Collaborative Research: Frameworks: Building a Collaboration Infrastructure: CyberWater2 -- A Sustainable Data/Model Integration Framework

点击翻译按钮获取中文摘要

Natural hazards, such as coastal and inland flooding caused by Hurricanes and severe drought and its associated wildfire, have been occurring with unprecedented frequency, induced by climate changes that encompass hydrological, biological, environmental, atmospheric, ocean, and other geosciences. Such hazards have caused not only profound damages to our environment and required tremendous efforts to recover, but also cost people's lives. To mitigate these potential disasters, it is a critical time to tackle their associated scientific questions both fundamental and large-scale that impact on the health, resilience, and sustainability of the Earth system we live in. The problems are complex and multidisciplinary, and researchers and practitioners from diverse fields must work together to find solutions. By its nature, Earth system models are comprised of component models – from land surface, to rivers, coastal regions, ocean, sea ice, and atmosphere, where each component model is coupled with one another. As science advances, a component model or its subsystems may have to be replaced because of new understanding, or because different perspectives must be explored and tested for the credence of different combinations to find the most credible predictions for different conditions at different locations. Such tasks often require substantial efforts and time and can become a bottleneck. This project is aimed at developing a new open-source cyberinfrastructure framework, Cyberwater2, in which model coupling is shifted from the current "code-coupling" approach to a new "information coupling" approach, and can be configured without writing glue code. This minimizes the need to access and modify each participating model's original code, and removes a major obstacle for large-scale cross-institutional collaborations and scientific investigations across disciplines and geographic boundaries. CyberWater2 is designed for diverse research communities including water, climate, coastal, engineering, and beyond. With our framework, researchers can devote their collaborative energy on problem solving and exploration of new frontiers, while using CyberWater2 to effectively achieve two-way open model couplings across platforms, model parameter calibration, data assimilation, testing/validations/comparisons, etc.The goal of this project is to make it easier to conduct large scale collaboration on complex problems and solve them efficiently, accurately and in-depth by developing a cyberinfrastructure, CyberWatyer2, that (1) significantly eliminates "glue" coding for two-way couplings across heterogeneous computing platforms, disciplines, and organizations; (2) automates complex model calibration and facilitates data assimilation processes applicable to various models; (3) supports task-based and in-situ hybrid workflow for greatly improved efficiency on two-way coupling across heterogeneous platforms; (4) provides a CyberWater2 server and web service framework for users in addition to the standalone systems; (5) enables sustainable data access from diverse sources by automatically adapting data agents to the changes (e.g., API interfaces) made to external data sources by providers; and (6) enables automated resource planning with intelligent site recommendation for High Performance Computing (HPC)/Cloud access on demand to maximize users' benefits.This project is supported by the Office of Advanced Cyberinfrastructure in the Directorate for Computer & Information Science & Engineering and the Division of Earth Sciences in the Directorate of Geosciences.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.