Concepts
Correction and cross-sensor comparison¶
SpectralBridge exists because airborne, drone, and satellite sensors do not observe the same landscape in directly comparable ways, even on the same day.
Different spectra
Band centers, widths, and response functions vary across sensors.
Different geometry
Solar and view angles change the reflectance signal seen by each platform.
Different scale
Ground sampling distance and aggregation change what a pixel represents ecologically.
The problem
Why raw reflectance is not enough¶
Integrating information from NEON imaging spectroscopy, drone systems such as MicaSense, and moderate-resolution satellites such as Landsat or Sentinel requires more than putting the files in one folder. Each platform encodes a different mix of spectral response, illumination geometry, viewing geometry, scaling, masking, and spatial support.
Those differences matter when you want to:
- validate satellite products using airborne data
- relate drone measurements to Landsat observations through a NEON airborne reference
- build cross-scale ecological models
- interpret reflectance change through time or across terrain
Two operations
Correcting and harmonizing are different steps¶
Physical corrections
SpectralBridge applies topographic and BRDF correction to reduce variation caused by terrain, illumination, and viewing geometry.
The result is a reflectance product that is more comparable across acquisition conditions.
Sensor harmonization
The NEON route applies configured sampled spectral weights to corrected hyperspectral spectra to construct sensor-like band representations. Recorded weight definitions and provenance must be reviewed; packaged centers and widths are not themselves measured sensor response curves.
Optional brightness adjustments are documented in the QA outputs rather than applied opaquely.
Why NEON
Why NEON is the core airborne foundation¶
NEON AOP directional reflectance products provide the ingredients the workflow needs to stay explicit and reproducible:
Rich spectral detail
NEON provides dense hyperspectral sampling with wavelength metadata that can be translated into target sensor bandspaces.
Per-pixel ancillary context
Solar geometry, sensor geometry, slope, and aspect travel with the reflectance cube, which is essential for correction and QA.
File-based reproducibility
The pipeline turns each flight line into restart-safe ENVI, parquet, JSON, and QA outputs that can be audited later.
Cross-scale bridge
NEON can provide a detailed spectral reference for a study that also includes drone and Landsat observations. Each observation retains its own acquisition and spatial support.
Common framework, sensor-specific treatment. A study may use NEON as a spectral reference for drone and Landsat comparisons. This is a study design, not a physical conversion chain or proof of sensor equivalence. Read the measurement concepts before the technical details.
Caveats
What still requires scientific judgment¶
SpectralBridge improves comparability, but it does not remove every source of uncertainty. Even after correction and harmonization, you still need to think carefully about:
- residual BRDF effects and unstable coefficients in low-signal regions
- atmospheric differences between sensors and acquisitions
- snow, smoke, water, shadows, and other masking edge cases
- scale mismatch between airborne, drone, and satellite footprints
- the quality flags and masking conventions used downstream
The package favors transparency over hidden convenience. Major steps write sidecars and QA artifacts so assumptions remain reviewable.
Where to go next