- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0) and
TIDAL FLATS = YES (default value) to improve mass conservation of water
- use of gradient conjugate + preconditioning 34 = 2*17 to solve PPE, more
efficient than GMRES for this example (very few iterations are needed)
- ACCURACY FOR DIFFUSION OF VELOCITIES = 1.E-15 as easily solved
+ PRECONDITIONING FOR DIFFUSION OF VELOCITIES = 34 to improve efficiency

- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0) and
TIDAL FLATS = YES (default value) to improve mass conservation of water

- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0)
to improve mass conservation of water + tracer
- use of gradient conjugate + preconditioning 34 = 2*17 to solve PPE, quite
more efficient than GMRES for this example
- switch to non-hydrostatic version (default) but no significant differences

- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0)
to improve mass conservation of water
- use of gradient conjugate to solve PPE (a little bit more efficient)

- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0)
to improve mass conservation of water (already chosen for NH example)
- use of default advection scheme for velocities (LIPS): better parallelisation
than with method of characteristics
- use of gradient conjugate to solve PPE (little bit more efficient)

Add comparison of the advection of tracers with PC1/PC2/LIPS

Same time step for non-hydrostatic and hydrostatic versions (now = 0.4 s as NH
before)

- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0)
to improve mass conservation of water+tracer
- preconditionings = 34 when possible (diffusion of velocities/tracers + PPE
as using conjugate gradients) for great efficiency to converge to the chosen
accuracy (down to 1.E-15 for the diffusion of tracers as easily reached for
this example)

- TIDAL FLATS = YES (default value) to improve mass conservation of water
- SOLVER FOR PPE = 1: conjugate gradient is a little bit more efficient than
CGSTAB (= 6) or GMRES (= 7) for this example

- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0) and
TIDAL FLATS = YES (default value) to improve mass conservation of water
for both examples gouttedo + thompson
- FREE SURFACE GRADIENT COMPATIBILITY : 0.8 for thompson example to decrease
spurious oscillations: 0.9 is not enough, 0.8 is OK, 0. is not so needed

- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0) and
TIDAL FLATS = YES (default value) to improve mass conservation of water
- FREE SURFACE GRADIENT COMPATIBILITY = 0.9 to decrease spurious oscillations
- SOLVER FOR PROPAGATION = 7 (better accuracy with quite same CPU time
rather than 2)
- switch to non-hydrostatic version (default)

- non-hydrostatic version (strange vertical velocity component W with
hydrostatic hypothesis)
- delete useless keywords with a direct solver (= 8)

- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0) and
TIDAL FLATS = YES (default value) to improve mass conservation of water
- switch to non-hydrostatic version (default)

- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0) and
TIDAL FLATS = YES (default value) to improve mass conservation of water
- SOLVER FOR PPE = 1 (conjugate gradient) + PRECONDITIONING FOR PPE = 34
(= 2*17) which combo can be used with finer accuracy with similar CPU time
as default SOLVER FOR PPE = 7 (GMRES)

- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0),
- IMPLICITATION OF VELOCITIES, ACCURACY FOR PPE, TIDAL FLATS set to their
default value,
- SCHEME FOR DIFFUSION OF VELOCITIES = 1 (with diffusion coefficients = 0)
rather than 0 to have no diffusion.
CPU time is a little bit shorter.
Mass conservation is improved (to machine precision, rather than 5 times
more of magnitude).

by using TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0
rather than 1)

- decrease SOLVER ACCURACY from 1.E-4 to 1.E-6
- TREATMENT OF NEGATIVE DEPTHS = 2 (new default value since release 9.0, rather
than 1)

for drogues files

punctual sources+control sections (T2D+T3D), sources by regions+fluxlines (T2D)

+ keyword RESULT FILE IN LONGITUDE-LATITUDE = YES (default choice)

to release drogues

in user_flot subroutines

[VnV] Update of axis labels for some figures where spherical coordinates are used

See merge request !231

A few additional descriptions + cosmetics

A little bit less diffusion (with the use of Predictor-Corrector scheme for
tracers rather than PSI)
+ use of "recent" default choices for implicitation coefficients
+ update of the documentation to show the improvements (and numerical keywords
to choose)
+ relax a few epsilons a little bit

Less diffusion (mainly due to refining the accuracy for PPE, then slightly
improved by the use of Predictor-Corrector scheme for tracers rather than PSI)
+ use of "recent" default choices for implicitation coefficients
+ update of the documentation to show the improvements (and numerical keywords
to choose)
+ refine epsilons (differences for temperature are now more acceptable, in
particular between scalar and parallel runs)

Missing or incorrect MNEMOS
Additional information for helps + grammar + cosmetics

Fix wrong time in time series files if using 2 different time sets in TELEMAC-2D

See merge request !230