Methodology for the estimation of terrestrial net primary production from remotely sensed data

TLDR: In this paper, the authors used the remote sensing of crop growth to estimate continental net primary productivity (NPP) as well as its seasonal and spatial variations, assuming a decomposition of NPP into independent parameters such as incident solar radiation, radiation absorption efficiency by canopies, and conversion efficiency of absorbed radiation into organic dry matter.

Abstract: Kumar and Monteith's (1981) model for the remote sensing of crop growth has been used to estimate continental net primary productivity (NPP) as well as its seasonal and spatial variations. The model assumes a decomposition of NPP into independent parameters such as incident solar radiation (S0), radiation absorption efficiency by canopies (ƒ), and conversion efficiency of absorbed radiation into organic dry matter (e). The precision on some of the input parameters has been improved, compared to previous uses of this model at a global scale: remote sensing data used to derive ƒ have been calibrated, corrected of some atmospheric effects, and filtered; e has been considered as biome-dependent and derived from literature data. The resulting global NPP (approximatively 60 GtC per year) is within the range of values given in the literature. However, mean NPP estimates per biome do not agree with the literature (in particular, the estimation for tropical rain forests NPP is much lower and for cultivations much higher than field estimates), which results in zonal and seasonal variations of continental NPP giving more weight to the temperate northern hemisphere than to the equatorial zone.