Scope and topics
Remote sensing of vegetation is increasingly moving beyond observing canopy greenness to assessing
plant function, photosynthetic activity, and environmental stress. Hyperspectral reflectance provides
detailed information on vegetation biochemical and structural properties, while solar-induced
chlorophyll fluorescence, or SIF, offers a more direct connection to photosynthetic energy allocation
and ecosystem productivity. The combined use of reflectance and SIF, therefore, has considerable
potential for agricultural monitoring, urban monitoring, vegetation stress detection, and estimation
of photosynthetic activity.
However, both signals are strongly affected by spatial and spectral mixing. Satellite and airborne
observations commonly integrate multiple crop types, plant species, soil backgrounds, shadows, and
non-vegetated surfaces within a single pixel. Even apparently uniform vegetation canopies contain
sunlit and shaded leaves, different canopy layers, and plants with contrasting physiological
conditions. The observed signal consequently represents a mixture of structural, biochemical, and
physiological components rather than a single homogeneous target.
This issue is particularly important for SIF. The measured fluorescence signal is weak and is controlled
not only by photosynthetic processes, but also by canopy structure, illumination, reabsorption,
scattering, viewing geometry, atmospheric conditions, and sensor characteristics. Reflectance
measurements can provide important complementary information on these factors, but mixed pixels
and variable canopy backgrounds also influence them. Joint interpretation of reflectance and SIF,
therefore, requires methods that can separate overlapping signal sources and account for their
different physical origins.
The proposed special session will bring together researchers working on SIF, imaging spectroscopy,
spectral unmixing, radiative-transfer modeling, and multi-sensor data fusion. Its central objective is
to examine how mixed-vegetation signals can be resolved across spatial scales and translated into
reliable information about vegetation properties and functions.
The session will emphasize the development and application of spectral and spatial unmixing
approaches for SIF and hyperspectral reflectance. It will cover linear and nonlinear unmixing,
physically constrained inversion, Bidirectional Reflectance Distribution Function (BRDF), machine
learning, spatial disaggregation, and hybrid methods that combine data-driven analysis with
radiative-transfer models. Particular attention will be given to approaches that use reflectance to
support the retrieval, interpretation, or downscaling of SIF.
The session will also address the integration of observations across spatial scales. Current SIF measurements range from leaf- and canopy-level instruments to airborne systems and coarse- resolution satellite products. Linking these observations remains challenging because changes in spatial resolution also change the measured signal’s composition. Contributions combining satellite
SIF with higher-resolution hyperspectral imagery, unmanned aerial vehicle observations, proximal sensing, thermal measurements, or LiDAR-derived canopy structure will therefore be especially relevant.
Applications may include separating vegetation and soil contributions, discriminating among crop or
vegetation types, distinguishing between sunlit and shaded canopy components, identifying subpixel
stress, and improving estimates of photosynthesis, gross primary productivity, and crop condition.
The session will also encourage critical discussion of uncertainty, validation requirements, and the
limits of what mixed SIF and reflectance observations can infer.
List of topics:
- Spectral and spatial unmixing of vegetation reflectance and SIF.
- Joint analysis of hyperspectral reflectance and chlorophyll fluorescence.
- Downscaling and spatial disaggregation of satellite SIF.
- Separation of vegetation, soil, shadow, and non-vegetated components.
- Unmixing of crop types, plant species, and physiological conditions.
- Sunlit and shaded canopy contributions to reflectance and fluorescence.
- Nonlinear mixing caused by canopy structure and multiple scattering.
- Physically constrained, machine-learning, and hybrid unmixing methods.
- Integration of unmixing algorithms with vegetation radiative-transfer models.
- Fusion of SIF with hyperspectral, thermal, LiDAR, and structural observations.
- Scale integration among proximal, UAV, airborne, and satellite measurements.
- Detection of drought, salinity, nutrient, and heat stress within mixed pixels.
- Uncertainty assessment, validation, and benchmarking of unmixing methods.
- Implications of mixed signals for estimating photosynthesis, productivity, and yield.
- Effects of bidirectional reflectance distribution function (BRDF) and illumination–viewing
geometry on reflectance, SIF interpretation, and unmixing.
Organizers
Dr. Oz Kira, Civil and Environmental Engineering, Ben Gurion University of the Negev.
Prof. Fadi Kizel, Civil and Environmental Engineering, Technion, Israel Institute of Technology