[SOLVED] How to model the zonal anisotropy

If you have problems using RGeostats package or if you simply need some help on an unlisted subject.

[SOLVED] How to model the zonal anisotropy

Postby Costu » Mon Aug 07, 2017 6:32 am

Hi,

I would like to know how to model the zonal anisotropy in RGeostats.

Thanks,

Kind regards,

Francky Fouedjio
Costu
 
Posts: 10
Joined: Sun Oct 14, 2012 8:37 pm

Re: How to model the zonal anisotropy

Postby Juan-Martin » Wed Jan 23, 2019 7:44 pm

Hi RGeostats pros/developers/maintainers!

I have the exact same question with fish densities measured in along hydroacoustic transects (and intertransects).

I found an answer with the package gstat here:
https://gis.stackexchange.com/questions ... gstat-in-r
(see post by Edzer Pebesma).

I would prefer continuing to use RGeostats. Could you please provide a simple 2-D example?

Thanks in advance!

Jean-Martin
Juan-Martin
 
Posts: 1
Joined: Wed Jan 23, 2019 5:57 pm

Re: How to model the zonal anisotropy

Postby Didier Renard » Mon Jan 28, 2019 10:36 pm

This file is to demonstrate the use of various types of anisotropies in RGeostats through 2-D examples.

We first create a 2-D grid 100 x100 (square lag=1) called *grid*

Code: Select all
grid = db.create(nx=c(100,100))


We define an isotropic model (model1) composed of a cubic basic structure with a range of 30 and a sill of 2. Using this model, we perform a non-conditional simulation that we visualize. Finally we calculate the experimental variogram along the two main axes of the grid (50 lags) and draw it together with the model

Code: Select all
model1 = model.create(vartype="Cubic",range=30,sill=2)
res1 = simtub(,grid,model1)
plot(res1,title="Isotropic")

Grid_isotropic.png
Grid_isotropic.png (30.22 KiB) Viewed 3433 times


Code: Select all
vario1 = vario.grid(res1,nlag=50)
plot(vario1)
plot(model1,vario=vario1,add=T)

Vario_Isotropic.png
Vario_Isotropic.png (19.2 KiB) Viewed 3433 times


Now we create a model (model2) with a cubic basic structure which follows a geometric anisotropy: the rotation of the main axis is 20 degrees (counted from east counterclockwise). The longest range is 30 and the shortest one is 10. Note that the directions for variogram calculations do not coincide with the anisotropy main directions: however the model fitting is reasonably correct.

Code: Select all
model2 = model.create(vartype="Cubic",range=c(30,10),sill=2,aniso.angles=20)
res2   = simtub(,grid,model2)
plot(res2,title="Geometric Anisotropy")

Grid_Geometric.png
Grid_Geometric.png (36.8 KiB) Viewed 3433 times


Code: Select all
vario2 = vario.grid(res2,nlag=50)
plot(vario2)
plot(model2,vario=vario2,add=T)

Vario_Geometric.png
Vario_Geometric.png (20.95 KiB) Viewed 3433 times


Now we create a model (model3) with a isotropic cubic component (range of 30 and sill of 1) and a second cubic basic structure which follows a zonal anisotropy with range of 10 and a sill of 1 in the direction 90). The range in the perpendicular direction is set to a huge value (the zonal isotropy is turned into a vary long geometric anisotropy).
Code: Select all
model3 = model.create(vartype="Cubic",range=30,sill=1,aniso.angles=20)
model3 = model.create(vartype="Cubic",range=c(10,10000),sill=1,aniso.angles=90, model=model3)
res3 = simtub(,grid,model3)
plot(res3,title="Zonal Anisotropy")

Zonal_Grid.png
Zonal_Grid.png (30.53 KiB) Viewed 3433 times


Code: Select all
vario3 = vario.grid(res3,nlag=50)
plot(vario3)
plot(model3,vario=vario3,add=T)

Zonal_Variogram.png
Zonal_Variogram.png (23.14 KiB) Viewed 3433 times


The last point is that the Automatic Fitting procedure is not able to fit a zonal model: nevertheless, it will possibly fit a model with a long geometric anisotropy.
Didier Renard
 
Posts: 337
Joined: Thu Sep 20, 2012 4:22 pm


Return to Other Troubleshooting

Who is online

Users browsing this forum: No registered users and 6 guests

cron