handled
   real,parameter :: weightair=28.97

src/concoutput_surf.f90

@@ -62,6 +62,7 @@ subroutine concoutput_surf(itime,outnum,gridtotalunc,wetgridtotalunc, &
   use outg_mod
   use par_mod
   use com_mod
+  use mean_mod
 
   implicit none
 
@@ -89,9 +90,10 @@ subroutine concoutput_surf(itime,outnum,gridtotalunc,wetgridtotalunc, &
   !integer sparse_dump_i(numxgrid*numygrid*numzgrid)
 
   !real sparse_dump_u(numxgrid*numygrid*numzgrid)
-  real :: auxgrid(nclassunc),gridtotal,gridsigmatotal,gridtotalunc
-  real :: wetgridtotal,wetgridsigmatotal,wetgridtotalunc
-  real :: drygridtotal,drygridsigmatotal,drygridtotalunc
+  real(dep_prec) :: auxgrid(nclassunc)
+  real(sp) :: gridtotal,gridsigmatotal,gridtotalunc
+  real(dep_prec) :: wetgridtotal,wetgridsigmatotal,wetgridtotalunc
+  real(dep_prec) :: drygridtotal,drygridsigmatotal,drygridtotalunc
   real :: halfheight,dz,dz1,dz2,tot_mu(maxspec,maxpointspec_act)
   real,parameter :: smallnum = tiny(0.0) ! smallest number that can be handled
   real,parameter :: weightair=28.97

src/concoutput_surf_mpi.f90

@@ -68,6 +68,7 @@ subroutine concoutput_surf(itime,outnum,gridtotalunc,wetgridtotalunc, &
   use par_mod
   use com_mod
   use mpi_mod
+  use mean_mod
 
   implicit none
 
@@ -95,9 +96,10 @@ subroutine concoutput_surf(itime,outnum,gridtotalunc,wetgridtotalunc, &
   !integer sparse_dump_i(numxgrid*numygrid*numzgrid)
 
   !real sparse_dump_u(numxgrid*numygrid*numzgrid)
-  real :: auxgrid(nclassunc),gridtotal,gridsigmatotal,gridtotalunc
-  real :: wetgridtotal,wetgridsigmatotal,wetgridtotalunc
-  real :: drygridtotal,drygridsigmatotal,drygridtotalunc
+  real(dep_prec) :: auxgrid(nclassunc)
+  real(sp) :: gridtotal,gridsigmatotal,gridtotalunc
+  real(dep_prec) :: wetgridtotal,wetgridsigmatotal,wetgridtotalunc
+  real(dep_prec) :: drygridtotal,drygridsigmatotal,drygridtotalunc
   real :: halfheight,dz,dz1,dz2,tot_mu(maxspec,maxpointspec_act)
   real,parameter :: smallnum = tiny(0.0) ! smallest number that can be handled
   real,parameter :: weightair=28.97

src/concoutput_surf_nest.f90

@@ -60,6 +60,7 @@ subroutine concoutput_surf_nest(itime,outnum)
   use outg_mod
   use par_mod
   use com_mod
+  use mean_mod
 
   implicit none
 
@@ -87,7 +88,7 @@ subroutine concoutput_surf_nest(itime,outnum)
   !integer sparse_dump_i(numxgrid*numygrid*numzgrid)
 
   !real sparse_dump_u(numxgrid*numygrid*numzgrid)
-  real :: auxgrid(nclassunc)
+  real(dep_prec) :: auxgrid(nclassunc)
   real :: halfheight,dz,dz1,dz2,tot_mu(maxspec,maxpointspec_act)
   real,parameter :: smallnum = tiny(0.0) ! smallest number that can be handled
   real,parameter :: weightair=28.97

src/concoutput_surf_nest_mpi.f90

@@ -63,6 +63,7 @@ subroutine concoutput_surf_nest(itime,outnum)
   use par_mod
   use com_mod
   use mpi_mod
+  use mean_mod
 
   implicit none
 
@@ -90,7 +91,7 @@ subroutine concoutput_surf_nest(itime,outnum)
   !integer sparse_dump_i(numxgrid*numygrid*numzgrid)
 
   !real sparse_dump_u(numxgrid*numygrid*numzgrid)
-  real :: auxgrid(nclassunc)
+  real(dep_prec) :: auxgrid(nclassunc)
   real :: halfheight,dz,dz1,dz2,tot_mu(maxspec,maxpointspec_act)
   real,parameter :: smallnum = tiny(0.0) ! smallest number that can be handled
   real,parameter :: weightair=28.97

src/makefile

@@ -81,7 +81,7 @@ interpol_mod.o          cmapf_mod.o \
 unc_mod.o               oh_mod.o \
 xmass_mod.o             flux_mod.o \
 point_mod.o             outg_mod.o \
-	 		random_mod.o
+mean_mod.o	 	random_mod.o
 
 MPI_MODOBJS = \
 mpi_mod.o
@@ -166,7 +166,7 @@ pbl_profile.o           readOHfield.o\
 juldate.o               \
 interpol_vdep.o         interpol_rain.o \
 hanna.o                 wetdepokernel.o \
-mean.o                  wetdepo.o \
+                  wetdepo.o \
 hanna_short.o           windalign.o \
 hanna1.o                	\
                         gridcheck_nests.o \
@@ -292,18 +292,18 @@ cmapf_mod.o: par_mod.o
 com_mod.o: par_mod.o
 conccalc.o: com_mod.o outg_mod.o par_mod.o unc_mod.o
 conccalc_mpi.o: com_mod.o mpi_mod.o outg_mod.o par_mod.o unc_mod.o
-concoutput.o: com_mod.o outg_mod.o par_mod.o point_mod.o unc_mod.o
+concoutput.o: com_mod.o outg_mod.o par_mod.o point_mod.o unc_mod.o mean_mod.o
 concoutput_mpi.o: com_mod.o mpi_mod.o outg_mod.o par_mod.o point_mod.o \
-	unc_mod.o
-concoutput_nest.o: com_mod.o outg_mod.o par_mod.o point_mod.o unc_mod.o
+	unc_mod.o mean_mod.o
+concoutput_nest.o: com_mod.o outg_mod.o par_mod.o point_mod.o unc_mod.o mean_mod.o
 concoutput_nest_mpi.o: com_mod.o mpi_mod.o outg_mod.o par_mod.o point_mod.o \
-	unc_mod.o
-concoutput_surf.o: com_mod.o outg_mod.o par_mod.o point_mod.o unc_mod.o
+	unc_mod.o mean_mod.o
+concoutput_surf.o: com_mod.o outg_mod.o par_mod.o point_mod.o unc_mod.o mean_mod.o
 concoutput_surf_mpi.o: com_mod.o mpi_mod.o outg_mod.o par_mod.o point_mod.o \
-	unc_mod.o
-concoutput_surf_nest.o: com_mod.o outg_mod.o par_mod.o point_mod.o unc_mod.o
+	unc_mod.o mean_mod.o
+concoutput_surf_nest.o: com_mod.o outg_mod.o par_mod.o point_mod.o unc_mod.o mean_mod.o
 concoutput_surf_nest_mpi.o: com_mod.o mpi_mod.o outg_mod.o par_mod.o \
-	point_mod.o unc_mod.o
+	point_mod.o unc_mod.o mean_mod.o
 conv_mod.o: par_mod.o
 convect43c.o: conv_mod.o par_mod.o
 convmix.o: com_mod.o conv_mod.o flux_mod.o par_mod.o
@@ -358,13 +358,15 @@ interpol_wind_nests.o: com_mod.o interpol_mod.o par_mod.o
 interpol_wind_short.o: com_mod.o interpol_mod.o par_mod.o
 interpol_wind_short_nests.o: com_mod.o interpol_mod.o par_mod.o
 juldate.o: par_mod.o
+mean_mod.o: par_mod.o
 mpi_mod.o: com_mod.o par_mod.o unc_mod.o
-netcdf_output_mod.o: com_mod.o outg_mod.o par_mod.o point_mod.o unc_mod.o
+netcdf_output_mod.o: com_mod.o outg_mod.o par_mod.o point_mod.o unc_mod.o mean_mod.o
 obukhov.o: par_mod.o
 obukhov_gfs.o: par_mod.o
 ohreaction.o: com_mod.o oh_mod.o par_mod.o
 openouttraj.o: com_mod.o par_mod.o point_mod.o
 openreceptors.o: com_mod.o par_mod.o
+outg_mod.o: par_mod.o
 outgrid_init.o: com_mod.o flux_mod.o oh_mod.o outg_mod.o par_mod.o unc_mod.o
 outgrid_init_nest.o: com_mod.o outg_mod.o par_mod.o unc_mod.o
 par_mod.o : $(wind_mod)
@@ -375,7 +377,7 @@ partoutput_mpi.o: com_mod.o mpi_mod.o par_mod.o
 partoutput_short.o: com_mod.o par_mod.o
 partoutput_short_mpi.o: com_mod.o mpi_mod.o par_mod.o
 pbl_profile.o: par_mod.o
-plumetraj.o: com_mod.o par_mod.o point_mod.o
+plumetraj.o: com_mod.o par_mod.o point_mod.o mean_mod.o
 psih.o: par_mod.o
 psim.o: par_mod.o
 raerod.o: par_mod.o

src/mean.f90

-!**********************************************************************
-! Copyright 1998,1999,2000,2001,2002,2005,2007,2008,2009,2010         *
-! Andreas Stohl, Petra Seibert, A. Frank, Gerhard Wotawa,             *
-! Caroline Forster, Sabine Eckhardt, John Burkhart, Harald Sodemann   *
-!                                                                     *
-! This file is part of FLEXPART.                                      *
-!                                                                     *
-! FLEXPART is free software: you can redistribute it and/or modify    *
-! it under the terms of the GNU General Public License as published by*
-! the Free Software Foundation, either version 3 of the License, or   *
-! (at your option) any later version.                                 *
-!                                                                     *
-! FLEXPART is distributed in the hope that it will be useful,         *
-! but WITHOUT ANY WARRANTY; without even the implied warranty of      *
-! MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the       *
-! GNU General Public License for more details.                        *
-!                                                                     *
-! You should have received a copy of the GNU General Public License   *
-! along with FLEXPART.  If not, see <http://www.gnu.org/licenses/>.   *
-!**********************************************************************
-
-subroutine mean(x,xm,xs,number)
-
-  !*****************************************************************************
-  !                                                                            *
-  !  This subroutine calculates mean and standard deviation of a given element.*
-  !                                                                            *
-  !      AUTHOR: Andreas Stohl, 25 January 1994                                *
-  !                                                                            *
-  !*****************************************************************************
-  !                                                                            *
-  ! Variables:                                                                 *
-  ! x(number)           field of input data                                    *
-  ! xm                  mean                                                   *
-  ! xs                  standard deviation                                     *
-  ! number              number of elements of field x                          *
-  !                                                                            *
-  ! Constants:                                                                 *
-  ! eps                 tiny number                                            *
-  !                                                                            *
-  !*****************************************************************************
-
-  implicit none
-
-  integer :: number,i
-  real :: x(number),xm,xs,xl,xq,xaux
-  real,parameter :: eps=1.0e-30
-
-  xl=0.
-  xq=0.
-  do i=1,number
-    xl=xl+x(i)
-    xq=xq+x(i)*x(i)
-  end do
-
-  xm=xl/real(number)
-
-  xaux=xq-xl*xl/real(number)
-
-  if (xaux.lt.eps) then
-    xs=0.
-  else
-    xs=sqrt(xaux/real(number-1))
-  endif
-
-end subroutine mean

src/mean_mod.f90

+!**********************************************************************
+! Copyright 1998,1999,2000,2001,2002,2005,2007,2008,2009,2010         *
+! Andreas Stohl, Petra Seibert, A. Frank, Gerhard Wotawa,             *
+! Caroline Forster, Sabine Eckhardt, John Burkhart, Harald Sodemann   *
+!                                                                     *
+! This file is part of FLEXPART.                                      *
+!                                                                     *
+! FLEXPART is free software: you can redistribute it and/or modify    *
+! it under the terms of the GNU General Public License as published by*
+! the Free Software Foundation, either version 3 of the License, or   *
+! (at your option) any later version.                                 *
+!                                                                     *
+! FLEXPART is distributed in the hope that it will be useful,         *
+! but WITHOUT ANY WARRANTY; without even the implied warranty of      *
+! MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the       *
+! GNU General Public License for more details.                        *
+!                                                                     *
+! You should have received a copy of the GNU General Public License   *
+! along with FLEXPART.  If not, see <http://www.gnu.org/licenses/>.   *
+!**********************************************************************
+
+module mean_mod
+  public
+
+! Interface to select single or double precision version of the 'mean'
+! function from type of input arguments ("function overloading")
+!********************************************************************
+  interface mean
+    module procedure mean_sp
+    module procedure mean_dp
+    module procedure mean_mixed_prec
+  end interface mean
+
+contains
+
+  subroutine mean_sp(x_sp,xm,xs,number)
+
+!*****************************************************************************
+!                                                                            *
+!  This subroutine calculates mean and standard deviation of a given element.*
+!                                                                            *
+!      AUTHOR: Andreas Stohl, 25 January 1994                                *
+!                                                                            *
+!      Single precision version ESO 2016                                     *
+!*****************************************************************************
+!                                                                            *
+! Variables:                                                                 *
+! x_sp(number)        field of input data                                    *
+! xm                  mean                                                   *
+! xs                  standard deviation                                     *
+! number              number of elements of field x_sp                       *
+!                                                                            *
+! Constants:                                                                 *
+! eps                 tiny number                                            *
+!                                                                            *
+!*****************************************************************************
+
+    use par_mod, only: sp
+
+    implicit none
+
+    ! integer :: number,i
+    ! real(sp) :: x_sp(number),xm,xs,xl,xq,xaux
+    ! real(sp),parameter :: eps=1.0e-30
+
+    real(sp), intent(in) :: x_sp(number)
+    real(sp), intent(out) ::xm,xs
+    integer,intent(in) :: number
+    real(sp) :: xl,xq,xaux
+    real(sp),parameter :: eps=1.0e-30
+    integer :: i
+
+    xl=0.
+    xq=0.
+    do i=1,number
+      xl=xl+x_sp(i)
+      xq=xq+x_sp(i)*x_sp(i)
+    end do
+
+    xm=xl/real(number,kind=sp)
+
+    xaux=xq-xl*xl/real(number,kind=sp)
+
+    if (xaux.lt.eps) then
+      xs=0.
+    else
+      xs=sqrt(xaux/real(number-1,kind=sp))
+    endif
+
+  end subroutine mean_sp
+
+  subroutine mean_dp(x_dp,xm,xs,number)
+
+!*****************************************************************************
+!                                                                            *
+!  This subroutine calculates mean and standard deviation of a given element.*
+!                                                                            *
+!      AUTHOR: Andreas Stohl, 25 January 1994                                *
+!                                                                            *
+!      Double precision version ESO 2016                                     *
+!*****************************************************************************
+!                                                                            *
+! Variables:                                                                 *
+! x_dp(number)        field of input data                                    *
+! xm                  mean                                                   *
+! xs                  standard deviation                                     *
+! number              number of elements of field x_dp                       *
+!                                                                            *
+! Constants:                                                                 *
+! eps                 tiny number                                            *
+!                                                                            *
+!*****************************************************************************
+
+    use par_mod, only: dp
+
+    implicit none
+
+    real(dp), intent(in) :: x_dp(number)
+    real(dp), intent(out) ::xm,xs
+    integer,intent(in) :: number
+    real(dp) :: xl,xq,xaux
+    real(dp),parameter :: eps=1.0e-30
+    integer :: i
+
+    xl=0._dp
+    xq=0._dp
+    do i=1,number
+      xl=xl+x_dp(i)
+      xq=xq+x_dp(i)*x_dp(i)
+    end do
+
+    xm=xl/real(number,kind=dp)
+
+    xaux=xq-xl*xl/real(number,kind=dp)
+
+    if (xaux.lt.eps) then
+      xs=0._dp
+    else
+      xs=sqrt(xaux/real(number-1,kind=dp))
+    endif
+
+  end subroutine mean_dp
+
+  subroutine mean_mixed_prec(x_dp,xm,xs,number)
+
+!*****************************************************************************
+!                                                                            *
+!  This subroutine calculates mean and standard deviation of a given element.*
+!                                                                            *
+!      AUTHOR: Andreas Stohl, 25 January 1994                                *
+!                                                                            *
+!      Mixed precision version ESO 2016 (dp input, sp output)                *
+!*****************************************************************************
+!                                                                            *
+! Variables:                                                                 *
+! x_dp(number)        field of input data                                    *
+! xm                  mean                                                   *
+! xs                  standard deviation                                     *
+! number              number of elements of field x_dp                       *
+!                                                                            *
+! Constants:                                                                 *
+! eps                 tiny number                                            *
+!                                                                            *
+!*****************************************************************************
+
+    use par_mod, only: sp,dp
+
+    implicit none
+
+    real(dp), intent(in) :: x_dp(number)
+    real(sp), intent(out) ::xm,xs
+    integer,intent(in) :: number
+    real(sp) :: xl,xq,xaux
+    real(sp),parameter :: eps=1.0e-30
+    integer :: i
+
+    xl=0._sp
+    xq=0._sp
+    do i=1,number
+      xl=xl+x_dp(i)
+      xq=xq+x_dp(i)*x_dp(i)
+    end do
+
+    xm=xl/real(number,kind=sp)
+
+    xaux=xq-xl*xl/real(number,kind=sp)
+
+    if (xaux.lt.eps) then
+      xs=0._sp
+    else
+      xs=sqrt(xaux/real(number-1,kind=sp))
+    endif
+
+  end subroutine mean_mixed_prec
+end module mean_mod

src/mpi_mod.f90

@@ -52,6 +52,7 @@ module mpi_mod
 !                         loops over particles. Will be all processes        *
 !                         unless a dedicated process runs getfields/readwind *
 ! lmp_sync                If .false., use asynchronous MPI                   *
+! mp_cp                   Real precision to use for deposition fields        *
 !                                                                            *
 !                                                                            *
 !                                                                            *
@@ -83,7 +84,7 @@ module mpi_mod
 
   integer :: mp_seed=0
   integer, parameter :: mp_sp=MPI_REAL4, mp_dp=MPI_REAL8
-  integer, parameter :: mp_pp=mp_sp
+  integer :: mp_cp
   integer, parameter :: id_root=0 ! master process
 
 ! MPI tags/requests for send/receive operation
@@ -162,14 +163,14 @@ contains
 !   mpi_mode    default 0, set to 2/3 if running MPI version
 !   mp_np       number of running processes, decided at run-time
 !***********************************************************************
-    use par_mod, only: maxpart, numwfmem
+    use par_mod, only: maxpart, numwfmem, dep_prec
     use com_mod, only: mpi_mode
 
     implicit none
 
     integer :: i,j,s,addmaxpart=0
 
-! each process gets an ID (mp_pid) in the range 0,..,mp_np-1
+! Each process gets an ID (mp_pid) in the range 0,..,mp_np-1
     call MPI_INIT(mp_ierr)
     if (mp_ierr /= 0) goto 100
     call MPI_COMM_RANK(MPI_COMM_WORLD, mp_pid, mp_ierr)
@@ -178,7 +179,7 @@ contains
     if (mp_ierr /= 0) goto 100
 
 
-! this variable is used to handle subroutines common to parallel/serial version
+! Variable mpi_mode is used to handle subroutines common to parallel/serial version
     if (lmp_sync) then
       mpi_mode=2 ! hold 2 windfields in memory
     else
@@ -189,6 +190,19 @@ contains
       lroot = .false.
     end if
 
+! Set MPI precision to use for transferring deposition fields
+!************************************************************
+    if (dep_prec==dp) then
+      mp_cp = MPI_REAL8
+      if (lroot) write(*,*) 'Using double precision for deposition fields'
+    else if (dep_prec==sp) then
+      mp_cp = MPI_REAL4
+      if (lroot) write(*,*) 'Using single precision for deposition fields'
+    else
+      write(*,*) 'ERROR: something went wrong setting MPI real precision'
+      stop
+    end if
+
 ! Check for sensible combination of parameters
 !*********************************************
 
@@ -1670,13 +1684,13 @@ contains
     end if
 
     if ((WETDEP).and.(ldirect.gt.0)) then
-      call MPI_Reduce(wetgridunc, wetgridunc0, grid_size2d, mp_sp, MPI_SUM, id_root, &
+      call MPI_Reduce(wetgridunc, wetgridunc0, grid_size2d, mp_cp, MPI_SUM, id_root, &
            & mp_comm_used, mp_ierr)
       if (mp_ierr /= 0) goto 600
     end if
 
     if ((DRYDEP).and.(ldirect.gt.0)) then
-      call MPI_Reduce(drygridunc, drygridunc0, grid_size2d, mp_sp, MPI_SUM, id_root, &
+      call MPI_Reduce(drygridunc, drygridunc0, grid_size2d, mp_cp, MPI_SUM, id_root, &
            & mp_comm_used, mp_ierr)
       if (mp_ierr /= 0) goto 600
     end if
@@ -1746,13 +1760,13 @@ contains
     end if
 
     if ((WETDEP).and.(ldirect.gt.0)) then
-      call MPI_Reduce(wetgriduncn, wetgriduncn0, grid_size2d, mp_sp, MPI_SUM, id_root, &
+      call MPI_Reduce(wetgriduncn, wetgriduncn0, grid_size2d, mp_cp, MPI_SUM, id_root, &
            & mp_comm_used, mp_ierr)
       if (mp_ierr /= 0) goto 600
     end if
 
     if ((DRYDEP).and.(ldirect.gt.0)) then
-      call MPI_Reduce(drygriduncn, drygriduncn0, grid_size2d, mp_sp, MPI_SUM, id_root, &
+      call MPI_Reduce(drygriduncn, drygriduncn0, grid_size2d, mp_cp, MPI_SUM, id_root, &
            & mp_comm_used, mp_ierr)
       if (mp_ierr /= 0) goto 600
     end if

src/netcdf_output_mod.f90

@@ -35,6 +35,10 @@
   !  - added option to not writeout releases information by changing 
   !    switch write_releases
   !  - additional updates for FLEXPART 9.x
+  ! 
+  ! ESO 2016
+  !  - Deposition fields can be calculated in double precision, see variable
+  !    'dep_prec' in par_mod
   !*****************************************************************************
 
 
@@ -47,7 +51,7 @@ module netcdf_output_mod
   use outg_mod,  only: outheight,oroout,densityoutgrid,factor3d,volume,&
                        wetgrid,wetgridsigma,drygrid,drygridsigma,grid,gridsigma,&
                        area,arean,volumen, orooutn
-  use par_mod,   only: dp, maxspec, maxreceptor, nclassunc,&
+  use par_mod,   only: dep_prec, sp, dp, maxspec, maxreceptor, nclassunc,&
                        unitoutrecept,unitoutreceptppt, nxmax,unittmp
   use com_mod,   only: path,length,ldirect,ibdate,ibtime,iedate,ietime, &
                        loutstep,loutaver,loutsample,outlon0,outlat0,&
@@ -70,6 +74,8 @@ module netcdf_output_mod
                        nested_output, ipout, surf_only, linit_cond, &
                        flexversion,mpi_mode
 
+  use mean_mod
+