Re: genapts

> Sometimes I wonder if the river/stream data is all the useful when it 
> doesn't match the surrounding terrain very well.  Once in a while I
> get tempted to calculate watershed myself from the SRTM terrain, but I
> suspect this is an overwhelming computational (and disk IO) problem to
> do on a global scale.



fyi,

Laura Toma @ Bowdion U. (again) might be someone to talk to about this:

TerraFlow: Computations on Massive Grids
"TerraFlow is 2 to 1000 times faster on large grids than the commonly
used software packages and scales up to grids of 1 billion elements."

about:
http://www.cs.duke.edu/geo*/terraflow/
http://www.cs.duke.edu/geo*/terraflow/terraflow-grass.html

man page + literature refs:
http://freegis.org/cgi-bin/viewcvs.cgi/*checkout*/grass51/raster/r.terraflow/description.html

it's written in C++ & should handle a single degree SRTM tile without
any problem; no idea how long it would take for all 10,000 or so SRTM
tiles..



Hamish

_______________________________________________
Terragear-devel mailing list
[email protected]
http://mail.flightgear.org/mailman/listinfo/terragear-devel
2f585eeea02e2c79d7b1d8c4963bae2d