Get in touch

Validation

6 min read

A measurement is only as useful as it is trustworthy. SIF is a faint signal, and a faint signal is easy to get wrong. For SIFcam data to support decisions in breeding trials, agronomy and research, every measurement has to be tied to absolute physical units and checked against instruments the scientific community already trusts.

Shoot dynamics group from Forschungszentrum Jülich setting up SIFcam on an Alta X drone in a field
The Shoot Dynamics group at Forschungszentrum Jülich setting up SIFcam on an Alta X drone for field imaging. FZJ validated SIFcam in its research workflows by comparing UAV-based fluorescence maps against established reference instruments.

Absolute measurements in physical units

SIFcam does not report an index or a relative score. It measures the light emitted by the canopy in absolute physical units, milliwatts per square metre per nanometre per steradian (mW m⁻² nm⁻¹ sr⁻¹), within the oxygen A-band. Reporting in physical units is what allows results to be compared across sites and instruments.

Reporting in absolute units demands exacting calibration. The European Space Agency's FLEX satellite mission, which will measure the same oxygen-band fluorescence from orbit, targets a retrieval accuracy of about 0.2 mW m⁻² nm⁻¹ sr⁻¹ [2]. Ground and airborne reference systems used to validate fluorescence imagers are held to comparably exacting standards.

Cross-validation against reference instruments

Internal calibration is necessary but not sufficient. The decisive test is whether SIFcam agrees with independent instruments measuring the same thing.

In a peer-reviewed UAV campaign over a mixed wheat and faba bean trial in Bonn, SIFcam was compared against two reference instruments on the same day [3]. Against the ground-based FloX point spectrometer, the community standard for continuous canopy SIF monitoring, the coefficient of determination is R² = 0.93 across 13 plots, with an RMSE of 0.18 mW m⁻² nm⁻¹ sr⁻¹. Against airborne HyPlant imagery, agreement reaches R² = 0.56. The lower figure is expected: HyPlant flies far higher than a drone, with metre-scale pixels that the centimetre-scale SIFcam map has to be aggregated to match.

Running the same dataset through our own SIFcam processing software gives results in line with the paper: against FloX, R² = 0.91 with a slope of 0.98 across the same 13 measurements. For HyPlant we compare 2 × 2 m blocks rather than single pixels, because at the scale of a single 0.5 m pixel, noise and small misalignments between the two maps dominate. With 2 × 2 m blocks the agreement is R² = 0.84 (0.68 pixel by pixel).

Ground reference

SIFcam vs. FloX

R²
0.91
Slope
0.98
RMSE
0.20
n
13
000.50.5111.51.5222.52.533FloX F₇₆₀SIFcam F₇₆₀
CanopyBare soilLinear fit1:1 line

Airborne reference

SIFcam vs. HyPlant

R²
0.84
Slope
1.15
RMSE
0.39
n
459
0011223344HyPlant F₇₆₀SIFcam F₇₆₀12510202 × 2 m blocks per hexagon
Linear fit1:1 line
The dataset of Salattna et al. (2026), the SIFcam flight of 13 June 2021 over the wheat and faba bean trial at Campus Klein-Altendorf, Bonn, evaluated with our own SIFcam processing software. Left: SIFcam F760 against 13 ground measurements with the FloX spectrometer (11 over canopy and 2 over bare soil); each SIFcam value is the mean of all pixels within the 1.5 m FloX footprint. Right: SIFcam against airborne HyPlant imagery. Because noise and small misalignments between the two maps dominate at the scale of single 0.5 m pixels, both maps are averaged to 2 × 2 m blocks before comparison; the colour of each hexagon shows how many blocks fall into it. Pixel by pixel, R² is 0.68. Blocks outside the plotted range (1 of 459) are included in the statistics. All values in mW m⁻² sr⁻¹ nm⁻¹. FloX and HyPlant reference measurements by Forschungszentrum Jülich, described in Salattna et al. (2026).

HyPlant is one of the world's leading airborne SIF imaging spectrometers, and it produced the first spatially resolved SIF maps of vegetation canopies [4]. Agreement with both a ground spectrometer and an airborne imager confirms that SIFcam is capturing genuine physiological signals of photosynthesis and stress, rather than instrument artefacts, and places its measurements within an established scientific lineage.

Anchored to the wider SIF science

Validation is not only about a single instrument. It is also about whether the quantity being measured is itself meaningful, and on that point the evidence base is broad. Fundamental reviews spanning decades of research have established SIF as a robust, quantitative proxy of photosynthetic efficiency and carbon uptake across a wide range of vegetation types and conditions [1][6].

The strongest external confirmation of the measurement's importance is that it has become the target of a dedicated satellite mission. The European Space Agency's FLEX mission, Earth Explorer 8, is designed to map chlorophyll fluorescence globally from space [5]. SIFcam measurements, delivered at centimetre resolution with traceable units and cross-checked against FloX and HyPlant, sit naturally within the calibration and validation infrastructure being built around missions of this kind. The same data that supports a breeding trial can also help connect ground observations to satellite-scale carbon and ecosystem models.

Reproducibility and traceability

Every SIFcam measurement carries the metadata needed to make it reproducible: the instrument identity, the calibration state, the atmospheric correction parameters and the retrieval settings used. This is what turns a set of images into comparable evidence. To keep that calibration state trustworthy, we recommend recalibrating SIFcam after every growing season.

A physiological signal at one time point is only meaningful in relation to earlier and later measurements, and that relation only holds if the measurements are traceable to the same physical reference. This chain, from a calibrated detector, to absolute physical units, to independent cross-validation, is what distinguishes a scientific instrument from a sensor.

References
  1. 1.Meroni, M., Rossini, M., Guanter, L., Alonso, L., Rascher, U., Colombo, R., & Moreno, J. (2009). Remote sensing of solar-induced chlorophyll fluorescence: Review of methods and applications. Remote Sensing of Environment, 113(10), 2037–2051. https://doi.org/10.1016/j.rse.2009.05.003
  2. 2.ESA (2015). Report for Mission Selection: FLEX. ESA SP-1330/2, European Space Agency, Noordwijk, The Netherlands.
  3. 3.Salattna, S., Elibol, A., Siegmann, B., Kneer, C., Scharr, H., Rascher, U., & Bendig, J. (2026). High Spatial Resolution Imaging of Solar-Induced Chlorophyll Fluorescence (SIF) From an Uncrewed Aerial Vehicle (UAV): Development and Evaluation of Image-Processing Workflows. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 19, 30350–30367. https://doi.org/10.1109/JSTARS.2026.3717082
  4. 4.Rascher, U., Alonso, L., Burkart, A., Cilia, C., Cogliati, S., Colombo, R., et al. (2015). Sun-induced fluorescence: a new probe of photosynthesis. First maps from the imaging spectrometer HyPlant. Global Change Biology, 21(12), 4673–4684. https://doi.org/10.1111/gcb.13017
  5. 5.Drusch, M., Moreno, J., Del Bello, U., Franco, R., Goulas, Y., Huth, A., et al. (2017). The FLuorescence EXplorer Mission Concept: ESA's Earth Explorer 8. IEEE Transactions on Geoscience and Remote Sensing, 55(3), 1273–1284. https://doi.org/10.1109/TGRS.2016.2621820
  6. 6.Mohammed, G. H., Colombo, R., Middleton, E. M., Rascher, U., van der Tol, C., Nedbal, L., et al. (2019). Remote sensing of solar-induced chlorophyll fluorescence (SIF) in vegetation: 50 years of progress. Remote Sensing of Environment, 231, 111177. https://doi.org/10.1016/j.rse.2019.04.030
Frequently asked
What is the FloX spectrometer?
+
FloX is a ground-based, continuously operating point spectrometer built by JB Hyperspectral Devices and widely used as a community standard for canopy SIF monitoring. Its high spectral resolution and controlled measurement geometry make it a primary reference for validating imaging SIF systems.
What is HyPlant?
+
HyPlant is a high-performance airborne imaging spectrometer for vegetation monitoring, developed by Forschungszentrum Jülich in cooperation with SPECIM Spectral Imaging Ltd. It produced the first spatially resolved maps of solar-induced fluorescence and is one of the main instruments used to bridge ground-based SIF measurements and satellite observations.
Why is the agreement with HyPlant lower than with FloX?
+
HyPlant is a much larger instrument, too heavy for a drone and carried by a manned aircraft, flying at 350 metres with a 1 by 0.5 metre pixel. The centimetre-scale SIFcam map therefore has to be aggregated to match it, and the two instruments use different retrieval methods. In a small experimental field, even a 2-pixel HyPlant misalignment (about 2 metres) blurs the comparison. The R² of 0.56 reflects those differences, not a failure of the measurement.
Can SIFcam data be compared across campaigns?
+
Within a single campaign, comparison is straightforward today. Across campaigns it is on our roadmap: reliable day-to-day and season-to-season comparability requires precise measurement of incoming sunlight, which is the focus of our current calibration work. SIFcam already records the per-measurement metadata that this kind of longitudinal comparison needs, so data captured now will fold into that workflow once the capability is fully validated.
Related reading