Inverse modelling of New Zealand's carbon dioxide balance estimates a larger than expected carbon sink

Abstract. Accurate national scale greenhouse gas source and sink estimates are essential to track climate mitigation efforts. Inverse models can complement inventory-based approaches for emissions reporting by providing independent estimates underpinned by atmospheric measurements, yet few nations have developed this capability for carbon dioxide (CO₂). We present results from a decade-long (2011–2020) national inverse modelling study for New Zealand, which suggests a persistent carbon sink in New Zealand’s terrestrial biosphere (-171 ± 29 Tg CO₂ yr⁻¹). This sink is larger than expected from either New Zealand’s Greenhouse Gas Inventory (-24 Tg CO₂ yr⁻¹) or prior terrestrial biosphere model estimates (-118 ± 22 Tg CO₂ yr⁻¹, Biome-BGCMuSo and CenW). The largest differences are in New Zealand’s South Island, in regions dominated by mature indigenous forests, generally considered to be near equilibrium, and certain grazed pasture regions. Relative to prior estimates, the inversion points to a reduced net CO₂ flux to the atmosphere during the autumn/winter period. The overall findings of this study are robust with respect to extensive tests to assess the potential biases in the inverse model due to transport error, prior biosphere, ocean and fossil-fuel estimates, background CO₂ and diurnal cycles. We have identified CO₂ exchange processes that could contribute to the gap between the inverse, prior and inventory estimates, but the magnitude of the fluxes from these processes cannot entirely explain the differences. Further work to identify the cause for the gap is essential to understand the implications of this finding for New Zealand’s inventory and climate mitigation strategies.