Organized by the Government of Lesotho, in partnership with FAO, this workshop will discuss the production of Lesotho National Land Cover annual maps and the extraction of land cover statistics. The course is intended for technical experts interested in developing or strengthening their GIS analysis and mapping skills as well as data science skills.

The Government of Lesotho, through the ReNOKA movement, is organizing a Land Cover Mapping national workshop under the framework of the Integrated Catchment Management Programme. Land cover data is the foundation information to sustainable resources management and agricultural development. An accurate, high-resolution and up-to-date land cover product serves as the basis to model soil erosion, vegetation productivity (cropland and grassland), hydrological processes and carbon sequestration. The advent of freely accessible and multi-temporal Earth Observation (EO) data, namely the Sentinel satellites, unlock the potential of frequent and accurate land cover map production. This workshop aims to train participants in using a state-of-the-art land cover methodolgy developed by FAO and IT tools to produce yearly land cover maps. A few practical examples of how the produced land cover data can be used to support sustainable land management will also be demonstrated.

Objective of the training is to build capacity in technical experts in:
1) Producing annual national land cover maps using FAO’s newly developed methodology 2020
2) Extracting statistics from the land cover map at national and subnational level
3) Sharing geospatial products and statistics using web gis technology

The workshop will consist of 4 modules held over 5 days:

Module I: Land Cover Mapping Recap and Principles

• In this module trainees will be provided with a review of the land cover mapping principle which will allow them to gain a full picture of the concepts and typical workflow,
• Commonly used satellite image products and tools will be introduced to the audience

Module II: Field Survey Design for Reference Data Collection

• The in-situ component will be addressed. The importance of the survey design for the scope of land cover mapping will be highlighted. GIS and statistical methods for optimal survey design will be explored as well as best practices in georeferencing data in the field
• A hands on exercise will be held outdoor in the surroundings of the workshop site to gather in-situ data using GPS, followed by migration of GPS data in GIS and QA-QC excercise

Module III: Land Cover Map Production with Google Earth Engine

In this module, hands-on exercise will be carried out to guide the trainees through the entire process of producing a national land cover map for Lesotho. Google Earth Engine will be used. The main steps will include:

• Connecting to EO data (Sentinel2 data)
• Pre-processing the data into Analysis Ready Data
• Uploading in-situ data
• Defining land cover classes
• Training a classification algorithm (random forest).
• Producing and visualizing the land cover map
• Assessing accuracy of map
• Exporting map in gis format outside of Google Earth Engine

Module IV: Explore the Land Cover Product(s) and Extract Statistics Statistics

This module will be dedicated to extracting statistics from the land cover map at national and subnational level using common GIS desktop software (free and open source) such as QGIS

To sign up for the training, please follow this link. 

Find out more: FAO enhancing land cover monitoring capacities in Lesotho using Earth Observations data