Remote Sensing for Climate Monitoring Training Course

Participants can select sensor types and acquisition parameters for defined climate-monitoring tasks.

• Passive optical, thermal and active radar sensing principles compared
• Spectral bands linked to vegetation, water, soil and built surfaces
• Spatial resolution, swath width and revisit interval trade-offs
• Landsat, Sentinel and MODIS mission characteristics for monitoring
• Viewing geometry, sun angle and terrain effects on reflectance
• Radar wavelength, polarization and incidence angle for surface observations

Participants can define a traceable imagery acquisition plan from monitoring requirements.

• Scene search parameters defined by area, dates and cloud thresholds
• USGS EarthExplorer, Copernicus Data Space and Google Earth Engine catalogs
• Metadata fields checked before downloading or processing satellite scenes
• Acquisition calendars aligned with growing seasons and hazard periods
• Baseline periods selected for climate normals and anomaly calculations
• Data licensing, access conditions and retention requirements documented

Participants can prepare compatible spatial datasets for climate-monitoring analysis.

• Raster cells, bands, extent and resolution interpreted correctly
• Coordinate reference systems chosen for area, distance and local mapping
• Reprojection effects assessed before combining multi-source climate datasets
• Resampling methods compared for categorical and continuous raster variables
• Vector boundaries, points and grids prepared for zonal analysis
• Spatial alignment checks applied before pixel-based change detection

Participants can prepare optical imagery suitable for consistent multi-date analysis.

• Top-of-atmosphere and surface reflectance products distinguished for analysis
• Atmospheric correction purpose, assumptions and residual limitations examined
• Cloud, cirrus and cloud-shadow masks applied using quality bands
• Terrain shadow and snow contamination identified in optical scenes
• Median composites built from multiple observations within monitoring windows
• Image quality records captured in a processing log

Participants can interpret and prepare Sentinel-1 radar data for climate applications.

• Sentinel-1 backscatter interpreted for roughness, moisture and structure
• VV and VH polarizations compared for land and water targets
• Speckle filtering trade-offs assessed before classification or change analysis
• Radiometric terrain correction applied in mountainous monitoring areas
• Permanent water, floodwater and wet soil signatures differentiated
• Radar limitations documented for vegetation, urban areas and wind

Participants can derive and interpret satellite vegetation indicators for drought monitoring.

• Normalized Difference Vegetation Index calculated from red and near-infrared bands
• Enhanced Vegetation Index selected where canopy saturation affects NDVI
• Vegetation Condition Index derived from historical NDVI extrema
• Seasonal phenology separated from anomalous vegetation stress signals
• Rainfall, soil moisture and vegetation observations compared during drought
• Administrative-area summaries produced using zonal statistics and percentiles

Participants can map changing water conditions using optical and radar observations.

• Normalized Difference Water Index calculated for open-water detection
• Modified Normalized Difference Water Index applied in built-up areas
• Pre-event and post-event radar scenes compared for flood extent
• Threshold selection tested against reference observations and visual interpretation
• Flood extent separated from permanent water using baseline masks
• Water-area time series summarized for reservoirs, wetlands and rivers

Participants can assess land-surface temperature patterns and their planning implications.

• Land-surface temperature distinguished from near-surface air temperature measurements
• Thermal infrared emissivity and atmospheric effects considered during retrieval
• Daytime satellite temperature patterns interpreted alongside land-cover classes
• Urban heat hotspots identified using temperature anomalies and population data
• Cooling effects of trees, water and reflective surfaces compared
• Heat-risk maps qualified using temporal coverage and exposure limitations

Participants can produce land-cover change evidence relevant to adaptation planning.

• Training samples designed for representative land-cover class signatures
• Random forest classification configured with spectral and ancillary predictors
• Object-based and pixel-based classification approaches compared for mapping
• Post-classification comparison used to identify land-cover transitions
• Change trajectories interpreted for deforestation, urban growth and wetland loss
• Class definitions documented using a clear land-cover legend

Participants can construct multi-year satellite time series and identify meaningful anomalies.

• Regular temporal composites created from irregular satellite acquisitions
• Seasonal decomposition separates trend, seasonality and residual variation
• Climate normals calculated from defined reference-period observations
• Standardized anomalies used to compare conditions across locations
• Theil-Sen slope estimates applied to non-normal environmental time series
• Mann-Kendall trend tests interpreted with serial-correlation cautions

Participants can create spatial hazard indicators from satellite-derived and climate datasets.

• Drought hazard surfaces combined from rainfall, vegetation and soil-moisture indicators
• Flood susceptibility factors assembled from terrain, drainage and land cover
• Heat hazard mapped using temperature persistence and nocturnal conditions
• Coastal erosion indicators derived from shoreline position time series
• Hazard thresholds justified through policy, science and local records
• Multi-criteria evaluation weights documented and tested for sensitivity

Participants can combine remote-sensing evidence with social and asset data for prioritisation.

• Exposure datasets assembled for population, assets, crops and infrastructure
• Vulnerability proxies assessed for relevance, bias and data currency
• Hazard, exposure and vulnerability combined using transparent indicator logic
• Grid-based prioritisation avoids misleading administrative-average risk conclusions
• Equity considerations incorporated when identifying adaptation priority locations
• Priority maps translated into intervention targeting and monitoring criteria

Participants can evaluate remote-sensing outputs using appropriate validation methods and metrics.

• Reference samples designed using random, stratified and spatial sampling methods
• Confusion matrices produced for classified land-cover and flood maps
• Overall, producer and user accuracy interpreted without overclaiming performance
• F1 score and balanced accuracy used for imbalanced classes
• Independent validation separated from model training and parameter tuning
• Field observations, high-resolution imagery and agency records evaluated as references

Participants can communicate uncertainty and limitations in climate-monitoring products.

• Measurement, classification and temporal uncertainty identified in monitoring workflows
• Cloud gaps, missing observations and sensor changes assessed explicitly
• Error propagation considered when combining multiple indicator layers
• Confidence classes assigned using evidence quality and validation results
• Maps annotated with caveats, source dates and appropriate claims
• Decision thresholds tested against false-positive and false-negative consequences

Participants can structure reproducible cloud-based workflows for satellite climate analysis.

• Image collections filtered by geometry, date and metadata properties
• Band calculations scripted for vegetation, water and thermal indicators
• Cloud masking functions applied consistently across Landsat and Sentinel scenes
• Reducer functions used for composites, zonal summaries and time series
• Export settings specified for rasters, tables and reproducible outputs
• Script documentation records assumptions, versioning and input datasets

Participants can create clear, decision-ready maps and summaries from analysis outputs.

• QGIS raster styling reveals meaningful thresholds without visual distortion
• Map projections selected to preserve relevant geographic properties
• Legends, scale bars and annotations designed for nontechnical readers
• Zonal statistics linked to districts, watersheds and programme areas
• Charts connect satellite trends with seasons, events and interventions
• Map layouts checked for dates, sources, uncertainty and classification logic

Participants can specify an operational climate-monitoring process that teams can maintain.

• Standard operating procedures define inputs, steps, outputs and responsibilities
• Indicator specification sheets record formulas, units, thresholds and update cycles
• Data provenance captured through source registers and processing records
• Quality assurance checks assigned before publication and decision use
• Data governance roles defined for access, review, retention and release
• Monitoring calendars align production cycles with operational decision windows

Participants can present remote-sensing findings with clear claims, limits and recommended actions.

• Decision briefs structure findings around choices, evidence and consequences
• Technical methods translated without concealing uncertainty or limitations
• Before-and-after imagery used carefully for intervention performance claims
• Maps paired with summary statistics, local context and reference dates
• Stakeholder questions anticipated for data quality, costs and relevance
• Recommendations distinguished from observations, assumptions and model outputs

Participants can build, review and present a complete remote-sensing climate-monitoring product for an adaptation decision.

• Monitoring question converted into indicators, sensor choices and acquisition windows
• Analysis-ready imagery prepared with documented preprocessing and quality checks
• Hazard indicator, exposure layer and priority analysis produced
• Validation evidence and uncertainty statements incorporated into final outputs
• Operational workflow specified with responsibilities, update frequency and quality controls
• Stakeholder briefing presents maps, findings, limitations and recommended decisions