[GRASSLIST:5105] Re: How to reduce processing time for s.vol.rst

[Maciek Sieczka]

From: "Michaela Kinberger" <kinb at atlas.gis.univie.ac.at>

Hi

I'm not the best person to answer this but as nobody else is trying to then
let me try.

> i'm using GRASS 5.4.0 and i want to interpolate sites data with s.vol.rst.
> i use 26 points as sites and a dhm of tyrol (austria) with a resolution of
> 200m. the interpolation nearly takes an hour.

Resolution is important - the finer the slower. The spatial extent - the
bigger area the slower. Also the more input points are taken into
interpolation the slower it runs. So lower "npmin" will speed-up the
interpolation, but will worsen it's quality. Also higher "dmin" will speed 
it up, as less input points are taken for interpolation then, but the output 
grid is more generalized. High "dmin" can also be some remedy for 
segmenting, because it let's you to have your input points distributed more 
homogenously, which all the RST modules like very much. But you have to 
remember that high "dmin" simply tells the algorithm to skip more input data 
points, which results in a less reliable output in some sense. You decide, 
trial and error followed by validating the output with input is required to 
understand how it works. I'm not saying that I understand it.

A nice example and more explanation than in the manual you'll find in the:

Cebecauer, Hofierka, Súri, 2002, Processing digital terrain models by
regularized spline with tension: tuning interpolation parameters for
different input datasets. Proceedings of the Open source GIS - GRASS users
conference 2002 - Trento, Italy, 11-13 September 2002.

Downloadable somewhere, Google it. It helped me a lot.

Maciek