30 GIS experts and soil scientists from various governmental institutions alongside university professors from Rwanda participated in the workshop.  The training was tailored to cover the complete cycle, from data management to data mining for predictive mapping of soil properties at the national scale. The workshop was opened by Dr Charles Bucagu (Deputy DG, Rwanda Agriculture and Animal Resources Development Board), Mr Jean Claude Musabyimana (Permanent Secretary, Ministry of Agriculture and Animal Resources) and Mr Hudson Wang (Representative of the Embassy of China).

The workshop took place over 40 hours spread over five days. The training consisted of practical exercises on DSM techniques using the statistical programming language R. The first exercise was based on a standard soil profile database from North Macedonia while the second one was based on an example dataset from Rwanda. The different workshop sessions were organised as follows:

• Day 1: Official opening and introduction to the workshop objectives and expected outcomes
• Day 2: Introduction to spatial data, DSM and open-source GIS software
• Day 3: Soil data and covariate preparation
• Day 4: Using Google Earth Engine to process and download covariates from the cloud
• Day 5: Use of a machine-learning algorithm to predict and map soil properties, final exercise and workshop closing

The training was designed so that participants with no prior experience in using a programming language could, by the end of the workshop, independently modify the pre-prepared R scripts for their own national data to generate national soil property maps. 

The training strengthened and further developed the country’s capacity in DSM. Furthermore, the workshop became an ideal setting for connecting national soil data institutions and research entities and fostering their collaboration to overcome data fragmentation. An informal community of participants from the relevant data-holding entities was established to coordinate activities with the overall objective to generate a national soil information system as well as soil property maps.

This workshop is part of the GSP’s capacity development programme, which constitutes the essential means for the establishment of a country-driven global soil information system (GloSIS). Over the last two years, online trainings have proven advantageous in terms of outreach and flexibility. In fact, the GSP has been able to significantly ramp up its capacity development programme by training over 700 national experts worldwide. However, the recent revival of in-person trainings such as this one, through its immediate and tangible impact has underscored the importance of live interaction among relevant stakeholders. The GSP looks forward to hosting many more virtual and in-person trainings to leverage the benefits of both.