Spatiotemporal reconstructions of global CO₂‐fluxes using Gaussian Markov random fields

Abstract

Atmospheric inverse modeling is a method for reconstructing historical fluxes of green-house gas between land and atmosphere, using observed atmospheric concentrations and an atmospheric tracer transport model. The small number of observed atmospheric concentrations in relation to the number of unknown flux components makes the inverse problem ill-conditioned, and assumptions on the fluxes are needed to constrain the solution. A common practice is to model the fluxes using latent Gaussian fields with a mean structure based on estimated fluxes from combinations of process modeling (natural fluxes) and statistical bookkeeping (anthropogenic emissions). Here, we reconstruct global CO 2 flux fields by modeling fluxes using Gaussian Markov random fields (GMRFs), resulting in a flexible and computational beneficial model with a Matérn-like spatial covariance and a temporal covariance arriving from an autoregressive model in time domain. In contrast to previous inversions, the flux is defined on a spatially continuous domain, and the traditionally discrete flux representation is replaced by integrated fluxes at the resolution specified by the transport model. This formulation removes aggregation errors in the flux covariance, due to the traditional representation of area integrals by fluxes at discrete points, and provides a model closer resembling real-life space-time continuous fluxes.

Bayesian spatio-temporal inference of trace gas emissions using an integrated nested Laplacian approximation and Gaussian Markov random fields

“…A potential extension to this work is global-scale modelling, for example a global study of methane emission from global satellite measurements. Spatio-temporal estimation of global CO 2 fluxes using Gaussian Markov random fields and a global measurement network gives promising results (Dahlén et al, 2019). The nature of the method makes it well-suited to parallel implementation and thus upscaling.…”

WOMBAT: A fully Bayesian global flux-inversion framework

“…, r t . This, in turn, suggests that a suitable model for α is the vector-autoregressive process, similar to that used by Peters et al (2005) and Dahlén et al (2020),…”

WOMBAT v1.0: A fully Bayesian global flux-inversion framework

“…, r t . This, in turn, suggests that a suitable model for α is the vector-autoregressive process, similar to that used by Peters et al (2005) and Dahlén et al (2020). Specifically, α k+1 = M(κ)α k + w k , for k = 1, .…”