MODFLOW USG - horizontal anisotropy

[Cptn. Nemo]

Hi everyone,

I'm trying to define a horizontal anisotropy in my model. At the moment, I want it to be very simple - the anisotropy is constant for all cells of the model grid. Kx/Ky ratio is 1:3 with angle of 90 deg (1.5708). I'm using LPF package. 

I defined CHANI as 0. Based on the documentation:

- the face angle array (ANGLEX) should be defined above HK arrays as one-dimensional array (NJAG elements). 

- HANI should be defined for every layer separately (NDSLAY elements for every model layer). It should be located under HK array

My LPF file looks like that:

# MODFLOW-USG Layer Property Flow (LPF) Package                                                     

 50  0.10000E+31 0 0     CONSTANTCV

 4 4 4 4                 <--- type of layer

 0 0 0 0                 <--- type of interblock transmissivity calcs

 -1. -1. -1. -1.         <--- CHANI

 0 0 0 0                 <--- VKA type 

 0 0 0 0                 <--- re-wetting parameter

 1.5707964           <--- this is where my ANGLEX array starts

 1.5707964

 1.5707964

 1.5707964

 1.5707964

 1.5707964

 1.5707964

 1.5707964

...

...

...

And obviously, I'm doing something wrong with my ANGLEX array - the *lst file is saying:

   INTERPRETATION OF LAYER FLAGS:

                        INTERBLOCK     HORIZONTAL    DATA IN

         LAYER TYPE   TRANSMISSIVITY   ANISOTROPY   ARRAY VKA   WETTABILITY

 LAYER      (LAYTYP)      (LAYAVG)       (CHANI)      (LAYVKA)      (LAYWET)

 ---------------------------------------------------------------------------

    1      UPSTREAM      HARMONIC      VARIABLE    VERTICAL K  NON-WETTABLE

    2      UPSTREAM      HARMONIC      VARIABLE    VERTICAL K  NON-WETTABLE

    3      UPSTREAM      HARMONIC      VARIABLE    VERTICAL K  NON-WETTABLE

    4      UPSTREAM      HARMONIC      VARIABLE    VERTICAL K  NON-WETTABLE

 WETTING CAPABILITY IS NOT ACTIVE IN ANY LAYER

 ERROR READING ARRAY CONTROL RECORD FOR               FACE ANGLE:

  1.5707964                                                                                                                                                                                              

... and there it ends.

Any suggestions? 

- what should be the proper format for the ANGLEX array? 

- where should it be located?

- does it have header? If so, what is the format?

- the same questions would be for the HANI array...

Any help would be really appreciated. 

I'm currently NOT using a GUI to build the model (apart from some pre-processing in GIS software and standalone mesh generator). Can you suggest a free GUI that I could use to create a simple working model so that I could look under the hood? As far as I know, MODELMUSE does not support USG yet?

Thanks!!

Pavel

[Julio Verdejo]

Hi Pavel,

Perhaps you can download a demo from a commercial GUI such as visual MODFLOW, sometimes the restrictions are about use time limits or number of cells limits. But if you just want to check how to look under the hood this maybe a good aproach

Julio

--
You received this message because you are subscribed to the Google Groups "MODFLOW Users Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to modflow+u...@googlegroups.com.
To post to this group, send email to mod...@googlegroups.com.
Visit this group at http://groups.google.com/group/modflow.
For more options, visit https://groups.google.com/d/optout.

[Cptn. Nemo]

OK, figured out the ANGLEX and HANI placement:

- ANGLEX is located outside of the NLAY loop, above the definition of HK. The data is read using U1DREL module, so the ANGLEX array needs 'reading directions' as specified in the on-line manual: http://water.usgs.gov/ogw/modflow-nwt/MODFLOW-NWT-Guide/index.html?array_reading_utility_modules.htm (CONSTANT, INTERNAL, EXTERNAL, OPEN/CLOSE). 

- HANI array is located just below HK array.

The model runs, but it is wrong, because I incorrectly interpreted the face angle parameter... Two steps forward, one step back...

So now I need to figure out, how to properly calculate the ANGLEX for every node-to-node connection within the mesh made of Voronoi polygons. 

Will have to read in the *.disu file and perhaps calculate the angle from node coordinates of adjacent cells???

Tips, anyone? :)

Thanks! P.

[Christian Langevin]

Hey Pavel, here is a python function I've used to calculate the USG anglex value for two cells, where Pn = (xn, yn) and Pm = (xm, ym).  Maybe you can adapt this.

def get_angle(Pn, Pm):

    """

    Fine angle of line n-m from horizontal x axis.  Return

    radians from 0 to 2 * pi

    """

    deltaX = Pm[0] - Pn[0]

    deltaY = Pm[1] - Pn[1]

    a = np.arctan2(deltaY, deltaX)

    if a < 0:

        a = 2 * np.pi + a

    return a

--

[Cptn. Nemo]

Hi Christian,

thanks for the tip, got it working - basically using the same approach - calculating deltaX and deltaY and calculating the angle as ATAN (deltaY/deltaX).

Still trying to figure out the cell-cell connections in unstructured grid, just don't understand why there is '+1' connection in the IAC array?

Cheers, P.