[GRASS-user] 3D unstructured Grid

Vaclav Petras wenzeslaus at gmail.com
Wed Oct 10 19:48:01 PDT 2018


On Wed, Oct 10, 2018 at 8:34 PM Francois Chartier <fra.chartier at gmail.com>
wrote:

> Hi Vaclav,
>
> Question:
> 1-can the value remain constant until the next underlying data point?
>

v.vol.rst is for interpolation, but a constant value can be achieved with
r.to.rast3elev or r.to.rast3, but I think in your case you will likely need
some prereprocessing to get the right rasters from your irregular points.

https://grass.osgeo.org/grass74/manuals/r.to.rast3elev.html
https://grass.osgeo.org/grass74/manuals/r.to.rast3.html

I think there was some addon or script for working with soil types (which
is how your case sounds). Perhaps somebody knows more about it.


> 2-by "the output is a regular grid"  does that mean that the points
> (vector) can be unstructured, but in the end the raster or mesh is a
> regular grid?
>

Yes, 3D points (XYZ tuples) and (regular) 3D raster (grid).


>   if that is the case, how does the value in the raster determined?
>

v.vol.rst is interpolating the value using "regularized spline with
tension" aka RST, i.e. it is applicable to properties which are a real
number, but it is not suitable for categorical data.


> Is it an interpolated value between the points that fall with in the
> raster cell footprint?
>

The interpolation happens for the raster cells (or you can think of them as
grid locations in the middle of the raster cell). Usually, there would be
much more raster cells than vector points, but the RST interpolation works
the same way in all cases.



>
> thanks,
> Francois
>
> Le jeu. 4 oct. 2018 à 22:48, Vaclav Petras <wenzeslaus at gmail.com> a
> écrit :
>
>>
>>
>> On Tue, Oct 2, 2018 at 9:38 PM Francois Chartier <fra.chartier at gmail.com>
>> wrote:
>>
>>> For example, in the example on the webpage, there is a 3D raster for
>>> soil moisture; if the soil moisture is measured only for the top of a
>>> cylinder along your vertical hole, the next moisture measurement is taken
>>> at a lower depth, and the intervals (cylinder length) are not the same
>>> along that vertical axis?
>>>
>>
>>
>> Yes, all are (can be) different. The values are interpolated in between.
>> The input are arbitrary places points. The output is a regular grid. Note,
>> you may need to explore zscale parameter of v.vol.rst because of the how
>> your data looks like.
>>
>>
>>
>>>
>>> Le lun. 1 oct. 2018 à 22:01, Vaclav Petras <wenzeslaus at gmail.com> a
>>> écrit :
>>>
>>>> On Sat, Sep 29, 2018 at 4:45 PM Francois Chartier <
>>>> fra.chartier at gmail.com> wrote:
>>>>
>>>>> Can Grass GIS create a 3D unstructured grid based on a 3D point data
>>>>> set ?
>>>>>
>>>>
>>>> All data and visualizations at the following website were created in
>>>> GRASS GIS including a 3D regular grid (3D raster in GRASS GIS) interpolated
>>>> from a set of 3D points (vector which is 3D, i.e. with z, in GRASS GIS).
>>>> Points were soil properties measurement at different depths, 3D raster is
>>>> then continuous representation of a selected property.
>>>>
>>>> https://geospatial.ncsu.edu/geoforall/multidim-geovis.html
>>>>
>>>> Module used for interpolation was v.vol.rst:
>>>>
>>>> https://grass.osgeo.org/grass74/manuals/v.vol.rst.html
>>>>
>>>> Vaclav
>>>>
>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/grass-user/attachments/20181010/e3eafe4f/attachment-0001.html>


More information about the grass-user mailing list