Prospects for chlorophyll fluorescence remote sensing from the Orbiting Carbon Observatory-2

Photosynthetic uptake of carbon dioxide by terrestrial vegetation is driven by absorbed photosynthetically active radiation (APAR) in the 400–700 nm wavelength region. The efficiency with which photons are used for actual photosynthesis varies because absorbed energy can also be dissipated into heat (non-photochemical quenching, NPQ). A small fraction is also re-radiated at longer wavelengths (660–800 nm), which is termed fluorescence and has been widely used in photosynthesis research for decades (e.g. Baker, 2008, Campbell et al., 2008, Corp et al., 2006, Genty et al., 1989, Krause and Weis, 1984, Moya et al., 2004). Active pulse-modulated fluorometry allows a direct quantification of photosynthetic efficiency by measuring fluorescence at steady state as well as upon illumination with a saturating light pulse. Even though this technique cannot be directly applied from space, steady-state sun-induced chlorophyll fluorescence (SIF) also offers the potential to place constraints on gross primary production (GPP) and thereby gain a more mechanistic understanding of ecosystem carbon exchange (Frankenberg, Fisher et al., 2011).

Recently, Joiner et al. (2011), Frankenberg, Butz, and Toon (2011), Frankenberg, Fisher, et al. (2011), and Guanter et al. (2012) showed that SIF can be retrieved based on in-filling of solar absorption features, so called Fraunhofer lines. The first real retrievals were performed using high-resolution spectra of the oxygen A-band region from the Tanso Fourier Transform Spectrometer onboard of the Japanese GOSAT satellite (Hamazaki et al., 2005, Kuze et al., 2009). Later, Joiner et al. (2012) showed that even moderate spectral resolution can be used to infer SIF using either a very broad Fraunhofer line at 866 nm (using SCIAMACHY) or applying a larger spectral range from 715 to 780 nm using GOME-2 (Joiner et al., 2013). In general, there is a trade-off between spectral resolution and spatio-temporal sampling. On the one hand, high spectral resolution enables more robust and accurate SIF retrievals but on the other hand it sacrifices spatial sampling. GOSAT retrievals (Frankenberg, Butz and Toon, 2011, Frankenberg, Fisher, et al., 2011, Guanter et al., 2012, Joiner et al., 2012, Joiner et al., 2011), for instance, are well characterized but cannot provide the spatial mapping of GOME-2. Neither will OCO-2 provide full spatial mapping even though it records about 8 times more spectra than GOME-2. Despite these limitations, GOSAT SIF retrievals have already been proven very useful in the analysis of carbon exchange in the tropical Amazon (Lee et al., 2013, Parazoo et al., 2013).

The Orbiting Carbon Observatory-2 (OCO-2), to be launched in July 2014, can alleviate some of the drawbacks of current SIF observations and combine the advantages of both GOSAT and GOME-2. It will enable SIF retrievals similar to GOSAT but will acquire 24 spectra per second instead of one every 4 s. This almost 100-fold increase in data density compared to GOSAT is combined with much smaller ground-pixels (1.3 ×ばつ 2.25 km² instead of a circular footprint with 10 km diameter for GOSAT or 80 ×ばつ 40 km² for GOME-2). As such, it has the potential to largely improve current SIF retrievals as well as bringing together scientific communities focused on top-down net ecosystem exchange inversions based on atmospheric CO₂ measurements with those using bottom-up estimates derived from remotely sensed vegetation parameters. Here, we will quantify the potential of OCO-2 using both retrievals based on a synthetic full repeat cycle of OCO-2 data as well as real thermal vacuum test data. The main purpose of this publication is to make potential users aware of datasets for vegetation-specific applications that can be expected from the OCO-2. The primary mission objective of OCO-2 is to measure atmospheric CO₂ with the precision and accuracy needed for source/sink inversions on regional scales. Thus, the current user community is composed of atmospheric inversion experts and global carbon cycle modelers. The new fluorescence product might be of interest to different scientific communities and this paper tries to bridge this gap as most previous publications on the OCO-2 focused on CO₂ only and scientists might be unaware of the potential of OCO-2 fluorescence retrievals and their applications in photosynthesis research and carbon cycle science. The use of atmospheric CO₂ in conjunction with fluorescence has just started (Parazoo et al., 2013) and we anticipate more synergistic applications in the future.