rsaga.ordinary.kriging {RSAGA}R Documentation

Local Ordinary Kriging

Description

Perform ordinary kriging using a local search neighborhood (local ordinary kriging). Also supports block kriging.

Usage

rsaga.ordinary.kriging(in.shapefile, out.grid, 
    out.variance.grid, field, 
    model = c("spherical", "exponential", "gaussian"),
    nugget = 0, sill = 10, range = 100,
    log.transform = FALSE, maxdist = 1000, blocksize, 
    nmin = 4, nmax = 20, target = rsaga.target(), ...)

Arguments

in.shapefile Input: point shapefile (default extension: .shp).
out.grid Output: filename for interpolated grid (SAGA grid file). Existing files will be overwritten!
out.variance.grid Output (optional): SAGA grid for kriging variances
field numeric(!): number (not name!) of attribute in the shapefile's attribute table to be interpolated; the first attribute is represented by a zero.
model character: variogram model to be used; defaults to "spherical".
nugget numeric (>=0): Nugget effect
sill numeric (>=0): Sill of the variogram
range numeric (>=0): Variogram range
log.transform logical: apply a log transformation to the observations? (default: FALSE).
maxdist numeric: maximum distance of nearest points to be used for kriging (search radius)
nmin numeric: Minimum number of points (within the local search neighborhood) required for interpolation.
nmax numeric: Maximum number of nearest points to be used for interpolation
blocksize numeric: block size for block kriging; block kriging is applied if this parameter is specified. If blocksize is missing (default), ordinary (point) kriging is used.
target list: parameters identifying the target area, e.g. the lower left corner and size of grid, or name of a reference grid; see rsaga.target.
... Optional arguments to be passed to rsaga.geoprocessor, including the env RSAGA geoprocessing environment.

Note

This function uses module 4 ("Ordinary Kriging") in SAGA library grid_gridding (users of the GUI of SAGA GIS should not be confused by the fact that the "Ordinary Kriging" module appears first in the GUI's module listing - it is in fact module 4).

The SAGA module support some other variogram models(?), but I am not quite sure what they are doing, so they (and the associated additional parameters) are currently not supported by this wrapper function. The module's usage page also mentions a FORMULA argument, but this seems to be a mistake.

Author(s)

Alexander Brenning (R interface), Olaf Conrad (SAGA module)

See Also

rsaga.inverse.distance, rsaga.target; see also krige in package gstat.

Examples

## Not run: 
# Krige attribute 0 from the points shapefile to
# a grid with the same extent and resolutionn as the
# (pre-existing) geology grid:
rsaga.ordinary.kriging("points", "dem", field = 0, maxdist = 1000,
    target = rsaga.target(target="target.grid",
    target.grid = "geology"))
# Specify a target grid manually (see rsaga.target):
rsaga.ordinary.kriging("points", "dem", field = 0, radius = 1000,
    target = rsaga.target("grid.system", 
        system.nx = 200, system.ny = 300,
        system.xy = c(604853,7465013), system.d = 50))
## End(Not run)

[Package RSAGA version 0.9-4 Index]