Global Evaluation of the Ecosystem Demography Model (ED v3.0)

Abstract

Abstract. Terrestrial ecosystems play a critical role in the global carbon cycle but have highly uncertain future dynamics. Ecosystem modelling that includes the scaling-up of underlying mechanistic ecological processes has the potential to improve the accuracy of future projections, while retaining key process-level detail. Over the past two decades, multiple modelling advances have been made to meet this challenge, including the Ecosystem Demography (ED) model and its derivatives including ED2 and FATES. Here, we present the global evaluation of the Ecosystem Demography model (ED v3.0), which likes its predecessors features the formal scaling of physiological processes of individual-based vegetation dynamics to ecosystem scales, together with integrated submodules of soil biogeochemistry and soil hydrology, while retaining explicit tracking of vegetation 3-D structure. This new version builds on previous versions and provides the first global calibration and evaluation, global tracking of the effects of climate and land-use change on vegetation 3-D structure, new spin-up process and input datasets, as well as numerous other advances. Model evaluation was performed with respect to a set of important benchmarking datasets, and model estimates were within observational constraints for multiple key variables including: (i) global patterns of dominant plant functional types (broadleaf vs evergreen); (ii) spatial distribution, seasonal cycle, and interannual trends of global Gross Primary Production (GPP); (iii) global interannual variability of Net Biome Production (NBP); and (iv) global patterns of vertical structure including leaf area and canopy height. With this global model version, it is now possible to simulate vegetation dynamics from local to global scales and from seconds to centuries, with a consistent mechanistic modelling framework amendable to data from multiple traditional and new remote sensing sources, including lidar.

Global evaluation of the Ecosystem Demography model (ED v3.0)

“…Code and data availability. All model simulation and source scripts can be found in https://doi.org/10.5281/zenodo.5236771 (Ma et al, 2021a). All benchmarking datasets are cited and publicly available.…”

Feasibility of Formulating Ecosystem Biogeochemical Models From Established Physical Rules

“…The ideas above are meant to be applied to individual microbes and plants (which can be representative of a population), forming the foundation to model population and community dynamics (e.g., population demography, community and ecosystem assembly). These individual-based formulations, when combined in "ecosystem demography" models (Koven et al, 2020;Ma et al, 2022;Medvigy et al, 2019), will contribute to better modeling of biogeochemistry-climate feedbacks regulated by plant and microbial physiology (a knowledge gap in existing models).…”

A machine learning approach targeting parameter estimation for plant functional type coexistence modeling using ELM-FATES (v2.0)

“…XGBoost has been successfully applied in a variety of fields within Earth and Environmental Sciences, such as urban temperature emulation (Zheng et al, 2021b), wildfire burned area (Wang et al, 2021), and emissions prediction (Wang et al, 2022), flash flood risk assessment (Ma et al, 2021), and aerosol property estimation (Zheng et al, 2021a, c).…”

“…However, explicit simulations of individual plants with stochastic processes suffer a substantial computational penalty and limit applicability over large or global scales (Fisher et al, 2018). To capture sufficient ecosystem dynamics and maintain relatively high computational efficiency, "cohort-based" models have been proposed (Haverd et al, 2013;Medvigy et al, 2009;Ma et al, 2021;Moorcroft et al, 2001;Longo et al, 2019). In a cohort-based approach, individual plants are grouped together as "cohorts" based on their similar properties, including size, age, and PFT (Fisher et al, 2018).…”