Pedometric mapping with R

Course title: Pedometric mapping with R
Course code: ISRIC-RDSM
Subject areas (Curriculum): Soil terrain mapping, Software tools
Duration: 2 days
Short description: This hands-on-software training course provides a systematic guide through various soil mapping steps using the open source software R + SAGA GIS. Topics covered: soil sampling, preparation of the soil covariate data, data import and formatting, fitting of splines to soil profile data, overlay and visualization of soil data (soil profiles, soil polygon maps, soil covariates), data screening, spatial analysis, spatial predictions and uncertainty assessment. The emphasise is put on how to implement specific analysis in R and how to extend the existing functionality. The objective of this course is to review the state-of-the-art pedometric methods and connected R packages and promote use of analytical techniques for automated analysis and mapping of soil variables.
Moderators: T. (Tom) Hengl,
Optional: D. (Dylan) Beaudette, G.B.M. (Gerard) Heuvelink
Target audience: MSc and PhD level students working primarily with soil data. Soil survey teams / GIS teams working with soil data.
Requirements: Knowledge requirements:
  • Soil mapping (advanced)
  • GIS modelling (advanced)
  • R programming (basic/advanced)
  1. DAY 1
    1. An overview of pedometric techniques (1.5 hrs)
    2. From spatial sampling to soil maps (1.5 hrs)
    3. Sources of uncertainty in soil data and the concept of mapping efficiency (1.5 hrs)
    4. Packages for soil mapping - aqp, gstat, GSIF and plotKML (1.5 hrs)
  2. DAY 2
    1. Introduction to aqp - soil profile data in R (1.5 hrs)
    2. Exercise: loading soil data and running various analysis steps (1.5 hrs)
    3. Exercise: generating maps and reports using participant’s data (1.5 hrs)
    4. Summary discussion (1.5 hrs)
Software / materials: Participants typically use their own laptop Software:
  • R (packages: gsif, plotKML, sp, raster, rgdal, gstat, geoR, aqp...)
  • RStudio or Tinn-R
  • Google Earth
Case studies: Ebergotzen case study;
Literature: Beaudette, D. 2010: Open Source tools for soil scientists.
Hengl, T. 2009: A practical guide to geostatistical mapping.
Hengl, T. 2012: Mapping efficiency and information content.
Minumum number of participants: 5
Maximum number of participants: 25
Further inquiries: Tom Hengl


Spatial prediction of sand content for the Ebergotzen case study (ordinary kriging vs regression-kriging).