This forum is shutting down! Please post new discussions at community.naturalcapitalproject.org

Generating input rasters using open-source GIS software

jellisojelliso Member
Hi all,

I'm working on a project in Ethiopia for which all my data and analyses until now have been done using open-source software (QGIS, etc). Because of this as well as my lack of access to ArcGIS, I'd like to see if there are instructions for generating some input rasters ("upslope source", "downslope retention", "riparian continuity" namely) outside of ArcGIS. 

On page 76 of the RIOS documentation, it says "instructions for calculating these inputs outside of ArcGIS are available on request." 

A big Thanks in advance!
James

Comments

  • DougDoug Administrator, NatCap Staff
    Hey James!

    I see this may have feel through the cracks. I'm going to ping a few folks to see if they can provide assistance!

    @BradEichelberger , @adrianvogl , any thoughts!?

    My apologies if this was already handled.
  • Thanks Doug! I'm still interested! -J
  • RichRich Administrator, NatCap Staff
    edited July 2015
    Sorry about this James.  I don't know that RIOS preprocessor inputs have been calculated on anything other than an ArcGIS client, either through scripting or by hand.  Not sure why the user's guide suggests otherwise.  For the moment you can look at the preprocessor ArcGIS code here to see how the inputs are calculated: http://data.naturalcapitalproject.org/rios_releases/rios_1.1.4_arcgis_preprocessor.zip

    For the very long term, RIOS 2.0.0 will implement these calculations directly so there will be no need to calculate "upslope source", "downslope retention", and such, but that's a way's out for the moment.
    Post edited by Rich on
  • adrianvogladrianvogl Member, NatCap Staff
    Hi James,
    We did generate a document that provides the step-by-step instructions to calculate these indices manually.  Unfortunately, the specific tools that are mentioned to run the calculations are based on the ArcGIS toolbox.  But hopefully with this guidance and some knowledge of QGIS you can work out how to replicate the steps using another method.  Here is a link to the document:

    Please let me know if you have any questions about the specific calculations.

    Best,
    Adrian
  • Thanks Rich and Adrian!
  • Hi all,

    I'm also trying to generate input rasters with open source tools; so far, I am sticking to python (aside: I am loving the pygeoprocessing module :D ). I'm working through the erosion control objective and have gotten as far as riparian continuity index, and now I'm confused :S

    Could you tell me if I've got the work flow correct? (I have no way of verifying this part as I'm unsure of the method and  don't have access to ArcGIS spatial analyst :/ )

    Riparian Continuity Index Workflow:
    i) generate stream network
    ii) buffer the stream network
    iii) clip the retention index (prepared earlier) to the buffer made in 2)
    iv) smooth each cell by taking the mean of the pixels in a 3x3 kenel around each cell, but do so without including pixels from the other side of the river.
    v) clip it to the buffers again.

    In your arcgis workflow, you have buffers separated by left and right to keep the mean from including pixels from the other side of the river. And in the Gura example the river has multiple channels. So.. 
    1) Does each channel have its left/right smoothed-LULC-R calculated separately or is left and right determined relative to the main river trunk?
    2) When smoothing by the mean 3x3 kernel, am I right in thinking that  the mean calculation only averages pixels within the buffer, such that:
    a) the LULC-R values that occur in stream pixels are included in the mean calculation, and 
    b) the LULC-R values immediately outside the buffer are NOT included in the mean calculation?

    Many thanks in advance! I do hope I'm not barking up the wrong tree!
Sign In or Register to comment.