This content has been marked as final. Show 10 replies
So, the areal interpolation tool is kinda like an intersect or spatial join but with geostats?
Ted, areal interpolation is fundamentally a kriging method. We assume that the starting polygons are the result of averaging some underlying Gaussian process, and we model this underlying process using the data in the polygons (this is often called the "change of support" problem). This modeling creates a smooth prediction surface for the variable of interest. This surface is then averaged (or integrated) in the new polygons to get the new predictions.
The big difference between this approach and proportional intersection methods is that the latter assumes the density of the variable is constant across each polygon, which isn't a realistic assumption. Our approach also provides standard errors for the predictions.
You can read a short paper about our method here. We're also working on making a longer version available. We'll also have a workflow topic up for Beta 2. I can send you a copy early if you want to try to see it now.
I would be keen on that workflow. I was looking to do a bunch of extension work at beta 2, so early would be nice, because I have a few projects that may be able to use Geostats.
I just sent the workflow.
We'll also have a workflow topic up for Beta 2. I can send you a copy early if you want to try to see it now.[/QUOTE]
I'd also be interested in seeing the workflow topic for areal interpolation.
How does the areal interpolation allow for unequal population sizes if I am interpolating some disease rates per tract?
I currently do not have access to Geostatistical Analyst 10.1. Are there any other interpolation features in 10.0 that I could use or do I have to convert to points?
The formula for Binomial (Rate) areal interpolation is in this paper (it's free to download):
It accounts for varying population sizes by doing a correction to the empirical semivariogram so that polygons with larger populations exert more influence on the model.
Thanks for the paper. I'm looking at GA 10.1. Do I include the rates at the polygons that do not have measured data as IsNull? I have more "missing" data as my sample is small compared to my population. Can I extrapolate? I have access to the "ozone" layer example,where can I get more help available on the stepwise process for rate/binomial inter/extrapolation? Thank you.
The software will automatically ignore polygons with missing data. If the polygons are coded correctly, you shouldn't have to do anything.
Also, be careful with extrapolation. Areal interpolation is based on simple kriging, so it is inherently bad at extrapolation. As you move away from the source polygons, the predictions will converge to the mean value of the polygons.
Read the two help topics we have to areal interpolation. The workflow topic shows a Rate (Binomial) example, but the workflow is basically the same for all three data types:
What would you suggest for extrapolating the polygon data?