Remote Sensing for Crop Health Monitoring Training Course

Participants can explain how sensors, reflectance and spatial resolution affect agricultural interpretation.

• Electromagnetic spectrum regions used for crop canopy and soil observation
• Reflectance, absorption and scattering across visible, near-infrared and shortwave bands
• Spatial, spectral, temporal and radiometric resolution trade-offs for crop monitoring
• Passive optical sensors compared with thermal and radar observation methods
• Mixed pixels and edge effects within smallholder agricultural landscapes
• Atmospheric conditions affecting surface reflectance and image interpretation

Participants can select suitable satellite data sources for defined crop monitoring tasks.

• Sentinel-2 band configuration, revisit frequency and agricultural applications
• Landsat Collection 2 products for historical crop condition analysis
• Planetary-scale archives compared by resolution, access and licensing conditions
• Cloud-cover thresholds set for seasonal imagery acquisition plans
• Acquisition schedules matched to crop phenology and intervention timing
• Metadata checks for processing level, dates, projection and coverage

Participants can prepare field boundaries and supporting spatial data for image analysis.

• Field boundary digitising rules for irregular parcels and fragmented holdings
• Coordinate reference systems selected for area measurement and local mapping
• Attribute tables structured with crop type, planting date and management records
• Vector geometry checks identify gaps, overlaps, slivers and duplicate fields
• Farm records joined to field polygons using stable identifiers
• Ancillary layers assembled for soils, irrigation, elevation and weather stations

Participants can prepare imagery and identify quality limitations before crop analysis.

• Surface reflectance products selected instead of uncorrected top-of-atmosphere imagery
• Cloud and cloud-shadow masks applied before vegetation index calculation
• Scene classification layers inspected for snow, cirrus and saturated pixels
• Mosaicking procedures manage adjacent scenes covering large agricultural areas
• Resampling choices preserve appropriate pixel values for index analysis
• Quality assurance logs record scene exclusions, masks and processing decisions

Participants can calculate and interpret common vegetation indices for crop monitoring.

• NDVI calculation using near-infrared and red reflectance band values
• NDRE interpretation for chlorophyll variation in dense crop canopies
• EVI comparison with NDVI under high biomass conditions
• SAVI adjustments reduce soil background effects during early crop growth
• Index ranges interpreted against crop stage, field history and local conditions
• Threshold maps distinguish low, moderate and high vegetation vigour zones

Participants can identify probable water stress patterns using optical and thermal indicators.

• NDWI and moisture indices derived from near-infrared and shortwave bands
• Thermal infrared data interpreted for canopy temperature and water stress
• Crop Water Stress Index concepts linked to reference temperature conditions
• Irrigation uniformity patterns identified through within-field vegetation variability
• Rainfall records compared with moisture signals and crop development stages
• Water stress differentiated from disease, nutrient deficiency and poor establishment

Participants can analyse seasonal image sequences to detect crop development and anomalies.

• Vegetation index time series assembled from repeated cloud-free observations
• Smoothing methods reduce noise from residual clouds and atmospheric variation
• Seasonal profiles compared with expected planting, peak growth and harvest dates
• Anomaly detection identifies departures from baseline crop development patterns
• Field-level trajectories grouped by similar growth timing and vigour
• Missing observations handled using interpolation limits and confidence annotations

Participants can stratify agricultural areas by crop type and observable field condition.

• Training samples selected from reliable field records and visual interpretation
• Supervised classification separates crop classes using spectral and temporal features
• Unsupervised clustering identifies internally similar crop condition zones
• Confusion matrices assess classification agreement against reference observations
• Producer's and user's accuracy interpreted for operational crop maps
• Stratified maps support targeted sampling, extension visits and resource allocation

Participants can distinguish plausible crop stress causes using multiple evidence sources.

• Nutrient deficiency patterns compared with water stress and soil variability
• Pest and disease damage interpreted through spatial pattern and timing
• Flooding signatures identified using moisture indices, terrain and field observations
• Poor emergence distinguished from late planting and cloud contamination
• Headlands, tracks and shade excluded from agronomic stress interpretation
• Multiple hypotheses documented before assigning likely causes to anomalies

Participants can design field verification plans that test satellite-derived interpretations.

• Stratified random sampling allocates visits across mapped crop condition zones
• Field observation forms capture crop stage, damage, management and photographs
• GPS collection protocols maintain positional accuracy and consistent field identifiers
• Reference plots selected for healthy, stressed and ambiguous crop conditions
• Timing field visits around satellite passes and critical crop stages
• Validation evidence documented with uncertainty, access constraints and observer notes

Participants can identify how remote sensing contributes to cautious crop yield estimation.

• Yield components related to crop area, plant density and seasonal biomass
• Regression models connect vegetation metrics with observed plot yields
• Training datasets checked for representative farms, seasons and management conditions
• Early-season forecasts separated from end-of-season production estimates
• Out-of-sample validation measures prediction error using independent yield records
• Yield uncertainty communicated through ranges, assumptions and stated limitations

Participants can use Google Earth Engine to assemble and analyse seasonal satellite collections.

• Image collections filtered by area, date range and cloud metadata
• Band selection and masking functions prepare Sentinel-2 agricultural composites
• Median composites reduce residual clouds across defined monitoring periods
• Index expressions calculate NDVI, NDWI and NDRE from image bands
• Zonal statistics summarise field-level values from raster time series
• Export settings preserve projection, scale and attributes for downstream mapping

Participants can create clear crop health maps and field-level analytical outputs in QGIS.

• Raster symbology uses meaningful colour ramps and documented class breaks
• Zonal statistics calculate mean, median and variability for each field
• Raster calculator combines indices into transparent crop condition rules
• Spatial joins connect field observations with satellite-derived condition metrics
• Map layouts include scale bars, legends, dates, sources and caveats
• GeoPackage files organise project layers, attributes and reusable styles

Participants can translate crop condition maps into prioritised agronomic management actions.

• Management zones delineated from vegetation variability and operational constraints
• Irrigation inspection priorities based on persistent low-vigour field areas
• Fertiliser response assessed through before-and-after vegetation index comparisons
• Variable-rate application concepts linked to validated management zone evidence
• Scouting routes prioritise high-risk anomalies and representative reference fields
• Action registers assign responsible staff, deadlines and verification indicators

Participants can apply crop monitoring methods to food security and development programme decisions.

• Seasonal monitoring supports rapid crop condition assessments after weather shocks
• Area-based sampling frames combine administrative boundaries, fields and livelihood zones
• Smallholder plot constraints considered when interpreting medium-resolution satellite pixels
• Crop condition evidence triangulated with market, rainfall and household survey data
• Programme indicators distinguish agricultural outputs from remotely observed outcomes
• Data protection considerations govern farm boundaries and household-linked records

Participants can document analytical uncertainty and apply responsible crop monitoring controls.

• Uncertainty sources traced through imagery, boundaries, models and field observations
• Confidence ratings assigned to findings using evidence quality criteria
• False precision avoided in maps, area estimates and management recommendations
• Data governance rules define access, retention, consent and permitted sharing
• Algorithmic bias risks considered for underrepresented crops and farming systems
• Peer review checklists test reproducibility, traceability and agronomic plausibility

Participants can produce decision-focused reports and briefings from remote sensing findings.

• Executive map products answer defined decisions rather than display every layer
• Finding statements distinguish observation, interpretation, assumption and recommended action
• Priority matrices rank fields by severity, urgency, area and operational feasibility
• Map annotations identify verification locations and likely management concerns
• Technical appendices record data sources, processing steps and validation evidence
• Briefing narratives explain uncertainty without weakening appropriate operational decisions

Participants can design a repeatable crop health monitoring system for an organisation.

• Monitoring calendars assign imagery review dates and seasonal decision points
• Standard operating procedures define roles, inputs, checks and approval steps
• Data dictionaries standardise field identifiers, crop codes and condition categories
• Escalation protocols specify thresholds for field visits and management response
• Dashboard requirements prioritise actionable metrics over excessive map displays
• Resource plans estimate staff time, software access and field verification costs

Participants can build, review and present a complete crop health assessment with maps, evidence, priorities and an action plan.

• Area of interest defined with crop calendar, fields and decision question
• Satellite imagery selected, masked and documented using quality assurance procedures
• Vegetation and moisture indicators calculated for relevant crop growth stages
• Condition zones mapped and interpreted against contextual agronomic evidence
• Field verification priorities established through risk, uncertainty and operational value
• Management actions proposed with responsible roles, timing and monitoring indicators