FLEXINVERT is a Bayesian inversion framework for optimizing surface-to-atmosphere fluxes of an atmospheric species (e.g. greenhouse gases and aerosols) using atmospheric measurements of this species and modelled atmospheric transport. It has been developed by Rona Thompson at the Norsk Institutt for Luftforskning (NILU).

FLEXINVERT is based on so-called Source Receptor Relationships (SRRs), which describe the relationship between changes in mixing ratios at a receptor point and changes in fluxes, calculated by the Lagrangian Particle Dispersion Model, FLEXPART. Fluxes may be optimized at any given temporal resolution and on a nested grid of variable spatial resolution.

About the code

FLEXINVERT is coded in Fortran90 and has been tested with the gfortran compiler and the Linux operating system and a makefile for gfortran is included. To run FLEXINVERT, the LAPACK and NetCDF libraries for Fortran must be installed. The current version of FLEXINVERT can be run with output from FLEXPART 9.2 (for other versions some modifications may be needed). For details about FLEXPART see here


FLEXINVERT may be used to estimate fluxes from continental to regional to local scale. It may be used for any atmospheric species as long as the chemistry (if there is any) can be described as a linear process.

FLEXINVERT has already been used to estimate fluxes of CH4, SF6, HFC-125, and HFC-134a, in Europe as part of the EU funded InGOS project, fluxes of CH4 in East Asia as part of the Norwegian Research Council (NFR) funded SOGG-EA project, and fluxes of CH4 and BC in the Arctic as part of the NFR funded SLICFONIA project.