Did it differently: time_index: returns in integer to be used in reading fields in ncdf-files

DONE: save jul and secs in bdy_2d.nc files and implement real times

Did it differently: use time_index in get_2d_bdy_ncdf.F90

DONE: update getm.inp to reflect real tidal simulation period - start with ns_coarse.

DONE: read initial fields of salinity and temperature

DONE: stub for internal_pressure - equation_of_state

we need sigma_to_z and z_to_sigma

Ready to implement: temperature.F90, salinity.F90, advection_3d.F90, eqstate.F90, internal_pressure.F90. Should use lock_exchange as test-bed. Within
#ifdef BAROCLINIC

#endif
hardcoding is permitted - e.g. variables that later should be read from namelists.

20010608: kbk
src/ncdf: really needs cleaning - just added ncdf_lon_lat() should maybe make
a module called ncdf_domain.F90

20010827: hb -- todo items discussed in Schwarzenbek

Put most hardcoded variable settings into the getm.inp file (Karsten) 

Horizontal tracer and momentum diffusion (Hans) 

Solar radiation into Temperature.F90 (Karsten)

Non-reflective lateral boundary conditions (Hans for algorithm, 
Karsten for data input) 

Work out straight-forward graphical system for GETM (Karsten) 

Source code documentation (Hans, Karsten) 

Run tides for North Sea (Hans, Karsten) 

Set up output routine for lateral boundary conditions for 
one-way nesting (Karsten)

Interpolation of vertical profiles of T and S into domain, horizontally
homogeneous and horizontally inhomogeneous (Karsten)

Run simulations for Lake Constance, unstratified, stratified (Hans)  

Data input for rivers (Karsten) 





