WOMBAT: A fully Bayesian global flux-inversion framework
Michael Bertolacci, University of Wollongong, michael_bertolacci@uow.edu.au (Presenter)
Andrew Zammit-Mangion, University of Wollongong, azm@uow.edu.au
Jenny Fisher, University of Wollongong, jennyf@uow.edu.au
Ann Stavert, Commonwealth Scientific and Industrial Research Organisation (CSIRO), ann.stavert@csiro.au
Yi Cao, University of Wollongong, ycao@uow.edu.au
Matt Rigby, University of Bristol, matt.rigby@bristol.ac.uk
Noel Cressie, University of Wollongong, ncressie@uow.edu.au
The WOllongong Methodology for Bayesian Assimilation of Trace-gases (WOMBAT) is a fully Bayesian hierarchical statistical framework for the estimation of fluxes of trace greenhouse gases from flask, in situ, and remotely sensed data. It incorporates inventory-based flux basis functions and uses an atmospheric chemical transport model to compute corresponding mole-fraction basis functions offline. The unknown scaling factors for the basis functions are assigned a hierarchical prior that encodes their variances and covariance length scales, while the measurement process incorporates a spatio-temporal correlated-error term, bias correction, and measurement-error scaling factors. Full uncertainty quantification is achieved by estimating the unknown parameters using a Markov chain Monte Carlo (MCMC) scheme.
We show, in an observing system simulation experiment (OSSE), that accounting for bias correction and correlated errors is crucial when the data have this structure. Using the GEOS-Chem chemical transport model, we show that WOMBAT is able to obtain posterior means and uncertainties on non-fossil-fuel CO2 fluxes from Orbiting Carbon Observatory-2 (OCO-2) data. The inferred fluxes are comparable to those from the Model Intercomparison Project (MIP) reported in Crowell et al. (2019, Atmos. Chem. Phys., vol. 19). We also find that our predictions of out-of-sample retrievals obtained from the Total Column Carbon Observing Network are, for the most part, more accurate than those made by the MIP participants. Subsequent versions of OCO-2 and in situ datasets will be ingested into WOMBAT as they become available.
Poster: Poster_Bertolacci__60_25.pdf
Presentation Type: Poster
Session: 3.5c Flux estimates and atmospheric inversions from space-based GHG measurements
Session Date: Wednesday (6/16) 12:00 PM