A challenge in plant ecophysiology has been quantifying photosynthesis across scales of biological organization. Recent advances in imaging spectroscopy of solar-induced fluorescence (SIF)—a physical flux linked to the machinery of the light reactions—may enable multiscale estimations of photosynthesis in natural systems. We performed an experiment to quantify changes in SIF due to the Kautsky effect in a potato canopy that was exposed to natural light. We use these data as a positive control to determine whether the SIF spectrum can be observed using a high-resolution imaging spectrometer, and to identify the signal-to-noise ratio (SNR) requirements for quantifying SIF.
Results/Conclusions
All estimates of SIF contain systematic error and random error. We removed systematic error by developing calibrations for spectral stray light and spatial and spectral non-uniformity. We found that systematic error due to spectral stray light is low (0.26% for a broadband LED light source), but still significant because the SIF signal is small (generally < 5% of vegetation reflectance). This underscores the importance of removing systematic error and producing measurements with high SNR. Binning measurements over gradients of space and time can dramatically increase SNR. We demonstrate that binning facilitates extraction of the SIF spectrum, but inherently involves tradeoffs with spatial or temporal resolution.