Advanced GIS and Remote Sensing Spatial Modelling Training Course

Course overview

Spatial decisions in land management, infrastructure planning, environmental monitoring and disaster risk reduction depend on models that are technically defensible, reproducible and transparent about uncertainty. Many GIS practitioners can create maps and run basic geoprocessing tools, but need stronger capability to integrate multi-source imagery, terrain, vector, climate and field data into spatial models that answer operational questions. This course addresses that gap: moving from visual interpretation and isolated analyses to calibrated, validated models that can withstand technical review and inform investment, policy and operational decisions.

Participants build advanced workflows for raster and vector analysis, suitability modelling, remote-sensing classification, change detection, terrain derivatives, spatial statistics and model validation. They work with ArcGIS Pro, QGIS, Google Earth Engine and GDAL to prepare data, select meaningful predictor variables, automate repeatable processing chains and assess the accuracy of outputs. The course also covers weighted overlay, multi-criteria decision analysis, supervised classification, vegetation indices, object-based concepts, spatial autocorrelation, sampling design and uncertainty reporting.

Instruction combines expert-led demonstrations with guided analysis of realistic spatial datasets. Each participant develops a documented spatial modelling project, including a problem statement, data inventory, processing workflow, model assumptions, validation results, maps and recommendations for decision-makers. The final day includes a technical review session in which participants test and refine their model against peer and instructor feedback, producing a reusable template for application in their own organisation.

The course is designed for GIS analysts, remote-sensing specialists, environmental professionals and technical teams already using GIS who need to design more rigorous models rather than simply operate mapping software.

Course objectives

By the end of this course, participants will be able to:

• Design a spatial modelling workflow that links a defined operational question to data, assumptions, processing steps and decision outputs
• Prepare and harmonise raster, vector, elevation and satellite datasets using projection, resampling, masking and quality-control methods
• Build a weighted suitability model using standardisation, constraint mapping and multi-criteria decision analysis
• Derive terrain, hydrological and proximity variables from digital elevation models and spatial feature datasets
• Classify multispectral imagery using supervised classification, training samples and spectral indices in Google Earth Engine
• Evaluate land-cover and change-detection outputs using confusion matrices, accuracy metrics and independent validation samples
• Apply spatial autocorrelation and hotspot analysis to identify clustering, bias and spatial dependence in model inputs
• Produce a documented spatial model report with reproducible workflows, uncertainty statements, maps and decision recommendations

Benefits of attending

Target competencies

Who should attend

• GIS Analysts — who need to turn mapped datasets into validated suitability, risk and predictive models
• Remote Sensing Specialists — who need stronger methods for image classification, change detection and accuracy assessment
• Environmental Scientists — who model habitat, land-use, watershed or environmental-risk conditions for evidence-based action
• Urban and Regional Planners — who must compare spatial development scenarios and justify site-selection recommendations
• Disaster Risk and Resilience Officers — who require defensible exposure, vulnerability and hazard-modelling workflows
• Geospatial Data Managers — who oversee reproducible analytical standards, data quality and technical review processes

Requirements and prerequisites

Participants should be comfortable working in a desktop GIS and should already understand layers, attribute tables, coordinate reference systems, basic vector geoprocessing, raster cell values and map layout. Experience with either ArcGIS Pro or QGIS is expected, including loading data, joining tables, clipping data and running simple analysis tools. Familiarity with satellite imagery, Python or statistical software is useful but not essential; coding is not a prerequisite. This is not a beginner mapping course. Participants do not need prior experience with Google Earth Engine, formal remote-sensing classification or spatial statistics, as these are taught through guided exercises.

Training methodology

The programme uses short instructor-led technical briefings followed by guided work in ArcGIS Pro, QGIS, Google Earth Engine and GDAL. Participants analyse a shared land-use and environmental planning case, progressively building data-preparation, suitability, classification and validation workflows. Exercises require participants to make and justify modelling choices, not merely follow tool steps. Group review sessions compare assumptions, weighting schemes and accuracy results. On the final day, each participant adapts the methods to a workplace use case and creates an implementation plan for data, governance and quality assurance.

Course outline

Day 1: Spatial modelling foundations and data engineering

• Framing decision questions as conceptual spatial models
• Data provenance, scale, resolution and modifiable areal unit effects
• Coordinate reference systems and spatial alignment diagnostics
• Raster resampling methods and implications for model outputs
• Vector topology checks and attribute-domain quality control
• GDAL workflows for format conversion, clipping and raster mosaicking
• Reproducible folder structures, metadata and processing logs

Workshop: Participants assemble and document a multi-source geodatabase for a land-suitability case, producing a data inventory and preprocessing workflow.

Day 2: Suitability, terrain and multi-criteria modelling

• Raster algebra and map algebra expressions
• Standardising continuous and categorical criteria
• Hard constraints, Boolean masks and exclusion zones
• Weighted overlay and analytic hierarchy process concepts
• Digital elevation model derivatives for slope, aspect and curvature
• Cost-distance, Euclidean distance and accessibility surfaces
• Sensitivity testing of criteria weights and thresholds

Workshop: Participants build a weighted site-suitability model for infrastructure development and produce a ranked suitability map with a sensitivity comparison.

Day 3: Remote sensing classification and change analysis

• Satellite sensor selection, spectral resolution and revisit cycles
• Cloud masking, compositing and image preprocessing in Google Earth Engine
• Normalized Difference Vegetation Index and built-up spectral indices
• Training-sample design for supervised land-cover classification
• Random forest classification and feature selection
• Post-classification filtering and class-area estimation
• Image differencing and post-classification change detection

Workshop: Participants classify two time periods of Sentinel-2 imagery and produce a land-cover change map with quantified class transitions.

Day 4: Validation, spatial statistics and uncertainty

• Probability and stratified sampling for reference data collection
• Confusion matrices, producer accuracy and user accuracy
• Overall accuracy, F1 score and Cohen's kappa interpretation
• Spatial autocorrelation using Global Moran's I
• Local clustering and hotspot analysis with Getis-Ord Gi*
• Residual mapping and spatial bias diagnostics
• Uncertainty communication through confidence maps and limitations statements

Workshop: Participants validate their classification and suitability outputs, producing an accuracy report, hotspot map and uncertainty statement.

Day 5: Operationalising and communicating spatial models

• ModelBuilder and QGIS graphical model workflows
• Batch processing and parameterised geoprocessing tools
• Google Earth Engine script structure and export workflows
• Scenario comparison and model version control practices
• Cartographic design for analytical and decision maps
• Technical reporting of assumptions, methods and validation evidence
• Governance controls for data updates, peer review and model approval

Workshop: Participants complete a capstone spatial model pack containing workflow documentation, validated outputs, decision maps and an implementation plan for their organisation.

Tools & standards covered

ArcGIS Pro, QGIS, Google Earth Engine, GDAL

A typical training day

Live online deliveries follow the same structure in the East Africa Time zone, with shorter screen blocks and longer breaks.

What the fee includes

How you can take this course

Certification

Participants who complete the full five days receive the Skillset Development Certificate of Completion, stating the course title, course code, dates and delivery format — suitable for professional-development records and employer reimbursement.

Frequently asked questions