Kriging
This window describes the Kriging step, which has the main goal of getting a
regular grid of values from the sample (points) data.
The Kriging module implemented in SPRING is based on the
"kt3d" GSLIB subroutine ( Deutsch and
Journel, 1992), which considers the simple Kriging, ordinary Kriging and
Kriging using several tendency models in two dimensions (2D) or three dimensions (3D).
Executing Kriging:
- Select in the "Control Panel"
an IL, from the Numeric model, which has samples of the POINTS TYPE;
- In the SPRING main menu press
Analysis Geoestatistics - Kriging...;
- In the "Kriging" window it is presented,
at the top, the variable corresponding to the Current Infolayer;
- Before executing the Kriging it is desirable to verify the model parameters. Thus
press the Model Verification...
button;
- Next, select the type of Kriging: Simple or Ordinary;
- If the Simple type was selected, type the Mean value in
the corresponding field. The Mean value can be computed using the
exploratory analysis.
- In Grid Definition, the Res.X and Res.Y fields
are filled with the default values giving an output grid with 200 columns and
200 rows (See the NOTE below);
- It is possible to limit the area over which the Kriging will be performed. For this,
press the Bounding Box...
button to establish the interest area. The Res.X and Res.Y fields are
automatically updated.
- The Minimum and Maximum Interpolation Parameters are related to the
Number of Points inside the Searching Ellipsoid. The default values are filled
(4 and 16 respectively). Next, the radius and orientation of the
Searching Ellipsoid are defined. The R.Min, R.Max and Angle
fields are initialized for an isotropic case, with the following default values:
R.min and R.max are equivalent to the Project bounding box and any Angle value, for instance, zero. Evidently if anisotropy is
present, these parameters should be adjusted and selected accordingly
( Deutsch e Journel, 1992);
- The Kriging output is a regular grid. Thus, it is necessary to define
a Category and an Infolayer;
- To finish the operation click on the Apply button.
[ Procedures ]
[ Concepts ]
[ Exploratory Analysis ]
[ Semivariogram Analysis ]
[ Semivariogram Modeling ]
[ Model Validation ]
[
Kriging
]
NOTE:
The Res.X and Res.Y fields are related to the output grid
X and Y resolution. These fields can be changed as far as the new values do not give
an output grid with number of columns or number of rows greater than 1000. Thus, the
Kriging module generate regular grids with, at most, 1000 columns by 1000 rows.
The Figure below summarizes the output grid parameters generated in the
Kriging module.

Figure - Output Grid.
See also:
SPRING - Spatial Analysis
The adjusting Theoretical Models for the
experimental variogram
How to Execute? - Model Validation
|