Merge pull request #4 from ubcgif/dev

theory and run the programs

content/examples.rst

@@ -1,9 +1,22 @@
 .. _examples:
 
-Example using five prisms
-=========================
-
-The example model is comprised of five anomalous rectangular prisms embedded in a uniform halfspace. There are three surface prisms simulating near-surface distortions, and two buried prisms simulating deeper targets, as shown in the figure below.
+Introduction
+============
+
+In this section, we present forward modelling and inversion examples for both DC and IP data
+types. The examples are freely available for download 3 MB and can also be found on the DCIP3D
+website (without the exe files). The example is synthetic and is a 10
+-m (0.1 S/m) block in a 1000
+
+-m (0.001 S/m) half-space with a pyramid-shaped topography (Figure 5). For added complexity,
+surface data on a 25-m grid is simulated along with borehole data so that the general data format
+is required. The intrinsic chargeability of the block is 0.15 (Figure 6). The electrode locations are
+denoted by the white dots in both Figures. After forward modelling, 5% Gaussian noise is added to
+each data set to create the \observed" data sets. The uncertainties assigned are 5% of each datum
+with a 0.0001
+oor (for both DC and IP). The last section shows the differences between the old
+code and version 5.0 with respect to the nodal-based finite difference scheme (shown in Figure 3).
+The mesh used for the example is 26  26  23 and the file is given by
 
 .. figure:: ../images/example/5prisms.png
 	:align: center

content/runPrograms.rst

@@ -56,6 +56,4 @@ Programs
         DCInv3D <runprog/dcinv>
         IPSEN3D <runprog/ipsen>
         IPInv3D <runprog/ipinv>
-        MAKE_WDAT <runprog/interfaceweights>
-
-
+        MAKE_WDAT <runprog/weights>

content/runprog/fwd.rst

@@ -56,7 +56,11 @@ vec
 
 **NOTE**: Formats of the files listed in this control file are explained :ref:`here <fileformats>`.
 
+**NOTE**: A sample input file can be obtained by executing the following line in the command prompt:
 
+.. code-block:: rst
+
+        dcipf3d -inp
 
 Output files
 ------------

content/runprog/ipsen.rst

@@ -4,7 +4,7 @@ IPSEN3D
 This program calculates the sensitivity matrix for the inversion of IP data. Command line usage:
 
 .. code-block:: rst
-        ipsen3d ipsen.inp [nThread]
+    ipsen3d ipsen.inp [nThread]
 
 where nThread is an optional integer argument to specify the number of OpenMP threads to use for the parallelization. If this value is missing, IPSEN3D will use the maximum number of threads based on the processor. The input file, ipsen.inp is described below.
 
@@ -51,6 +51,14 @@ tol
 vec
     Specifies how solution vectors are to be stored in the computer's memory. Use -1 to store all vectors in memory.
 
+**NOTE**: Formats of the files listed in this control file are explained :ref:`here <fileformats>`.
+
+**NOTE**: A sample input file can be obtained by executing the following line in the command prompt:
+
+.. code-block:: rst
+
+        ipsen3d -inp
+
 Output Files
 ------------
 

content/runprog/weights.rst

@@ -0,0 +1,67 @@
+.. _weights:
+
+MAKE_WDAT
+=========
+
+``MAKE_WDAT`` is a utility used to make a ``w.dat`` file that has smoothing in the x- and y-directions for the first few layers that underlie the topography surface. It suppresses the tendency of the algorithm to make highly variable structure in these top layers of the model.
+
+**NOTE**: We stress that the above weighting should be applied with care. A strong horizontal smoothing can often eliminate horizontal changes in the conductivity, even if the earth model truly had these. The inclusion, and details, of the surface weighting therefore involves subjective decision by the u er. Without weighting, the model at the surface can be quite rough (due to electrod effects). On the other hand, local geology is also sometimes quite rough and hence the tr e earth model should exhibit a large heterogeneity.
+The code is called by the command
+
+.. code-block:: rst
+
+		make_wdat mkw.inp
+
+This utility works with an arbitrary input control file name and does not have to be "weight.inp".
+
+Control parameters and input files
+----------------------------------
+
+The following shows the control file format:
+
+.. figure:: ../../images/interfaceweights.PNG
+	:align: center
+	:figwidth: 75%
+
+mesh
+	The 3D mesh file used in the DC or IP inversion
+topo
+	A topography file that is in the general format. Use null to apply to the top layers of the mesh.
+
+n
+	An integer specifying the number of layers to implement the weighting. The air layers will not be weighted and the remaining portion of the model will have a weight of 1.
+
+vals
+	Values of weights for the n layers specied on the previous line.
+
+name
+	Name of output file that writes the topography in the discrete format.
+
+
+**NOTE**: Formats of the files listed in this control file are explained :ref:`here <fileformats>`.
+
+**NOTE**: A sample input file can be obtained by executing the following line in the command prompt:
+
+.. code-block:: rst
+
+        make_wdat -inp
+
+Output files
+------------
+
+The program ``MAKE_WDAT`` outputs two files:
+
+w.dat
+	A 3D weighting file to use in the inversion
+
+name
+	The topography file in discrete format named from the last line in the input file. This file is not generated if the name line in the input file was null.
+
+Example files
+-------------
+
+Here is an example of the MAKE WDAT input file that places extra weights the first six layers beneath the topography. The remaining weights for the model are 1 and the air cells do not get weighted. Additionally, the program returns topo.idx that can be used with the general observations format.
+
+.. figure:: ../../images/interfaceweightsfile.PNG
+        :figwidth: 75%
+        :align: center