topic Re: dotted raster pattern II in ArcGIS GeoStatistical Analyst Questions

dotted raster pattern II

BrookeHodge — Thu, 14 Oct 2010 16:24:53 GMT

I wrote in a few weeks ago about dotted pattern in my interpolated raster in geostatistical analyst. I was able to fix it using a different model type. I am now using a different dataset and am seeing the same pattern (see attachment), but this time choosing a different model type does not rectify the problem. The pattern seems to occur from my point dataset used for the interpolation. No matter which model type I use, the pattern still occurs.

Does anyone have any ideas on how I can fix this?

Thanks,
Brooke

Re: dotted raster pattern II

EricKrause — Fri, 15 Oct 2010 00:52:22 GMT

I don't see the dot pattern in the pdf you posted. It's certainly nowhere near as distinct as either of the topics you've recently posted on the subject of strange rasters.

Re: dotted raster pattern II

BrookeHodge — Fri, 15 Oct 2010 13:13:52 GMT

Hi Eric,

Thanks for getting back to me. You are correct in that the pattern is not as distinct as my last post. The data in my last post was evenly distributed and there were a lot more points closer together, so I think the pattern was more dramatic. Here my point dataset is a bit more sparse and not evenly distributed, and at that scale I think it was a bit hard to see. I have attached a new pdf that is zoomed in more and shows the dots a bit better. There are two screen shots, and one includes the point dataset used for the interpolation. As you can see, it looks as though the dots on the raster are forming at where the points are.

Also, while I have you here and sense you're a GA product engineer, I was wondering if you knew of any materials that go into depth on GA, in particular determining which method/parameter/model/transformation etc. to use. I have read the GA tutorial provided by ESRI, but it doesn't go into detail on each of the parameter options for the different interpolation methods. I usually find myself using what others have used in similar situations, or just guessing and trying different options until I find something that works, but when I'm writing up a report/paper, it's nice to have reasons why I chose certain parameters. If there are any materials, books, etc. that you think would be useful, I would greatly appreciate any recommendations.

Thanks Eric!
Brooke

Re: dotted raster pattern II

SteveLynch — Wed, 27 Oct 2010 15:12:40 GMT

Brooke

The Smooth option in the searching neighborhood was developed to fix this problem, see explanation in
Gribov, A., and Krivoruchko, K. 2004. Geostatistical Mapping with Continuous Moving Neighborhood. Mathematical Geology, Volume 36, Number 2, February 2004.

or on-line
http://training.esri.com/campus/library/bibliography/recorddetail.cfm?id=30571

Steve Lynch
ESRI