Hands on Digital Soil Mapping

Share on:
2015-05-18 12.39.23.jpg Namibia_OCS250m.png

Overview  

This course introduces methods and software for management, analysis and mapping of soil variables within the R environment for statistical computing. The course alternates between lectures and computer practicals and covers a variety of subjects, such as geostatistics, linear regression and machine learning for soil mapping, quantification of uncertainty and soil map validation. The course aims at soil geographers and environmental scientists who want to learn more about the theory and practice of digital soil mapping. After this course participants will be able to apply the methods learnt to their own datasets. Lecturers are experienced pedometricians and soil data analysis specialists.  

  • Gerard Heuvelink, course coordinator (geostatistics for soil mapping, uncertainty assessment)

  • Bas Kempen (introduction to R, data preparation, machine learning for soil mapping, validation of soil maps)

  • Eloi Ribeiro (soil databases)

  • Madlene Nussbaum, guest lecturer (machine learning and data mining)

  • Titia Mulder, guest lecturer (soil spectroscopy)

 

Who is it for? 

This course is intended for soil and environmental professionals, researchers and students interested in producing soil maps and/or using local, regional and global soil datasets for digital soil mapping. Participants must have a basic level of statistics, geo-information science and soil/environmental science. Experience with computer programming in R is advantageous but not required. A number of recorded sessions from a previous Spring School can be accessed via the ISRIC YouTube channel at: http://youtube.com/c/ISRICorg.

 

Software installation

Please note that to participate in the DSM course you have to bring your own laptop computer, with R and RStudio software installed. Also, do not forget a power plug (travel) adapter if you need one. In the Netherlands the power sockets are of type C and F. We will not be able to provide you with these.

Software installation instructions can be found here, and the associated R code to test your installation here. Please read and follow these carefully to ensure you come fully prepared to the spring school.

 

Programme*

*Working programme subject to changes.

DAY 0 (OPTIONAL, Friday, 25 May 2018):

Block Topic Room/ type Lecturer
9.00 – 17.30 Introduction to the statistical software R* Lumen 1&2 Guided self-study Bas Kempen

(* A minimum number of participants is required for the optional day)

 

DAY 1 (Monday, 28 May 2018):

Time Topic Room/ type Lecturer
8.30 – 9.00 Registration and Coffee Hall Gaia building  
9.00 – 9.45

Official opening of the ISRIC Spring School

Group photo

Gaia 1 Rik van den Bosch (ISRIC director), Bas Kempen
9.45 – 10.15

Course introduction and overview

Lumen 1&2 Gerard Heuvelink
10.15 – 10.45 Geostatistics for soil mapping

Lumen 1&2

lecture

Gerard Heuvelink
10.45 – 11.15 Coffee break    
11.15 – 12.30 Geostatistics for soil mapping

Lumen 1&2

lecture

Gerard Heuvelink
12.30 – 13.30 Lunch GAIA first floor  
13.30 – 15.00 Geostatistics for soil mapping

Lumen 1&2

computer practical

Gerard Heuvelink, Titia Mulder
15.00 – 15.30 Coffee break    
15.30 – 17.30 Geostatistics for soil mapping

Lumen 1&2

computer practical

Gerard Heuvelink, Titia Mulder

 

DAY 2 (Tuesday, 29 May 2018):

Block

Topic

Room/ type

Lecturer

9.00 – 09.45

Data preparation for DSM

Lumen 1&2

lecture

Bas Kempen

09.45 – 10.30 Data preparation for DSM

Lumen 1&2

computer practical

Bas Kempen,

Titia Mulder

10.30 – 11.00

Coffee break

   

11.00 – 12.30

Data preparation for DSM

Lumen 1&2

computer practical

Bas Kempen,

Titia Mulder

12.30 – 13.30

Lunch

   

13.30 – 14.15

WOSIS soil database and preparation of soil point data

Lumen 1&2

lecture

Eloi Ribeiro

14.15 – 15.00

WOSIS soil database and preparation of soil point data

Lumen 1&2

computer practical

Eloi Ribeiro, Bas Kempen

15.00 – 15.30

Coffee break

   

15.30 – 17.30

Remote and proximal sensing for natural resource inventories

Lumen 1&2

lecture and exercise

Titia Mulder

19.00 – 22.00

Dinner in town (Colours World Food restaurant)

 

 

DAY 3 (Wednesday, 30 May 2018):

Block

Topic

Room/ type

Lecturer

9.00 – 10.30

Machine learning 1: Introduction to Random Forest Modelling

Lumen 1&2

lecture

Bas Kempen

10.30 – 11.00

Coffee break

   

11.00 – 12.30

Machine learning 1: Introduction to Random Forest Modelling

Lumen 1&2

computer practical

Bas Kempen

Gerard Heuvelink

12.30 – 13.30

Lunch

   

13.30 – 15.00

Uncertainty quantification and propagation

Lumen 1&2

lecture

Gerard Heuvelink

15:00 – 15:30

Coffee Break

   

15.30 – 17:30

Uncertainty quantification and propagation

Lumen 1&2

computer practical

Gerard Heuvelink

Bas Kempen

 

DAY 4 (Thursday, 31 May 2018):

Block

Topic

Room/ type

Lecturer

9.00 – 10.30

Machine learning 2: Understanding methods, model selection and interpretation

Lumen 1&2

lecture

Madlene Nussbaum

10.30 – 11.00

Coffee break

   

11.00 – 12.30

Machine learning 2: Understanding methods, model selection and interpretation

Lumen 1&2

computer practical

Madlene Nussbaum,

Bas Kempen

12.30 – 13.30

Lunch

   

13.30 – 15.00

Validation of digital soil maps

Lumen 1&2

lecture

Bas Kempen

15.00 – 15.30

Coffee break

   

15.30 – 17.30

Validation of digital soil maps

Lumen 1&2

computer practical

Bas Kempen, Gerard Heuvelink

 

DAY 5 (Friday, 1 June 2018):

Block

Topic

Room/ type

Lecturer

9.00 – 10.00

Visit to World Soil Museum

WSM

Stephan Mantel

10.00 - 10.30

Workshop

Opportunity for course participants to present their work and receive feedback

Lumen 1+2 Gerard Heuvelink, Bas Kempen, Titia Mulder

10.30 – 11.00

Coffee break

   

11.00 – 12.30

Workshop

Opportunity for course participants to present their work and receive feedback

Lumen 1+2

Gerard Heuvelink, Bas Kempen, Titia Mulder

12.30 – 13.30

Lunch

   

13.30 – 14.30

Course evaluation, certificates

Lumen 1+2

Gerard Heuvelink

14.30 – 15.30

Digital soil resource inventories
Status and prospects (guest lecture)

Lumen 1+2

David Rossiter

15.30 – 17.00

Closing words and ISRIC borrel (drinks and snacks)

Lumen 1+2

Rik van den Bosch

 

Materials

The materials for the DSM course are available for download below. The materials are compressed in zip files. We recommend to create a folder on the hard drive of your computer at an easy-accessible location (e.g. D:/springschool) and store the materials in that folder. Do not store the materials in a folder on your desktop. After unzipping, do not change the names of the folders and files because code might not run anymore.

  1. Introduction to R
  2. Geostatistics
  3. Data Preparation
  4. WoSIS
  5. Soil Sensing
  6. Machine Learning 1
  7. Uncertainty Assessment
  8. Machine Learning 2
  9. Validation