Global Mapping of Leaf Photosynthetic Capacity Using Remote Sensing Methods

The maximum rate of carboxylation (Vcmax) determines the leaf photosynthetic capacity and is an important parameter in Farquhar's enzyme-kinetic leaf photosynthesis model. In regional and global applications of Farquhar's model, Vcmax at 25C (Vcmax25) is generally held constant according to plant functional type (PFT), although it has been found to vary greatly within the same type, and between seasons. Here we explore ways to map Vcmax25 using remote sensing (RS) methods. One way is to derive Vcmax25 using satellite measurements of Sun-induced chlorophyll fluorescence (SIF). The total SIF measured from a canopy is first separated into two components emitted from sunlit and shaded leaf groups according to sun-target-sensor geometry. The sunlit SIF component is then converted into the average sunlit leaf photosynthetic rate, from which Vcmax25 is derived using a data assimilation technique in an ecosystem model. The second way is to derive Vcmax25 from leaf chlorophyll content (LCC) retrieved from satellite data. LCC is responsible for light harvesting for plant photosynthesis, and empirical data showed close relationships between LCC and Vcmax25. Using a two-step model inversion algorithm, a time series of global LCC maps are retrieved using MERIS data from 2003 to 2011 at 300 m resolution and 7-day intervals. A validation using 133 ground sites in 5 PFTs suggests that this product is reliable (r2=0.5, p<0.01, RMSE=9.17 μg cm-2 or mean error 23%). Using empirical relationships between LCC and Vcmax25 for various PFTs, this global LCC time series is converted into Vcmax25. Vcmax25 maps from SIF and LCC are then compared at the global scale (r2=0.40, p<0.001, RMSE=24.2 μmol m-2 s-1). They are also compared with a Vcmax map derived using an ecological optimality theory (EOT), which provides a basis for estimating the optimum Vcmax of leaves under given radiation and temperature conditions. Vcmax values derived from EOT at the mean growth temperature is highly correlated with those derived from SIF (r2=0.85, p<0.001, RMSE=11.6 μmol m-2 s-1) and from LCC (r2=0.76, p<0.001, RMSE=16.2 μmol m-2 s-1), giving higher confidence in SIF-derived Vcmax than in LCC-derived Vcmax. However, there are insufficient ground-based data to validate these RS-Vcmax results at coarse and moderate resolutions (9 km for SIF and 300 m for LCC).