Inferring non-steady-state terrestrial vegetation carbon turnover times from multi-decadal space-borne observations on global scale

The global biomass is determined by the vegetation turnover times (τ) and carbon fixation through photosynthesis. Vegetation turnover time is a central parameter that not only partially determines the terrestrial carbon sink but also the response of terrestrial vegetation to the future changes in climate. However, the change of magnitude, spatial patterns and uncertainties in τ as well as the sensitivity of these processes to climate change is not well understood due to lack of observations on global scale. In this study, we explore a new dataset of annual above-ground biomass (AGB) change from 1993 to 2018 from spaceborne scatterometer observations. Using the long-term, spatial-explicit global dynamic dataset, we investigated how τ change over almost three decades including the uncertainties. Previous estimations of τ under steady-state assumption can now be challenged acknowledging that terrestrial ecosystems are, for the most of cases, not in balance. In this study, we explore this new dataset to derive global maps of τ in non-steady-state for different periods of time. We used a non-steady-state carbon model in which the change of AGB is a function of Gross Primary Production (GPP) and τ (ΔAGB = α*GPP-AGB/ τ). The parameter α represents the percentage of incorporation of carbon from GPP to biomass. By exploring the AGB change in 5 to 10 years of time step, we were able to infer τ and α from the observations of AGB and GPP change by solving the linear equation. We show how τ changes after potential disturbances in the early 2000s in comparison to the previous decade. We also show the spatial distributions of α from the change of AGB. By accessing the change in biomass, τ and α as well as their associated uncertainties, we provide a comprehensive diagnostic on the vegetation dynamics and the potential response of biomass to disturbance and to climate change.