Interpolate climate surface from points

[PatrickSchaefer]

Hello, I am relatively new to GIS, so forgive me if the solution to this problem is actually very easy.

I am attempting to create a climate surface from an xy table produced by ClimateWNA (http://www.genetics.forestry.ubc.ca/cfcg/ClimateWNA/ClimateWNA.html). This program quires several databases for various climatic parameters and required imputing x, y, and z coordinates from a DEM (GTOPO30 was sufficient for me). This first step wasn't too difficult, and I successfully created a table with the relevant climatic parameters at over 2 millions points. The area I am interested in the the western Canadian Arctic Archipelago (Mainly Victoria Island and the surrounding mainland).

I am now having difficulty interpolating these points into a surface. When I look at the locations of the points, they are restricted to the landmasses, as they should be, but when I attempt to interpolate across the surface, it only interpolates over the water leaving the landmasses blank. I have downloaded several country outlines and attempted to include them as a barrier feature, but keep getting this error: "Input barrier polyline features: The input is not within the defined domain. ERROR 000863: Invalid GP data type".

Any tips or suggestions would be greatly appreciated,
Patrick

[RamB]

Just a small hint, you can 'mask' your area of analysis. Please refer extent parameter for spatial analyst.

regards,

[PatrickSchaefer]

Thanks for the tip.

If I understand your hint correctly, I have already defined the extent to be identical in all the files I am using. Otherwise, there would be much more than 2 million data points.

I suspected the error might somehow have to do with extent, but it doesn't seem to be the case (as far as I can tell). The error message refers to 'domain' which I guess is different from 'extent'?

[RamB]

Hi

Domain is usually the range of values allowed for a field. Please check. Probably the calculated values are not falling within the domain. may be some integer-float-double-short-long sort of error.

mask is different to extent. mask i think can also be a irregular shape, extent is always a right-angled 4 sided polygon.

regards,

[SteveLynch]

ERROR 000863: Invalid GP data type".

Patrick
The barrier feature class must be polyline. You might be trying to use a polygon FC.

Steve

[PatrickSchaefer]

Thanks for the tips. I did successfully create the polyline feature from the polygon.

I started running the IDW with barriers interpolation, and it has been running for about a day and a half. Is it normal for the analysis to be taking this long? or has something gone terribly wrong? My computer is not that sluggish either (AMD 4 core, 3.5MHz processor and 8gb RAM).

Just to reiterate what I am trying to do: I want to create a surface from a series of data points representing average summer temperatures. However, my area of interest includes coastline and several islands. I have attached below an image showing the area, with the points coloured by their values, and the outline showing the polyline feature I am trying to use in the barriers field. I created this image in Quantum GIS, as I only have one license for arcGIS, which is why it looks awkward (I dont know how to use Quantum GIS.

Thanks again,
Patrick

[SteveLynch]

Patrick

Yes it can take a long time. The barrier implementation is using a line of sight approach. I.e. from the location where you want a value it'll "look" to see which points are "visible" and then use only those.

It'll also depend on the complexity of the barrier featureclass, i.e. how many vertices and on the number of input points and then on the number of output cells.

You should, however, first test how long it takes without the barrier.

Steve

[MarkBoucher]

Could one do this by interpolating first and then clipping the results? or would the lack of a barrier skew the results?

[PatrickSchaefer]

Yes, I was wondering about that as well, (and I think that would be equivalent to using a mask). I worry that some of the islands may be too close together, which would influence the interpolation.

The analysis is still running. I did run it without barriers and it took only a few minutes to complete. I think the barrier polyline I used may be too complex. I did not appreciate that a more complex barrier would result in longer running time.

[SteveLynch]

The mask is applied post interpolation so it'll not have the same effect as a barrier.

Try to reduce the number of vertices in your barrier. See tools in the Generalization toolset can be used to create a new feature class by smoothing or deleting some of these features.

Steve