topic Re: Darcy Flow output rasters have strange checkerboarded surface in Spatial Data Science Questions

Darcy Flow output rasters have strange checkerboarded surface

Anonymous User — Thu, 05 Dec 2013 04:58:11 GMT

Original User: zack.benedetto

Hello,

I am attempting to make a groundwater flow model using the Darcy Flow tool in Arcmap 10.1. However, the output I'm getting has a very strange checkerboard/gridded surface. Its not just the grid of the raster, each "checkboard box" is about 10 cells by 10 cells. See attached images below.

Does anybody know why it is creating this pattern? I made a flow vector field from the direction raster and it didn't seem to be greatly affected by it. Particle track seemed to have normal results too.

I've also read that if the residuals are high then the output might be meaningless. All of my values are at least in the thousands though. How large is too large for the residuals to be? Everything I used was in units of feet and days.

Other info:
I created the head input raster as a surface from groundwater well levels. For the thickness, transmissivity and porosity input rasters, I assigned values to the underlying geology using data collected from pump tests from a couple different sources, as well as some tabulated values.

Thanks!
Zack Benedetto

P.S. I'm doing this project for my undergraduate senior thesis, which I have a presentation on next week, so any quick answers would be greatly appreciated.

residuals
[ATTACH=CONFIG]29593[/ATTACH]

direction
[ATTACH=CONFIG]29594[/ATTACH]

magnitude
[ATTACH=CONFIG]29595[/ATTACH]

Re: Darcy Flow output rasters have strange checkerboarded surface

Anonymous User — Thu, 05 Dec 2013 08:08:56 GMT

Original User: mark909

It would be interesting to see your DEM with the hillshade function applied.

Re: Darcy Flow output rasters have strange checkerboarded surface

MarkBoucher — Thu, 05 Dec 2013 16:41:11 GMT

I'm following these two forum posts that present questions about "stripe" patterns. There is some discussion on solutions that move in the direction of raster size as a problem for some older versions. The "grid" pattern is not part of the discussion. I don't have time to filter through the forum content for a solid recommendation, but maybe there is one in there that can help you. The second link is imbedded in a comment on the first one.

http://forums.arcgis.com/threads/17237-Problem-running-Region-Group-tool

http://forums.arcgis.com/threads/97714-Problem-with-query-builder-and-output-raster-for-the-quot-Extract-by-attribute-quot-tool?p=347326#post347326

Re: Darcy Flow output rasters have strange checkerboarded surface

Anonymous User — Fri, 06 Dec 2013 18:54:46 GMT

Original User: zack.benedetto

Thanks, it seems like my issue might be that there is too much data, since I think this type of groundwater analysis is usually done for much smaller areas. I'll have to look into a script that can parse the data up, run the tool, and then patch it back together.

Any idea about how large (and in what units) do the residuals become too large for the output to make sense/be useful?

Thanks,
Zack

Re: Darcy Flow output rasters have strange checkerboarded surface

MarkBoucher — Fri, 06 Dec 2013 19:27:25 GMT

Rasters can be created/converted in/to different formats. http://resources.arcgis.com/en/help/main/10.2/index.html#//001200000032000000

I've heard that the grid format is more efficient. Also, having the raster in a gdb may make some difference.

Frankly, I don't deal with that many different raster formats, but possibly a certain raster format may take less memory to process and so that might get around the size issue. Might be worth some research.