Satellite retrievals of leaf chlorophyll and photosynthetic capacity for improved modeling of GPP

This study investigates the utility of in situ and satellite-based leaf chlorophyll (Chl) estimates for quantifying leaf photosynthetic capacity and for constraining model simulations of Gross Primary Productivity (GPP) over a corn field in Maryland, U.S.A. The maximum rate of carboxylation (Vmax) represents a key control on leaf photosynthesis within the widely employed C3 and C4 photosynthesis models proposed by Farquhar et al. (1980) and Collatz et al. (1992), respectively. A semi-mechanistic relationship between Vmax25 (Vmax normalized to 25 °C) and Chl is derived based on interlinkages between Vmax25, Rubisco enzyme kinetics, leaf nitrogen, and Chl reported in the experimental literature. The resulting linear Vmax25−Chl relationship is embedded within the photosynthesis scheme of the Community Land Model (CLM), thereby bypassing the use of fixed plant functional type (PFT) specific Vmax25 values. The effect of the updated parameterization on simulated carbon fluxes is tested over a corn field growing season using: (1) a detailed Chl time-series established on the basis of intensive field measurements and (2) Chl estimates derived from Landsat imagery using the REGularized canopy reFLECtance (REGFLEC) tool. Validations against flux tower observations demonstrate benefit of using Chl to parameterize Vmax25 to account for variations in nitrogen availability imposed by severe environmental conditions. The use of Vmax25 that varied seasonally as a function of satellite-based Chl, rather than a fixed PFT-specific value, significantly improved the agreement with observed tower fluxes with Pearson's correlation coefficient (r) increasing from 0.88 to 0.93 and the root-mean-square-deviation decreasing from 4.77 to 3.48 μmol m−2 s−1. The results support the use of Chl as a proxy for photosynthetic capacity using generalized relationships between Vmax25 and Chl, and advocate the potential of satellite retrieved Chl for constraining simulations of GPP in space and time.