We test the relationship between canopy photosynthesis and reflected near-infrared radiation from vegetation across a range of functional (photosynthetic pathway and capacity) and structural conditions (leaf area index, fraction of green and dead leaves, canopy height, reproductive stage, and leaf angle inclination), weather conditions, and years using a network of field sites from across central California. We based our analysis on direct measurements of canopy photosynthesis, with eddy covariance, and measurements of reflected near-infrared and red radiation from vegetation, with light-emitting diode sensors. And we interpreted the observed relationships between photosynthesis and reflected near-infrared radiation using simulations based on the multilayer, biophysical model, CanVeg.Measurements of reflected near-infrared radiation were highly correlated with measurements of canopy photosynthesis on half-hourly, daily, seasonal, annual, and decadal time scales across the wide range of function and structure and weather conditions. Slopes of the regression between canopy photosynthesis and reflected near-infrared radiation were greatest for the fertilized and irrigated C4corn crop, intermediate for the C3tules on nutrient-rich organic soil and nitrogen fixing alfalfa, and least for the native annual grasslands and oak savanna on nutrient-poor, mineral soils. Reflected near-infrared radiation from vegetation has several advantages over other remotely sensed vegetation indices that are used to infer canopy photosynthesis; it does not saturate at high leaf area indices, it is insensitive to the presence of dead legacy vegetation, the sensors are inexpensive, and the reflectance signal is strong. Hence, information on reflected near-infrared radiation from vegetation may have utility in monitoring carbon assimilation in carbon sequestration projects or on microsatellites orbiting Earth for precision agriculture application
