Remote Sensing for Forest Management Training Course

Participants can choose suitable earth-observation datasets for a defined forest-monitoring purpose.

• Landsat, Sentinel-2 and Planet imagery for optical forest observation
• Sentinel-1 radar backscatter for cloud-prone tropical forest monitoring
• LiDAR and digital elevation models for canopy and terrain analysis
• UAV imagery strengths, constraints and regulatory operating considerations
• Comparing spatial, spectral, radiometric and temporal sensor characteristics
• Data portals, licences and acquisition records for reproducible analysis

Participants can organise forest spatial data using appropriate coordinate systems and metadata.

• Geographic and projected coordinate reference systems for forest mapping
• UTM zone selection, datum consistency and coordinate transformation checks
• Vector, raster and tabular data structures in forest information systems
• Geodatabases structured around compartments, plots, boundaries and observations
• Metadata fields recording source, date, resolution, processing and limitations
• File naming, version control and folder structures for repeatable projects

Participants can prepare satellite imagery for consistent forest analysis.

• Radiometric calibration converting raw digital numbers into analysis-ready values
• Atmospheric correction methods for surface reflectance image preparation
• Cloud, shadow and haze masking using quality-assurance bands
• Geometric correction checked against roads, rivers and surveyed control points
• Mosaicking adjacent scenes while managing acquisition-date and illumination differences
• Image quality checks using histograms, band statistics and visual inspection

Participants can interpret forest features from imagery using systematic visual cues.

• Tone, colour, texture, pattern and context in forest interpretation
• Recognising plantations, natural forest, scrub, agriculture and settlements
• Canopy gaps, logging roads and skid trails as disturbance indicators
• False-colour composites for vegetation vigour and moisture interpretation
• Interpreting terrain effects, topographic shadows and seasonal phenology
• Photo interpretation keys documented with reference examples and decision rules

Participants can calculate and interpret vegetation indices for forest-condition assessment.

• NDVI calculation, interpretation and saturation limitations in dense forests
• EVI for improved canopy sensitivity under high biomass conditions
• Normalized Burn Ratio for fire severity and post-fire recovery
• NDMI and moisture indicators for drought and vegetation stress
• Index threshold selection supported by field observations and image review
• Time-series profiles separating seasonal variation from persistent canopy loss

Participants can design field reference data collection to support forest mapping and validation.

• Reference data requirements for training, validation and field verification
• Stratified random sampling across forest classes and change categories
• GPS waypoint collection, positional accuracy and field-recording protocols
• Plot forms capturing canopy cover, species, disturbance and photograph references
• Separating training samples from independent accuracy-assessment samples
• Managing access constraints, safety risks and seasonal fieldwork timing

Participants can produce a supervised forest-cover classification from labelled reference data.

• Training polygons representing homogeneous and spectrally distinct land-cover classes
• Maximum likelihood classification assumptions, outputs and common failure modes
• Random forest classification using spectral bands and derived indices
• Class definitions separating forest, non-forest, plantation and water
• Feature selection using bands, indices, texture and terrain variables
• Post-classification filtering to remove isolated pixels and mapping artefacts

Participants can assess when unsupervised and object-based approaches suit forest mapping tasks.

• K-means clustering for exploratory grouping of image spectral signatures
• ISODATA clustering parameters, convergence criteria and class interpretation
• Image segmentation creating objects from spectral, shape and scale attributes
• Object-based classification for plantation blocks, crowns and clearings
• Comparing pixel-based and object-based outputs against mapping requirements
• Recoding spectral clusters through imagery review and field evidence

Participants can quantify classification accuracy and report map-area estimates with appropriate caution.

• Confusion matrices comparing mapped classes with independent reference observations
• Producer's, user's and overall accuracy calculations for forest classes
• Omission and commission errors explained for management decision-makers
• Stratified validation sampling for rare classes and change events
• Area-adjusted estimates correcting mapped forest areas using reference samples
• Accuracy statements, confidence intervals and limitations in technical reports

Participants can combine forest inventory and remotely sensed variables for biomass assessment.

• Forest inventory variables including diameter, height, species and plot expansion
• Allometric equations for above-ground biomass estimation from plot measurements
• Linking plot coordinates with satellite pixels and canopy metrics
• Regression models relating biomass observations to spectral and radar predictors
• Spatial prediction maps, residual diagnostics and extrapolation warnings
• Biomass uncertainty sources from plots, models and remote-sensing inputs

Participants can detect and characterise forest change using multi-date imagery.

• Image-to-image comparison using aligned dates, sensors and preprocessing methods
• Post-classification comparison for forest loss, gain and class transition maps
• Spectral differencing and change-vector analysis for disturbance screening
• Distinguishing deforestation, degradation, seasonal clearing and regeneration signals
• Change thresholds tested against reference imagery and field observations
• Change maps summarised by administrative area, concession and protected zone

Participants can interpret radar data for forest observation where optical imagery is constrained.

• SAR backscatter, wavelength, polarization and incidence angle fundamentals
• Sentinel-1 VV and VH channels for forest-cover change screening
• Speckle filtering and terrain correction for radar image preparation
• Radar limitations in biomass estimation and dense-canopy discrimination
• Combining optical and radar predictors for cloud-resilient classification
• Interpreting moisture, roughness and structure effects on backscatter values

Participants can map fire extent and assess disturbance severity from remote-sensing evidence.

• Active-fire products, thermal anomalies and detection timing limitations
• Burned-area delineation using Normalized Burn Ratio and differenced indices
• Severity classes calibrated against vegetation damage and field evidence
• Mapping windthrow, insect damage and storm disturbance signatures
• Prioritising post-disturbance field inspections using spatial severity patterns
• Recovery monitoring through multi-season vegetation-index trajectories

Participants can apply GIS spatial analysis to forest planning and protection questions.

• Buffer analysis around rivers, roads, settlements and protected boundaries
• Slope, aspect and elevation derived from digital elevation models
• Least-cost paths for patrol routes and field-access planning
• Fragmentation metrics measuring patch size, edge density and connectivity
• Overlay analysis identifying encroachment risks within management zones
• Zonal statistics summarising forest indicators by compartment and jurisdiction

Participants can prepare remote-sensing activity data for forest carbon monitoring and reporting.

• Activity data, emission factors and stock changes in forest carbon accounting
• IPCC land-use categories and forest land conversion reporting rules
• Representing land-use transitions through consistent classification legends
• Sampling design requirements for area estimation and uncertainty reporting
• Forest reference levels, baselines and monitoring-period consistency
• Audit trails linking source imagery, methods, validation and reported estimates

Participants can use spatial indicators to assess habitat condition and restoration progress.

• Habitat suitability variables derived from canopy cover, elevation and proximity
• Forest connectivity analysis supporting corridor and protected-area planning
• Restoration monitoring using survival, canopy recovery and land-cover indicators
• Mapping invasive species risk from disturbance, roads and spectral evidence
• Selecting biodiversity indicators appropriate to remotely sensed data limitations
• Combining camera-trap, survey and habitat layers for ecological interpretation

Participants can prepare clear forest-monitoring maps and technical evidence for decision-makers.

• Map scale, projection, legend and symbology selected for intended users
• Colour schemes distinguishing forest classes, change categories and uncertainty
• Layout design including north arrow, scale bar, sources and disclaimers
• Charting forest area, loss rates and validation results accurately
• Technical methods sections documenting data, processing and quality controls
• Executive map briefs translating spatial findings into management actions

Participants can specify repeatable processing workflows for routine forest monitoring.

• ModelBuilder and QGIS processing models for repeatable geoprocessing sequences
• Google Earth Engine scripts for image filtering and composite creation
• Parameter files recording dates, thresholds, classes and processing decisions
• Batch processing procedures for annual imagery and multiple management units
• Quality-control checkpoints embedded before classification, validation and publication
• Workflow documentation enabling review, handover and future updating

Participants can build, review and present a complete forest remote-sensing project with documented methods, validated outputs and management recommendations.

• Problem statement defining forest-management decision, geography and reporting audience
• Data inventory matching imagery, field evidence and GIS boundaries
• Processing workflow documenting preprocessing, classification and change-analysis choices
• Validation plan specifying samples, accuracy metrics and uncertainty treatment
• Map portfolio presenting forest extent, change, priority areas and statistics
• Peer review testing assumptions, limitations, traceability and decision relevance