[GRASS-user] Error increasing in r.gwflow

[Nils G]

Hello all,
I am a student working on my master thesis right now. I just started using
GRASS and therefore probably making obvious mistakes.

With that said, I work on a 2d groundwater model. I have all of my raster
maps ready for the calculation, but the error is increasing with each
iteration and each linearization approach. This results in maps with
extremly high pheads (up to 10^23m) and budget errors. Sometimes the first
linerization approach ends with an error of 10^-7 while the second starts
with 900 and counts up to 10^40.

I observed that smaller dtimes reduce the error for the first few
linearization approaches. A simplification of the model (consistent kf and
effectice porosity values for the model, smoothed raster maps) also reduces
the error, but never leads to a sufficent result, because the dtime
manageable for the program without running into obscure budget errors is too
small. 

My size of my model is ~22x24km with a cell size of 100x100m. I use the
Kandawala coordinate system (EPSG 5234) in which one "coordinate step"
correlates with one meter. 

After I tried changing every possible parameter I have no idea of what is
left to try. Therefore I would be very grateful for every idea.



--
Sent from: http://osgeo-org.1560.x6.nabble.com/Grass-Users-f3884509.html