Performance of the Ecosystem Demography model (EDv2.2) in simulating gross primary production capacity and activity in a dryland study area

Authors: DASHTI, HAMID - Boise State University; PANDIT, KARUN - Boise State University; GLENN, NANCY - Boise State University; SHINNEMAN, DOUGLAS - Boise State University; Flerchinger, Gerald; HUDAK, ANDREW - Us Forest Service (FS); DE GRAAF, MARIE ANNE - Boise State University; FLORES, ALEJANDRO - Boise State University; USTIN, SUSAN - University Of California, Davis; ILANGAKOON, NAYANI - Boise State University; Fellows, Aaron

Submitted to: Agricultural and Forest Meteorology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/28/2020
Publication Date: 12/15/2020
Citation: Dashti, H., Pandit, K., Glenn, N., Shinneman, D., Flerchinger, G.N., Hudak, A., De Graaf, M., Flores, A., Ustin, S., Ilangakoon, N., Fellows, A.W. 2020. Performance of the Ecosystem Demography model (EDv2.2) in simulating gross primary production capacity and activity in a dryland study area. Agricultural and Forest Meteorology. 297. Article 108270. https://doi.org/10.1016/j.agrformet.2020.108270.
DOI: https://doi.org/10.1016/j.agrformet.2020.108270

Interpretive Summary: Dryland ecosystems play an important role in the global carbon cycle, including regulating the global inter-annual carbon sink. Ecosystem models are important tools for modeling ecosystem dynamics, but they often struggle to reproduce changes in plant productivity along meteorological gradients, such as those found with changes in elevation. The aim of this study is to enhance model representation of ecosystem processes in drylands to facilitate improved understanding of elevation-dependent plant productivity, which is an important component of the carbon cycle. Our results showed good performance of Ecosystem Demography dynamic global vegetation model at simulating monthly plant producitivity at lower elevations, but model performance degraded at higher elevations having greater productivity. Results suggest that more diversity (e.g., more plant functional types) and modifying processes for simulating plant phenology may improve simulation of climate effects on carbon cycling.

Technical Abstract: Dryland ecosystems play an important role in the global carbon cycle, including regulating the global inter-annual carbon sink. Dynamic global vegetation models play an important role in improving our understanding of carbon cycle in different ecosystems. Currently, there is a poor understanding of the performance of these models in drylands due to rapid changes of carbon cycle components derived by variable hydrometeorological conditions (e.g. along an elevation gradient). The aim of this study is to enhance model representation of ecosystem processes in drylands to facilitate improved understanding of elevation-dependent gross primary production (GPP) as one of the important components of the carbon cycle. We performed sensitivity analysis and calibrated the Ecosystem Demography (ED.v2.2) dynamic global vegetation model to simulate GPP over an elevation gradient in a drylands watershed of the western US. GPP capacity and activity were investigated by comparing model simulations with GPP estimated from eddy covariance data and remote sensing products. Time series analysis was based on Bayesian model averaging for trend analysis and cross-correlogram spectral matching for phenometrics (start/end of the season) retrieval. Our results show good performance of EDv2.2 at monthly timestamps (RMSE˜0.38 [kgC/m^2/year]) between simulated and measured GPP in lower elevations. Moreover, remote sensing analysis showed that EDv2.2 captures the long-term trend in this ecosystem, however it doesn’t perform well in capturing phenometrics. The performance of the model degrades in higher elevations with greater GPP which requires introducing more diversity (e.g., more plant functional types) and modifying plant processes (e.g., plant hydraulics and phenology) to improve the model performance.