PHOTOSYNSAT, photosynthesis from space: Theoretical foundations of a satellite concept and validation from tower and spaceborne data

We develop herein the theoretical foundations for a new satellite concept, utilizing multi-angle, along track spectral measurements to infer photosynthesis and gross primary production, at the landscape level over time. We validate the theory using both tower and space-borne sensors. The concept, originated in Hall et al. (2008), and Hilker et al. (2008a) and is based on two principles: (1) The first derivative of the photochemical reflectance index (PRI) with respect to shadow fraction viewed by the sensor ∂PRI/∂αs, is proportional to light-use efficiency ε. (2) This behavior can be shown both theoretically and empirically to be independent of vegetation structure and optical properties. These two principles provide the basis for a robust photosynthesis algorithm that can be applied consistently both spatially and temporally. We develop the general theoretical concept using a canopy reflectance model that incorporates a dependence of leaf reflectance on illumination strength, permitting the leaf reflectance at 531 nm to depend on the intensity of photosynthetic down-regulation. Using this model we are able to show that using PRI alone to infer ε is confounded by the shadow fraction viewed by a sensor, the PRI value in a non-down-regulated physiological state, and the sunlit canopy reflectance. We are able to demonstrate that these difficulties are mitigated by using ∂PRI/∂αs—not PRI—as the primary measure of canopy level ε. We demonstrate our concept using tower and satellite data acquired over three years, in two distinct biomes and vegetation types to show that PRI/∂αs and ε are related by a single function. Building on these ideas we propose the development of a new satellite concept that can utilize a spatially and temporally robust algorithm to map photosynthesis at landscape scales and its temporal variation.