1Research Center for Global Change and Ecological Forecasting, School of Ecological and Environmental Sciences, East China Normal University, Shanghai 200241, China
2Ministry of Education Key Laboratory for Earth System Modelling, Department of Earth System Science, Tsinghua University, Beijing, 100084, China
3oint Center for Global Change Studies (JCGCS), Beijing, 100875, China
4Center for ecosystem science and society, Northern Arizona University, Arizona, Flagstaff, USA
1Research Center for Global Change and Ecological Forecasting, School of Ecological and Environmental Sciences, East China Normal University, Shanghai 200241, China
2Ministry of Education Key Laboratory for Earth System Modelling, Department of Earth System Science, Tsinghua University, Beijing, 100084, China
3oint Center for Global Change Studies (JCGCS), Beijing, 100875, China
4Center for ecosystem science and society, Northern Arizona University, Arizona, Flagstaff, USA
Received: 20 Mar 2020 – Discussion started: 27 Apr 2020
Abstract. The synchronous increase of model complexity and data volume in Earth system science challenges using observations to evaluate Earth system models (ESMs). The challenge mainly stems from the untraceable of model outputs, the lack of automatic algorithms, and the high computational costs. Here, we built up an online Traceability analysis system for Model Evaluation (TraceME), which is traceable, automatic and shareable. The TraceME (v1.0) can trace the structural uncertainty of simulated carbon (C) storage in the state-of-the-art ESMs into gross primary production (GPP), carbon use efficiency (CUE), baseline residence time and environmental scalars (temperature and precipitation). The cloud-based framework used in TraceME provides the scientific workflows and a shareable platform to achieve the automated analysis and distributed data storage to greatly improve the efficiency of model evaluation. Then, we set up a worker node in TraceME (v1.0) to store the data from Coupled Model Intercomparison Project (CMIP6), and submitted tasks through browser to analyze the uncertainties of CMIP6 models in the TraceME system. Overall, this new tool can greatly facilitate model evaluation to identify sources of model uncertainty and provide some new implications for the next generation of model evaluation.
The increase of model complexity and data volume challenges the evaluation of Earth system models (ESMs), which mainly stems from the untraceable, unautomatic, and high computational costs. Here, we built up an online Traceability analysis system for Model Evaluation (TraceME), which is traceable, automatic and shareable. The TraceME (v1.0) can trace the structural uncertainty of simulated carbon (C) storage in ESMs and provide some new implications for the next generation of model evaluation.
The increase of model complexity and data volume challenges the evaluation of Earth system...