Partitioning of Net Fluxes

Inferring two dependent variables (R_eco and GEP) from one observation (NEE) is an ill-posed problem; the same net flux can result from an indefinite number of combinations of R_eco and GEP if both are simultaneously occurring or have occurred over the temporal averaging interval used to describe NEE. Hence, additional constraints or information about flux processes are needed. Most flux partitioning strategies are based on the notion that only R_eco occurs at night in ecosystems dominated by C₃ and/or C₄ photosynthesis, while GEP is virtually zero [but not with CAM photosynthesis, San-José et~al. (2007)]. The challenge comes in extrapolating these nighttime R_eco measurements to daytime conditions to estimate GEP by difference using Eq. 9.1. These difficulties are compounded by the fact that nighttime flux measurements are often compromised by stable atmospheric conditions with insufficient turbulence to satisfy the assumptions of the eddy covariance measurement system. These observations must be filtered from the eddy covariance data record (Sect. 5.3), leaving incomplete information about R_eco and thereby GEP.