IN037-12
Advances in CyberWater development — an open source framework for data and model integration
Tuesday, 15 December 2020: 09:03
Virtual
Daniel Luna1, Ranran Chen2, Yao Liang2 and Xu Liang3, (1)University of Pittsburgh, Department of Civil and Environmental Engineering, Pittsburgh, PA, United States, (2)Indiana University Purdue University Indianapolis, Department of Computer and Information Science, Indianapolis, IN, United States, (3)University of Pittsburgh, Civil and Environmental Engineering, Pittsburgh, PA, United States
Abstract:
To address scientific questions about health, resilience, and sustainability of water resources, which include hydrological, environmental, biological, atmospheric, and other geosciences that define the Earth system, researchers need to easily access various data sources and to effectively incorporate these data into heterogeneous models. At present, many of these models are used in isolation due to their specific and complex requirements in model configurations. CyberWater, a sustainable and easy-to-use Open Data and Open Modeling framework under development since 2018, has been designed to undertake such cyberinfrastructure challenges. Its Open Data architecture offers a common internal data model, which facilitates diverse data source integration. Moreover, its Open Modeling architecture enables models to be integrated into CyberWater via Model Agents. CyberWater adopts a graphical scientific workflow system (VisTrails), and offers generic tools, enabling users to incorporate their specific models into the CyberWater cyberinfrastructure with little or no coding, which significantly enhances their models’ usability by the water community and graduate students.
In this presentation, we report our recent development of CyberWater via a land surface model (LSM) of the Variability Infiltration Capacity (VIC) model and a high-resolution hydrological model of the Distributed Hydrology Soil Vegetation Model (DHSVM) that have been integrated into the CyberWater framework. The model integration is demonstrated by using both user’s developed Model Agents and the generated Model Agents from CyberWater’s Model Agent tools. We furthermore illustrate how users can quickly carry out hydrological model simulations for six different watersheds via building automatic workflows from accessing online data sources, to obtain model simulation results, and to compare results with observations in a user-friendly workflow environment with CyberWater. These functionalities of CyberWater are being thoroughly tested by the development team and the water community. The CyberWater is aimed at significantly enhance the accessibility, usability, and reproducibility of the various water models through an open data and modeling framework in cyberinfrastructure for broad earth science communities.