Towards consistent assessments of in situ radiometric measurements for the validation of fluorescence satellite missions

• DOI: 10.1016/j.rse.2022.112984
• Publication Date: 2022-03-19T03:53:21Z
• Authors: Bastian Buman; Andreas Hueni; Roberto Colombo; Sergio Cogliati; Marco Celesti; Tommaso Julitta; Andreas Burkart; Bastian Siegmann; Uwe Rascher; Matthias Drusch; Alexander Damm
• Keywords: info:eu-repo/classification/ddc/550; 1903 Computers in Earth Sciences; Soil Science; Geology; 04 agricultural and veterinary sciences; Bias; FLEX; FloX; Measurement variability; Spectral shift; Spectroradiometer; Sun-induced chlorophyll fluorescence; Uncertainty;; 10122 Institute of Geography; 13. Climate action; 0401 agriculture, forestry, and fisheries; 910 Geography & travel; Computers in Earth Sciences; 1111 Soil Science; 1907 Geology

The upcoming Fluorescence Explorer (FLEX) satellite mission aims to provide high quality radiometric measurements for subsequent retrieval of sun-induced chlorophyll fluorescence (SIF). The combination of SIF with other observations stemming from the FLEX/Sentinel-3 tandem mission holds the potential to assess complex ecosystem processes. The calibration and validation (cal/val) of these radiometric measurements and derived products are central but challenging components of the mission. This contribution outlines strategies for the assessment of in situ radiometric measurements and retrieved SIF. We demonstrate how in situ spectrometer measurements can be analysed in terms of radiometric, spectral and spatial uncertainties. The analysis of more than 200 k spectra yields an average bias between two radiometric measurements by two individual spectrometers of 8%, with a larger variability in measurements of downwelling radiance (25%) compared to upwelling radiance (6%). Spectral shifts in the spectrometer relevant for SIF retrievals are consistently below 1 spectral pixel (up to 0.75). Found spectral shifts appear to be mostly dependent on temperature (as measured by a temperature probe in the instrument). Retrieved SIF shows a low variability of 1.8% compared with a noise reduced SIF estimate based on APAR. A combination of airborne imaging and in situ non-imaging fluorescence spectroscopy highlights the importance of a homogenous sampling surface and holds the potential to further uncover SIF retrieval issues as here shown for early evening acquisitions. Our experiments clearly indicate the need for careful site selection, measurement protocols, as well as the need for harmonized processing. This work thus contributes to guiding cal/val activities for the upcoming FLEX mission.