0

I am making a prospective map of fossil deposits using the "Weighted Overlay" ArcGIS tool. Input factor rasters are mostly topographic in nature (elevation, roughness, TPI, etc). Initially, I had planned on assigning the weightings of each factor based on expert opinion. However, due to the large number of input rasters it is now apparent that a less subjective, statistical technique of determining these weights would be preferable. I want to know if the analytical hierarchy process (AHP) can be suitably applied to determine the weights of the input rasters in this case.

If not: can you suggest any other methods to reduce the subjectivity of the designation of these weightings?

I am using ArcGIS 10.2

Thanks

1

In my education in GIS, I met some statistical methods, however I never encountered better, than AHP for weighting objectively. Unfortunately, there was only one software which provided built-in AHP support (surprisingly an entire wizard) for this purpose: IDRISI. There are some user scripts for ArcGIS, like AHP 1.1 or AHP-OWA 2.0, which has Ordered weighted averaging. You should try these scripts and if they refuse to work, than you can always use a spreadsheet, a pen and a piece of paper or an online AHP calculator.

Just remember, you consistency ratio must be under 10% or 0.1. :)

3

This tool might be of benefit to you if you want to convert your rasters to vector. There is a OWA function and a WLC(weighted linear combination) there appears to be support for ArcGIS 10.1 and above.

http://mcda4arcmap.codeplex.com/ and a paper http://www.ryerson.ca/~crinner/pubs/Pages12-13_from_Cartouche86_Winter-Spring2013.pdf

0

Not entirely sure what you are after experimentally but you may want to investigate a Weights of Evidence approach.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.