Carlos wrote: > I remember that some time ago, there was some discussion on > interpolation using faultlines. Any advances in that subject?
After many years of not publishing my GRASS interpolation with barriers module, I am today finally posting it publicly. I hope to publish an article on the module soon. It is now posted to the wiki addons page and housed in the GRASS addons SVN. http://grass.gdf-hannover.de/wiki/GRASS_AddOns#Vector_add-ons http://trac.osgeo.org/grass/browser/grass-addons/vector/v.surf.icw screenshot: http://grass.gdf-hannover.de/wiki/Image:Inlets_03_SurfSal_icw_big.png It is designed to work with no more than a few dozen data points and is very computationally expensive*. But results are worth the wait :) For "simple" inverse true-distance-around-barriers interpolation make your cost surface a 1/NULL map. For molasses barrier cases use a more dynamic cost surface input raster (e.g. harder for invasive ground critter to climb a sheer cliff than to walk around the back of the mountain and spread up the gentle way; or harder (but still possible) to diffuse upstream than downstream, ...). Unlike other half-baked (IMO badly wrong) IDW barrier interpolation methods the input points do not need to have line of sight to the raster cell to be considered. I have just updated it from GRASS 6.0 to GRASS 6.2 coding style, and had previously ported it from GRASS 5, so let me know if it breaks due to introduced typos etc. There is still much internal code cleanup I could do. [*] Adding a 'r.cost max_cost=' step could help speed it up, but the level for that would need to be dynamically calculated. Running in a RAM-disk mapset then copying out the result could help save a lot of r.cost disk thrashing and again help speed it up. GRASS> v.surf.icw --help Description: IDW interpolation, but distance is cost to get to any other site Usage: v.surf.icw [-rv] input=string output=string column=string cost_map=string [post_mask=string] [friction=value] [null_string=string] [--overwrite] [--verbose] [--quiet] Flags: -r Use (d^n)*log(d) instead of 1/(d^n) for radial basis function -v Verbose mode Parameters: input Name of existing vector points map containing seed data output Name of new raster file to be used for output column Column name in vector points map that contains data values cost_map Name of existing raster file containing cost information post_mask Name of existing raster file to be used as post-processing MASK friction Friction of distance, (the 'n' in 1/d^n) options: 1-6 default: 2 null_string Text string that represents "no data" in input data default: nan enjoy, Hamish _______________________________________________ grass-user mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/grass-user
