Geospatial Analysis Using Python and QGIS Training Course

Participants can assess, acquire and organise spatial data from appropriate sources.

• Open geospatial data portals, licensing terms and authoritative source assessment
• OGC web services: WMS, WFS and WMTS connections in QGIS
• CSV, GeoJSON, GeoPackage, Shapefile and GeoTIFF format selection
• STAC catalog searches for satellite scenes and raster assets
• Attribute schemas inspected for keys, domains, units and null values
• File naming and directory conventions supporting repeatable analysis

Participants can prepare reliable vector and tabular datasets for analysis.

• Geometry validity checks using QGIS Check Validity and GEOS rules
• Duplicate features, sliver polygons and multipart geometries identified systematically
• Field Calculator expressions standardise text, dates and coded values
• Joins, relates and unique identifiers preserve attribute-table integrity
• Missing values profiled before interpolation, classification or aggregation decisions
• GeoPandas dataframes read, inspect and filter spatial datasets

Participants can select, apply and document coordinate systems appropriate to spatial measurements.

• Geographic and projected coordinate systems distinguished for distance calculations
• EPSG definitions selected from area, purpose and measurement requirements
• On-the-fly projection checked against layer metadata and basemap alignment
• Reprojection with GDAL and QGIS preserves geometry and attributes
• UTM zones, national grids and equal-area projections compared
• Datum transformations documented where source datasets use different realizations

Participants can answer location and access questions using vector geoprocessing methods.

• Buffer, clip, dissolve and intersection operations answer proximity questions
• Spatial joins transfer polygon attributes to points and lines
• Overlay outputs checked for topology errors and unexpected area changes
• Network distance distinguished from Euclidean distance for access studies
• Python GeoPandas overlays reproduce QGIS vector geoprocessing workflows
• Service-area inputs defined from facilities, roads and travel assumptions

Participants can calculate spatial indicators and integrate operational reference data.

• QGIS expressions calculate conditional fields, ranks and grouped summaries
• SQL WHERE clauses filter attributes using explicit selection criteria
• Group-by summaries quantify counts, rates and category distributions
• Field types, precision and units controlled before calculated outputs
• Python pandas aggregation produces auditable indicator tables
• Quality checks compare totals before and after spatial joins

Participants can prepare raster data and derive raster-based analytical indicators.

• Raster bands, pixels, nodata values and cell-size implications
• Raster Calculator combines surfaces through documented map algebra expressions
• Clipping, resampling and reprojection align rasters before comparison
• Nearest-neighbour and bilinear resampling selected by data type
• Zonal statistics summarise raster values within administrative polygons
• GDAL raster metadata inspected for extent, resolution and band structure

Participants can select satellite imagery and calculate interpretable spectral indicators.

• Electromagnetic spectrum concepts connect sensor bands to land-cover signals
• Sentinel-2 and Landsat scenes selected by date, cloud and resolution
• Cloud masks, scene classification layers and quality flags screened
• NDVI and NDWI calculated from reflectance bands with QGIS
• False-colour composites reveal vegetation, water and built surfaces
• Surface reflectance products distinguished from raw digital-number imagery

Participants can classify imagery, assess accuracy and compare land-cover conditions over time.

• Training samples designed with class definitions and spatial independence
• Supervised classification concepts: features, labels and decision boundaries
• Random Forest parameters tuned using held-out validation samples
• Confusion matrices report producer, user and overall accuracy
• Post-classification filtering removes isolated pixels without masking evidence
• Change-detection rasters compare classified dates using consistent class schemes

Participants can identify and interpret geographic clustering, concentration and spatial dependence.

• Point patterns mapped using counts, rates and denominator populations
• Choropleth classifications compared using quantiles, natural breaks and equal intervals
• Kernel density estimation identifies concentrated incidents and service demand
• Spatial autocorrelation assessed with Global Moran's I statistic
• Local indicators flag clusters and spatial outliers for investigation
• MAUP and ecological fallacy considered when interpreting area statistics

Participants can model continuous surfaces and derive terrain-based spatial measures.

• IDW interpolation parameters selected from sampling density and surface behavior
• Kriging concepts: variograms, range, sill and nugget interpretation
• Cross-validation compares prediction errors before choosing interpolation methods
• Contour generation converts continuous surfaces into interpretable map products
• Terrain derivatives calculate slope, aspect and hillshade from DEMs
• Hydrological conditioning identifies sinks, flow direction and accumulation paths

Participants can write clear Python scripts for vector, raster and tabular processing.

• Python functions separate inputs, transformations, outputs and reusable parameters
• GeoPandas workflows read, reproject, join and export vector datasets
• Rasterio opens bands, masks arrays and writes derived rasters
• pandas merges spatial indicators with operational reference tables
• Exception handling and assertions expose failed assumptions early
• Configuration files store paths, thresholds and project-specific variables

Participants can automate recurring geoprocessing tasks through scripts, models and command-line tools.

• PyQGIS scripts call processing algorithms and manage project layers
• QGIS Model Designer chains geoprocessing steps with named parameters
• Batch processing applies identical transformations across multiple source files
• Command-line GDAL utilities convert, warp and mosaic raster datasets
• Script logs capture run dates, inputs, settings and output locations
• Version control principles protect tested scripts and analytical decisions

Participants can create maps that communicate analytical findings accurately and clearly.

• Visual variables: colour, size, value and hierarchy communicate analytical meaning
• Classification choices align legend intervals with decision thresholds
• Accessible colour palettes accommodate common forms of colour-vision deficiency
• QGIS layouts combine maps, charts, scale bars and source credits
• Labels, callouts and annotation rules prevent ambiguous map interpretation
• Export settings produce print-ready PDF and web-ready image files

Participants can review geospatial work for quality, reproducibility and responsible data use.

• Analytical question, unit of analysis and decision threshold documented
• Metadata records source, date, licence, CRS and transformation history
• Positional, thematic, temporal and logical accuracy assessed separately
• Peer review checks parameters, code, maps and stated limitations
• Privacy risks assessed before mapping sensitive people, sites or events
• Reproducibility package organises data dictionaries, scripts and outputs

Participants can build, review and present a complete geospatial analysis addressing a defined decision question.

• Problem statement defines stakeholders, decision, area and success measures
• Source inventory records provenance, licences, dates and fitness criteria
• QGIS and Python workflow produces validated spatial indicators and maps
• Methods note explains CRS, assumptions, transformations and quality checks
• Findings distinguish observed patterns, uncertainty and recommended action
• Peer review and presentation address technical questions and decision implications