DESCRIPTION

r.pi.odc is part of the patch based fragmentation analysis package r.pi.* (Patch Index). It computes omnidirectional connectivity analysis between patches.

NOTES

Several output raster are generated with the defined output file name and a suffix of the information provided. All files named *.FP.* are providing information concerning the focus patch. All files named *.TP.* are providing informaton about the target patches.
TODO: ???

The user must specify the names of the raster map layers to be used for input and output, the keyval the ratio (area/odd or odd/area) and stats used (i.e., average).

Within r.pi.odc the following setting have to be set:

keyval setting:

The keyval operator determines which category value is taken for the Patch Index analysis.

Ratio setting:

The ratio operators determine what measure is applied.

Neighbourhood level:

The neighbor_level operator determines which neighbourhood level is used. 0 produces output for the focus patch itself, 1 assigns the connectivity information of the first omnidirectional neighbours to the focus patch, hence the connectivity of the surrouding fragments. This value can be increased for analysing the more distant neighbours.

Output:

Various output files are autmatically created with the pattern $output.* The
FP describes attributes of the fokus patch (area and area of the odd) TP describes attributes of the target patch (all neighbouring patches around the FP) - separated by the statsmethod (average, median, variance, stddev) ratio describes which ratio is taken for all TPs. The output raster files are named accordingly:
*.FP.area: size of the patch
*.FP.odd: size of the isolation area
*.FP.odd_area: ratio of size of patch and size of isolaton area
*.TP.no: amount of neighbouring patches
*.TP.area.avg: average size of all neighbouring patches
*.TP.odd.avg: average size of all isolation areas of neighbouring patches
*.TP.odd_area.avg: average ratio of isolation area to patch size
*.diagram: (if flag -d active) isolation areas and border are depicted

EXAMPLE

An example for the North Carolina sample dataset:

r.pi.odc input=landclass96 output=odc keyval=5 ratio=odd_area stats=average neighbor_level=0 -d

AUTHORS

Programming: Elshad Shirinov
Scientific concept: Dr. Martin Wegmann
Department of Remote Sensing
Remote Sensing and Biodiversity Unit
University of Wuerzburg, Germany

Last changed: $Date$