Forest Inventory Mapping Using GIS Training Course

Participants can select and document appropriate spatial reference systems for forest inventory data.

• Geographic and projected coordinate systems used in forestry mapping projects
• Datum selection, map projection distortion and local grid considerations
• UTM zones, national grids and coordinate transformations evaluated
• Horizontal accuracy, vertical accuracy and positional tolerance distinguished
• Raster cell size matched to inventory scale and forest heterogeneity
• Coordinate reference metadata recorded for field, vector and imagery datasets

Participants can design a relational spatial data model for forest inventory records.

• Entity relationships between plots, trees, stands, strata and compartments
• Unique identifiers preserving links across repeated inventory measurements
• Domain values and coded lists controlling attribute entry consistency
• Geodatabase feature classes for boundaries, plots, access routes and observations
• Tree, plot and stand tables structured for inventory calculations
• Join and relate operations connecting field measurements with mapped features

Participants can select a sampling design suited to stated inventory precision and cost requirements.

• Simple random, systematic and stratified sampling designs compared
• Plot size, shape and spacing matched to forest structure
• Sample allocation using proportional, equal and optimum allocation methods
• Cluster plots and nested plots for multi-size tree measurement
• Sampling intensity calculated from variance, precision and confidence requirements
• Access constraints, slope and safety incorporated into plot placement

Participants can prepare field data collection procedures and assess GNSS location quality.

• Diameter at breast height, height and species measurement protocols
• Plot centre coordinates recorded using GNSS averaging procedures
• Waypoint, track and polygon capture for forest field surveys
• Differential correction concepts and GNSS error sources examined
• Field forms, digital validation rules and mandatory attributes configured
• Chain-of-custody records maintained for field observations and photographs

Participants can identify, correct and document common errors in forest inventory field datasets.

• Range checks for diameter, height, age and stocking attributes
• Duplicate plot identifiers and repeated tree records detected
• Missing values, impossible dates and invalid species codes resolved
• Outlier screening using scatterplots, summaries and field notes
• Plot relocation rules applied to inaccessible or disturbed locations
• Quality-control logs documenting corrections, reasons and reviewer approval

Participants can select and prepare satellite imagery appropriate for forest inventory mapping.

• Landsat, Sentinel-2 and commercial imagery characteristics compared
• Acquisition dates selected around phenology, cloud cover and inventory timing
• Surface reflectance products distinguished from top-of-atmosphere reflectance
• Cloud, shadow and haze masking applied before classification
• Image mosaics clipped to study boundaries and analysis masks
• Spectral bands, indices and texture layers prepared for modelling

Participants can derive terrain variables and integrate supporting datasets into inventory analyses.

• Digital elevation models assessed for resolution, voids and artefacts
• Slope, aspect and terrain position index derived from elevation
• Hydrography, roads and settlements screened for mapping constraints
• Soil, geology and climate layers aligned with inventory boundaries
• Distance surfaces calculated for roads, streams and forest edges
• Ancillary variables evaluated for ecological relevance and data quality

Participants can create forest cover classes using supervised image classification methods.

• Class definitions separating forest, shrubland, plantation, water and agriculture
• Training samples selected from field plots and interpreted imagery
• Spectral signatures compared using band combinations and vegetation indices
• Maximum likelihood and random forest classifiers contrasted
• Object-based image analysis concepts for crown and stand delineation
• Classification rules documented with source data and decision thresholds

Participants can produce a classified forest cover map and refine its spatial outputs.

• Training and validation samples separated before model fitting
• Random forest classification run using spectral and terrain predictors
• Post-classification filtering removes isolated pixels and speckle noise
• Minimum mapping units applied to operational forest cover classes
• Manual interpretation resolves ambiguous boundaries and mixed pixels
• Class labels, symbology and attribute codes prepared for delivery

Participants can delineate inventory strata and stands from imagery, terrain and management evidence.

• Stratification criteria based on forest type, age, density and accessibility
• Stand boundaries interpreted from imagery, contours and field evidence
• Segmentation parameters adjusted for heterogeneous forest canopy conditions
• Compartments subdivided using natural features and operational access routes
• Stand attributes populated from plot summaries and interpreted characteristics
• Boundary confidence recorded where imagery or field evidence remains uncertain

Participants can calculate plot and stand inventory attributes from tree measurement records.

• Basal area calculated from diameter at breast height measurements
• Stem density expanded from plot counts to hectare estimates
• Quadratic mean diameter derived for stand structure reporting
• Merchantable volume estimated using species-specific volume equations
• Height-diameter models used where measured heights are incomplete
• Plot summaries grouped by strata, stand and management compartment

Participants can estimate above-ground biomass and carbon using inventory plots and spatial predictors.

• Allometric equations selected by species group and forest type
• Above-ground biomass calculated from diameter, height and wood density
• Root-to-shoot ratios applied for below-ground biomass estimates
• Carbon fractions and carbon dioxide equivalents calculated transparently
• Plot biomass linked with spectral indices and canopy metrics
• Prediction surfaces generated with documented model assumptions and limits

Participants can model continuous forest attributes from plot data and mapped predictor variables.

• Response variables selected for volume, biomass, height or canopy cover
• Predictor screening identifies collinearity, missingness and non-linear relationships
• Regression models fitted using plot observations and spatial covariates
• Random forest regression compared with linear model performance
• Cross-validation estimates prediction error across sampled forest conditions
• Raster predictions masked and summarized by management unit boundaries

Participants can evaluate forest map accuracy and produce adjusted area estimates.

• Reference samples designed independently from classification training samples
• Confusion matrices summarize producer, user and overall accuracy
• Commission and omission errors interpreted for forest reporting decisions
• Stratified accuracy sampling improves estimates for rare land-cover classes
• Area-adjusted estimates calculated from error matrices and map proportions
• Confidence intervals reported with sample design and reference-source limitations

Participants can detect, quantify and explain forest cover changes between inventory periods.

• Image co-registration checked before comparing multi-date forest scenes
• Post-classification comparison identifies persistence, loss, gain and transition
• Normalized difference vegetation index trajectories interpreted for canopy disturbance
• Change thresholds tested against reference imagery and field observations
• Annual forest loss summarized by compartment, watershed and protection zone
• False change caused by seasonality, clouds and sensor differences investigated

Participants can quantify key inventory uncertainties and apply quality assurance controls.

• Sampling error, measurement error and model error distinguished
• Error propagation considered across biomass calculation and map production
• Confidence intervals calculated for means, totals and area estimates
• Quality assurance plans assign checks, tolerances and responsible reviewers
• Independent review procedures test reproducibility of spatial analysis outputs
• Version control and audit trails protect inventory data integrity

Participants can adapt inventory mapping outputs for carbon, restoration and conservation reporting needs.

• Forest reference levels linked with activity data and emission factors
• Biomass maps support carbon stock baselines and restoration monitoring
• Restoration priority areas identified using degradation, slope and connectivity criteria
• Canopy cover metrics used for habitat condition assessment
• Permanent sample plots planned for repeat carbon stock measurements
• Reporting tables align mapped evidence with project monitoring indicators

Participants can prepare clear forest inventory map products and supporting technical documentation.

• Map layouts include scale, projection, legend, sources and production date
• Symbology communicates forest classes, uncertainty and management priorities clearly
• Metadata records lineage, spatial reference, processing and quality information
• ISO 19115 elements applied to inventory geospatial dataset documentation
• Technical reports present methods, estimates, limitations and recommendations
• Web-ready exports prepared with appropriate resolution and file formats

Participants can build, review and present a complete forest inventory mapping package for a defined management area.

• Project brief translated into inventory objectives, classes and reporting outputs
• Geodatabase, plot records, imagery and ancillary layers assembled coherently
• Sampling rationale and field-data quality evidence documented for review
• Forest cover classification, stand strata and inventory estimates produced
• Accuracy results, uncertainty statements and management implications synthesized
• Final map package presented with methods, metadata and recommendations