Skip to content

Concepts

Why cross-sensor calibration?

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

SpectralBridge then integrates corrected hyperspectral spectra against published sensor response functions to translate the data into another sensor's bandspace.

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 serves as the translating intermediary between fine-scale field or drone measurements and coarser Landsat observations.

Drone → NEON → Landsat is the conceptual relationship. The package does not treat the drone workflow as a direct one-step Landsat conversion: it preserves drone-native outputs and builds corrected, sensor-harmonized NEON products as the inspectable cross-scale reference.

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

Continue into the workflow