|
|
| GeoCommunity Mailing List |
| |
| Mailing List Archives |
| Subject: | Re: [gislist] Filling gaps in grids (was: "grid cuting") |
| Date: |
04/11/2005 10:55:01 AM |
| From: |
Quantitative Decisions |
|
|
At 09:27 AM 4/11/2005 -0500, Bing Du wrote:
>Dose any one know how to fill in nearby data into no data cells in a
>grid file of arcview3.2?
How this is done depends on the purpose. For grids that represent
continuous surfaces, you want to interpolate across the gaps. I describe
the technique in general, using ArcView 3.2 (+Spatial Analyst) examples, at
http://www.quantdec.com/SYSEN597/GTKAV/section9/chapter_29d.htm#Patching
. This method can be enhanced with regression techniques if you have
additional covariate datasets, but the mechanics of it will be quite
similar: you create a grid to interpolate across gaps and then either fill
in only the NoData cells or some small neighborhood around each one with
the new values.
For grids that represent categorical variables, the process depends
strongly on how much and what kind of data you have and how you are
modeling them. There are ad-hoc methods, such as majority neighbor fills,
there are statistical methods, such as logistic regression, there are
physical models of growth, proximity, size, and so on, and there are
probabilistic models, such as Markov models with spatial correlation, that
can also be used. Most raster GISes, including AV 3.2's Spatial Analyst,
support enough "map algebra" operations to carry out any of these methods,
but few include the statistical exploration, diagnostic, and graphing tools
needed to find the right coefficients to use or to estimate the possible
errors, so you would likely have to resort to a good statistical package
for that part of the work.
--Bill Huber
Quantitative Decisions
_______________________________________________
gislist mailing list
gislist@lists.geocomm.com
http://lists.geocomm.com/mailman/listinfo/gislist
_________________________________
This list is brought to you by
The GeoCommunity
http://www.geocomm.com/
|
|
Sponsored by:
For information
regarding
advertising rates
Click Here!
|
