End Member Mixing Analysis or Principal Component Analysis

[kazi rahman]

Hello!
Does any one have any experience on End Member Mixing Analysis or Applying Principal Component Analysis for river discharge?
I made a model of 40 km2 watershed using SWAT, I calibrate it and validate it
In my study area there is three different sources of water that makes the stream flow generation
they are ground water feeded tributaries snow melted tributaries and Glacier melted tributaries. I want to see each components contribution

I have collected some water sample and made chemical analysis of each of the sources and I want to plot and see the status of each tributaries
I believe I understand the theory behind the End Member Mixing Analysis but just need some excel spreadsheet or Matlab code to develop those triangular representation and hydrograph separation (slide no 8 and 48)

If you have plz share with me
kind regards
Kazi Rahman
PhD Student
University of Geneva, Switzerland

[Sheila Christopher]

EMMA is a very complex analysis that first involves a principal component analysis (multivariate analysis) that helps you determine if your watershed does indeed have three water sources.  A lot of people assume this.  Have you read the Rick Hooper papers?  Also look up one by Doug Burns.  I do have a spreasheet with examples.  I will try to dig it up.  It is in excel.  But you first have to do the PCA in a stats package first (to get eigen values).

Sheila

--
You received this message because you are subscribed to the Google Groups "ArcSWAT" group.
To post to this group, send email to arc...@googlegroups.com.
To unsubscribe from this group, send email to arcswat+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/arcswat?hl=en.