Contour water quality parameter around headland

[DouglasHersh]

I am new to these forums. Pretty new to ArcGIS too. I apologize in advance if the following question has already been answered.

I need to produce contour maps of sparsely sampled, water quality data (e.g. chlorophyll, bacteria) for an area with a complex coastline that includes islands and headlands. It is important that measured values on one side of a peninsula do not "leak" under the land to the water on the other side. On the other hand, I do want measurements on one side of a headland to influence points nearby (as the fish swims) on the other side of the headland. I don't expect a simple canned solution, but I imagine someone can describe the workflow that would get me there or point me to sources describing how this might be done. I would prefer not to use arbitrary, hand-drawn, line-of-sight barriers as that would be susceptible to artistic license.

Thank you for your help.

- Doug

[Luke_Pinner · Accepted solution]

If you have the ArcGIS GeoStatistical Analyst extension, have a look at the diffusion and kernel interpolation with barriers tools.

Otherwise, perhaps try the spline interpolation with barriers tool (requires ArcGIS Spatial Analyst extension).

Once you have interpolated a surface (raster), you can create contours with the contour tool.

Edit: Do you have coastline data already? What format are your samples in?

[DouglasHersh]

Luke,

Thanks for the information. I do have a license for the Geostatistical and Spatial Analyst extensions.

I already have coastline data. My sample locations are in Oracle SDO and my water quality data are in Oracle tables that I can query using query layers. I will follow up with the results of my work.

Doug Hersh
Environment Quality, Water and Wastewater
Massachusetts Water Resource Authority

[SteveLynch]

Doug

I suggest you look at using KernelInterpolationWithBarriers as it'll also produce an output of prediction standard errors.

Steve

[DouglasHersh]

Thank you, Steve.

That's two votes for kernel interpolation with barriers with the additional benefit of prediction standard errors.

- Doug