===============================================
 Amanzi Tests
===============================================

This directory contains a collection of verification and benchmark
tests for Amanzi.  Specifically, verification tests exercise Amanzi on
model problems for which an analytic solution is available.  As a
result, verification tests tend to have relatively simple geometries,
and uniform or piecewise constant parameters. Nevertheless, these problems
can significantly stress the nonlinear solvers and discretization methods.
For more complex and realistic scenarios, an analytic or high-quality 
reference solution may not be available.  In this cases we build confidence
in Amanzi by benchmarking it with other well known simulation codes.

The high-level directory structure is as follows:

  * verification   -  Collection of tests with analytic or reference solutions

    * flow         -  Flow tests 
      * saturated  -  Saturated flow tests (including pump tests)
      * richards   -  Variably saturated tests using Richards' model
 
    * transport    - transport (advection-dispersion-diffusion) tests 
      * saturated  - transport in a saturated flow field
      * ????       - transport in a variably saturated flow field

  * benchmarking   - Collection of benchmark tests (coming soon)

The naming conventions of subsequent directories is always up for debate
but the convention is to embed as a reasonable amount of information in
the path and directory name, e.g.,

 richards/transient/infiltration_dry_sand{1d|2d}

for directory names. Although the structured and unstructured input
files are very similar, it seems best to explictily call out which an
input file will invoke.  In most cases this distinction will be made
only a the file level, with files of both types in a single generically
named directory.


