Including git version output into tracking branch.

Tracking cloud particles are allocated, created, values interpolated, saved,
moved and reset. Information required for restart is saved in standard output
file, track logs in other directory.
No post-processing of output files is done yet.

working similar to svn version update

Since the refined octree is needed for the reinjection, the first
creation of the tracking cloud belongs there.

Save history on disk in separate directory
Update fields,write and move cloud all in one file

Working on the creation and update of tracking cloud particles. The type
sufaced will be deleted for memory reasons, but here a backup will be
kept.

Minor updates - vertical exaggeration, description in input file.

Defining tracking_cloud and tracking_surfaced types. The arrays are
defined, allocated and filled with dummy values. Deallocation is
included at the end of DOUAR.f90.
No calculation is done with the tracking cloud yet.

A shell unconstrained in temperature and velocity is used for a natural
evolution of temperatures above the indenter. A second shell with
linear-radial increasing temperatures to 1 set along the bottom ensures
a more reasonable temperature profile along the bottom.
The top surface nodes are fixed in velocity, so that no material from
the surface is subducted, which leads to too low temperatures.

Squashed commit of the following:

commit 9177a243c5439f5cf9fc7bd084efe8fdf121c823
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Tue Dec 9 13:30:00 2014 +0100

    Fixing calculation

    Changed the temperature calculation in relax-shell

commit 406363377a64dcae62cb71064f40c88cdd11538d
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Mon Dec 8 11:13:45 2014 +0100

    Fixed typo in xshell calculation

    minus sign switched

commit b29f80c745acd2ddd1aba72e84f081e8e3a6b0bb
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Fri Dec 5 10:16:46 2014 +0100

    Adding parameter read and updating input file

commit 9ef0d88dcb4e9433d0c84b1aa6c2534d40dfdffc
Author: Matthias Schmiddunser <mschmiddunser@argand.geowissenschaften.uni-tuebingen.de>
Date:   Thu Dec 4 17:55:48 2014 +0100

    Adding a temperature relaxation zone

    Second spherical shell (relaxshell), in which temperatures along bottom
    bi-linearly increase to mimic relaxation

commit f0bbcb0641160e29dae13a7f93c517ddec36e54e
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Thu Dec 4 14:12:12 2014 +0100

    Fix bottom temperature in 2 shellwidth distance

    Bottom nodes in twice the shellwidth distance are fixed to temperature 1
    in order to get a smooth tranisition

    Also adjusting the surface fixation.

commit 61cacda33de538ac546b8eee5929734b61b3c018
Merge: 8485230 a1b5f83
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Thu Dec 4 11:10:27 2014 +0100

    Fixing Surface nodes for initalisation

    Merge commit 'a1b5f83' into tinit2
    Previously discarded because did not converge, but might have been due to other
    problem. Trying anew.

commit 848523074de2440200558ccdb4500aa8fd32600e
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Thu Dec 4 10:26:11 2014 +0100

    Leavig bottom temperature entirely unconstrained

    Temperature gradient along bottom beyond the spherical shell is too
    steep, so trying not to constrain the bottom at all.

commit 71f872b8de52dce10655768750b94312eac4ed06
Merge: a2ab6e6 c8473b72
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Thu Dec 4 10:22:18 2014 +0100

    Merge branch 'douarw_rotsub' into tinit

    Conflicts:
    	src/define_bc_rot_sub_init.f90
    	src/read_input_file.f90
    	test/input_schmiddu.txt

commit a1b5f83b5b97e26101b24d317d44dab7957b2334
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Mon Dec 1 09:54:43 2014 +0100

    Fixing surface  in spinup

    For wider spherical shells, fix two topmost nodes

commit a2ab6e6b1d55b62c0e83285259b8064cb213f8cd
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Thu Nov 27 13:48:05 2014 +0100

    Added input parameter for T=0 elevation

    Using parameters%ztemp creates error, since this value is usually chosen
    to be above uppermost surface.

commit d95ee1204978053b305f577adf9d624a4afb1b97
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Tue Nov 25 14:44:32 2014 +0100

    New Temperature Init setup

    Rotating indenter, nonmoving bulk and a spherical shell without
    temperature or velocity constraints in between

commit 1436995156dac0a11afd94abc2c81389a2c13cc3
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Fri Nov 21 13:59:26 2014 +0100

    kfixt for part of bottom option

    bottom temperature gets fixed to 1 at the bottm after a belt of
    width fixbelt from the indenter outline

commit 59245b5146df8f97a30b8177f8a4b43b0b6e5b09
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Thu Nov 20 15:04:10 2014 +0100

    Extending options for fixed velocity in domain

    domainfix now allows choosing of which section (indenter, rest of
    domain, or both) will be fixed in velocity

commit 3b89c4d8d273269ee5e9eecff4c75288cb523a62
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Tue Nov 18 16:18:08 2014 +0100

    Removing basal temperature contraint

    kfixt = 0 for entire base.
    Also adding geotherm for x=1 and fixing temp.

Changing write_global_output so that the current output file is always kept
temporarily and deleted after the next output file was written. When
time = n*savstep+savoffset the files are kept permanently.

removing superflous variables and function parameters

Rotating indenter, nonmoving bulk and a spherical shell without
temperature or velocity constraints in between

Temperature is set at the left side and unset in the indenter footprint,
use to "pull" temperature from ther into indenter.

Also addition to elliptical dome surface: spherical shells of given
offset can be created.

Squashed commit of the following:

commit 01af6100baccbd389dca45e170c578cf642d73e7
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Tue Nov 11 11:42:49 2014 +0100

    Adding option not to set temperature at bottom

    Use for restart after sstemp

commit 3f0d2ab899ffba26030e56353b2c5a12beb550a4
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Fri Nov 7 16:18:42 2014 +0100

    Adding option domainfix and using old vex scaling

    domainfix will set velocity in entire domain 2 dz above indenter to 0
    returning to apply vex to osolve%z, since scaling the input parameters
    leads to different and incorrect shape of the indenter

commit 00a4fcd59cef94f7efc227f96297fa4d705bfce7
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Thu Nov 6 16:40:12 2014 +0100

    scaling input parameters with vex

    moving vex out of actual code and scaling input parameters at the start

commit 44171a3be6b1bc42eb4d80af3453bd6b3fa2e3cf
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Thu Nov 6 14:09:36 2014 +0100

    Changing Surface Calculation

    Trying a supposedly better spherical shell implementation

commit 1ee86ae001542ac5b37c7f41190b2d0cb913b8bd
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Thu Nov 6 11:12:51 2014 +0100

    Fixed geotherm calculation

commit 8ef7e3c3fe5e6631e4629cce003e4242c408b5bd
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Wed Nov 5 11:07:45 2014 +0100

    Added vex for geotherm calculation

commit 11e0f37f90b125014e254b3be9bab925b71590ef
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Wed Nov 5 10:44:29 2014 +0200

    Fixed typos, removed pi

    Pi defined by use constants, set comparison variables in if clauses to
    double precision

commit 15bb1a8e8fa9944d694f0f7fee2354ef8b3109d7
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Wed Nov 5 09:33:18 2014 +0100

    fixed minor typos

commit 9e9f1c7b9285f9497e4f1558179427520801eccf
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Tue Nov 4 10:30:49 2014 +0100

    Adding rot_sub_tinit for Temperature Initialisation

    New set of boundary conditions similar to rot_subduction, but with
    kfixt=1 at the left boundary and kfixt=0 within indenter footprint.
    Strictflow=True by default.

commit aa2fc2c179ca5b378995879c097bd92ce332b2c7
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Thu Oct 30 16:07:01 2014 +0100

    Added spherical shells description

commit 0c7c2532df3f0dd44c57e52a2c32a725b4cecc30
Author: Matthias Schmiddunser <matthias.schmiddunser@uni-tuebingen.de>
Date:   Thu Oct 30 16:02:47 2014 +0100

    Added spherical shells for indenter surface

    parallel surfaces of given distance to originally defined plane

deleting declared but unused variables, removing dummy arguments,
changing variables to double precision.
Not touching: uninitialized and more complicated warnings.

This sets velocity and temp for nodes inside the indenter by setting
osolve%kfix... Option to set either velocity (rotation with vin),
temperature (uniform)  or both.

The input template has been updated accordingly.

To eliminate warnings, declared but unneeded variables were removed from
define_bc and define_bc_rot_subduction

Makefile.inc.esd.argand for compiling on Todd Ehler's work group's
(ESD's) Argand (Linux machine).

Also tidied up input file template input_schmiddu.txt.

5th temperature option added:
Radial Gradient from (0,y0,0) to indenter edge
Updated input template

using strictflow=1 (bc_pamam12) will set the boundary conditions
at the sides and back to (vback,0,0) and prevent vertical movement.

module version updated to version after compilation

New boundary conditions:
 - A cylindrical or ellipsoidical indenter that is rotating downward set
   in horizontal flow field
 - Temperature conditions at the base can be set to a gradient or
   different for indenter (uniform or gradient)

New Surface:
 - Top of a cylinder with axis along y or of an ellipsoid rising above plane

Input files edited to accomodate new BCs and new surface

Example input file describing new options provided

Makefile for Taito updated (August 2014)

added variable calc_comp_velo, which is used to determined when compaction velocity should (or should not) be calculated

Fixed a few compaction problems: Deallocating certain osolve fields after compaction; Storing densityp=density in initialize_compaction; Several small bugs related to calculating the compaction velocity