Interpolation using Known points with Identical Value

[PolashBanerjee]

I am using IDW to interpolate the value of known points which coincidently have identical value. On interpolation, the map is showing only one value i.e. the value of the known point all over the map. Ideally, value per pixel should taper off  with increase in distance from the known points. I have tried to interpolate with other tools like Ordinary Kriging and Empirical Bayesian Kriging.  But they all are showing the same result. 

[ChrisDonohue__GISP]

Some questions to help clarify the issue:

• Can you label the two images you provided so viewers can discern what differentiates them?
• Does the issue involve any of the known points with the same pixel values also having the same X,Y values?
• Can you provide an image where the point values are labeled on the image by the point.
• Also, any chance there are data points to work with besides the very linear collection currently depicted?  I mention this as without more data points to the "sides" of the linear set you have, the interpolation out to the "sides" won't be meaningful.  I.e. interpolation doesn't provide anything useful in vast areas without data points.

Chris Donohue, GISP

[DanPatterson_Retired]

duplicate points need to be removed .... consider the case of an IDW with 12 points set as the sampling framework to perform the interpolation.  If 4 of those points are duplicates, then they still would participate in the interpolated values within the potential sampling area.  If 3 of the duplicates were removed, then the program would be forced to look a bit further to replace them... this would affect the numeric output of the interpolated value... the degree to which it would differ would depend upon where the duplicates are located with respect to the cell that is being calculated.