GIS and Remote Sensing Integration Training Course

Participants can select imagery and sensor products that meet temporal, spatial and spectral requirements.

• Electromagnetic spectrum regions related to land, water and infrastructure features
• Passive optical, thermal and active radar sensors compared
• Spatial, spectral, radiometric and temporal resolution trade-offs evaluated
• Sentinel, Landsat and commercial imagery products compared systematically
• Acquisition dates matched to phenology, seasonality and operational events
• Cloud cover, viewing geometry and revisit intervals assessed

Participants can define a structured geospatial project with traceable datasets, questions and outputs.

• Decision questions translated into spatial criteria, outputs and acceptance measures
• Data inventory records source, licence, date, resolution and known limitations
• Geodatabase schemas designed for assets, observations, boundaries and derived layers
• Unique identifiers maintained across field, tabular and spatial records
• Version control practices applied to project files and processing outputs
• Data quality checks identify duplicates, gaps, invalid geometry and attributes

Participants can create and maintain reliable vector datasets from imagery, plans and field records.

• On-screen digitising uses snapping, topology rules and defined capture scales
• Feature classes created with coded domains and controlled attribute values
• Topology validation detects overlaps, gaps, dangles and duplicate features
• GPS observations imported, checked and linked to operational records
• Attribute joins and spatial joins combine tabular and mapped evidence
• Editing logs document source imagery, interpretation dates and confidence

Participants can prepare raster datasets for consistent analysis within a common spatial framework.

• Raster formats, bands, bit depth and NoData values inspected
• Mosaics constructed using matching projections, cell sizes and radiometric rules
• Clipping, resampling and alignment preserve analytical comparability between layers
• Nearest-neighbour, bilinear and cubic resampling methods compared
• Raster compression choices balanced against storage, speed and information loss
• Pyramids and statistics built for efficient display and processing

Participants can create readable maps and interpret image patterns using documented visual cues.

• Visual interpretation uses tone, texture, shape, pattern and association
• False-colour composites reveal vegetation, water stress and built surfaces
• Classification symbology distinguishes categories without misleading visual hierarchy
• Map layouts apply scale bars, legends, north arrows and source statements
• Colour ramps selected for sequential, diverging and categorical data
• Map review checks positional accuracy, labels, hierarchy and audience relevance

Participants can interpret spectral responses and calculate indices for targeted feature extraction.

• Spectral signatures compared for vegetation, soil, water and impervious surfaces
• Band combinations selected for specific materials and sensor characteristics
• NDVI calculated to measure vegetation greenness and cover variation
• NDWI and MNDWI applied to surface-water delineation tasks
• NDBI and built-up indices tested against urban land-cover classes
• Band ratios and threshold values documented for repeatable extraction

Participants can prepare imagery for multi-date analysis using appropriate correction and masking methods.

• Digital numbers converted to radiance, reflectance or surface reflectance values
• Atmospheric correction methods compared for optical sensor products
• Geometric correction assessed using ground control points and RMSE
• Cloud, shadow and haze masks created before spectral analysis
• Image co-registration checked before multi-date change detection
• Terrain correction applied where topography affects image brightness

Participants can create representative training data and run supervised land-cover classification methods.

• Classification schemes define mutually exclusive, observable land-cover classes
• Training samples stratified across classes, locations and image conditions
• Reference polygons separated from independent validation samples
• Maximum likelihood classification applied using spectral training statistics
• Random forest classification configured with predictor bands and features
• Class imbalance identified before model fitting and accuracy reporting

Participants can refine image-classification results using segmentation, ancillary data and post-processing rules.

• Image segmentation groups neighbouring pixels into meaningful objects
• Scale, shape and compactness parameters calibrated for target features
• Object attributes include mean spectra, texture, geometry and context
• Ancillary GIS layers improve separation of confusing spectral classes
• Majority filters remove isolated pixels while retaining legitimate small features
• Rule-based post-classification edits documented with transparent decision criteria

Participants can quantify classification quality and communicate uncertainty to decision makers.

• Confusion matrices compare mapped classes against independent reference observations
• Overall accuracy, producer accuracy and user accuracy calculated correctly
• Kappa statistics interpreted cautiously alongside class-specific performance measures
• Stratified random sampling supports valid class-level accuracy estimates
• Positional, thematic and temporal uncertainty documented in final outputs
• Limitations statements distinguish measured error from interpretation judgement

Participants can detect, quantify and explain material spatial change across multiple observation dates.

• Image dates harmonised for season, sensor characteristics and preprocessing level
• Post-classification comparison identifies transitions between mapped land-cover classes
• Image differencing measures reflectance or index changes between dates
• NDVI time series reveal crop cycles, disturbance and vegetation recovery
• Change thresholds tested against reference samples and operational requirements
• Transition matrices summarise losses, gains and persistent land-cover areas

Participants can derive terrain and drainage information from digital elevation models.

• Digital elevation model sources compared for resolution, coverage and vertical accuracy
• Sink filling prepares elevation surfaces for flow-routing analysis
• Slope, aspect and curvature derived for terrain suitability assessment
• Flow direction and accumulation identify drainage pathways and catchments
• Watersheds delineated from pour points and validated against mapped channels
• Hillshade and terrain visualisation reveal landform and drainage anomalies

Participants can interpret core SAR products and identify suitable applications for radar imagery.

• SAR backscatter explained through wavelength, polarisation, incidence angle and roughness
• VV and VH polarisation compared for vegetation, water and built surfaces
• Speckle filtering balances noise reduction against feature preservation
• Radar geometry effects include layover, shadow and foreshortening
• SAR coherence supports change detection and infrastructure monitoring
• Flood mapping combines low backscatter, terrain and permanent-water masks

Participants can build transparent multi-criteria spatial models for site, risk and priority decisions.

• Buffer, overlay and proximity analysis quantify spatial relationships between features
• Reclassification converts diverse criteria into comparable suitability scores
• Weighted overlay combines environmental, technical and operational selection criteria
• Constraint masks exclude prohibited, unsafe or infeasible locations
• Sensitivity analysis tests the influence of alternative criterion weights
• Suitability results validated against known sites and stakeholder requirements

Participants can integrate field observations with imagery and GIS layers while maintaining traceability.

• Field forms specify mandatory attributes, coded values and photo requirements
• Mobile data collection captures coordinates, observations, timestamps and attachments
• Positional accuracy assessed against device settings and collection conditions
• Ground-truth points linked to classification training and validation datasets
• Quality assurance rules flag missing attributes, outliers and improbable locations
• Field verification plans prioritise uncertain classes and high-risk areas

Participants can automate repeatable geoprocessing steps and publish controlled geospatial outputs.

• ModelBuilder and graphical workflows document repeatable geoprocessing sequences
• GDAL commands batch convert, reproject and clip raster datasets
• Processing parameters externalised for consistent reruns with new data
• OGC WMS, WFS and WMTS services compared for publishing needs
• Web map layers configured with scale ranges, pop-ups and metadata
• Publishing checks address access controls, performance and data sensitivity

Participants can build, review and present a complete GIS and remote sensing analysis addressing a defined operational decision.

• Problem statement defines decision users, geography, criteria and success measures
• Data register records sources, permissions, quality checks and processing choices
• Integrated workflow combines imagery, GIS layers, field evidence and analysis
• Accuracy evidence and limitations support defensible interpretation of results
• Final maps, dashboards or reports communicate findings and recommended actions
• Peer review tests reproducibility, usability and decision relevance before presentation