On Mon, 09 Jan 2012 10:34:52 +0000 Giovanni Manghi <[email protected]> wrote:
> Hi, > > On Thu, 2012-01-05 at 02:22 -0800, Maximiliano wrote: > > Hi, > > does anybody know if there’s an application of Qgis to perform SITE > > FIDELITY TEST, like in arcview3? > > I’d like to know if I calculate the harmonic mean of all my fixes > > and the distance of each fixes from this “home range” centre… > > thanks a lot!! > > > don't think there is something ready for QGIS, but it would be a nice > addon for the Home Range or for the triangulation plugin. Any interest > in supporting its development? Hello Maximiliano, Giovanni, the literature on harmonic mean is quite old: the method dates back to 1980 and the Animal Movements documentation [0] warns about the known problems in home range estimates. I won't implement this specific method in HomeRange plugin, as I don't support the idea of implementing a method that has such known biases. Also, as HR plugin is a interface to adehabitat, I would ask Clément Calenge his opinion about adding this method to adehabitat directly. As an alternative to harmonic means, to individuate the core centres of the home range, I suggest to calculate homeranges with kernel method and select the 90% isopleth. (90% is a common value to identify core areas. However, better run an incremental core area analysis [1] to identify the correct limit of the core area for each animal). Then, calculate the center of each core area polygon and use it instead of the harmonic mean. This is my suggestion, of course comments and improvements are welcome! [0] http://alaska.usgs.gov/science/biology/spatial/gistools/animal_mvmt.php#harm_mean [1] http://wiki.faunalia.it/dokuwiki/doku.php/public/animove_howto#analyses Best regards, Anne -- http://gis.cri.fmach.it/ghisla/
signature.asc
Description: PGP signature
_______________________________________________ AniMov mailing list [email protected] http://lists.faunalia.it/cgi-bin/mailman/listinfo/animov
