Taking CyberWater one step further to solve the climate crisis

Natural hazards like hurricanes and wildfires are rising at an unprecedented frequency because of climate change. Researchers from diverse fields are coming together to mitigate these disasters – and potentially save lives.

The University of Pittsburgh will lead a four-year, $1.8 million, multi-institution and multidisciplinary collaborative project sponsored by the Cyberinfrastructure for Sustained Scientific Innovation (CSSI) program of the National Science Foundation to solve climate-related issues. Pitt’s share of the funding is approximately $1.08 million.

Xu Liang, principal investigator and professor of civil and environmental engineering at the University of Pittsburgh Swanson School of Engineering, and Jeen-Shang Lin, co-principal investigator and associate professor of civil and environmental engineering at Pitt, will collaborate with colleagues from other institutions on the project, known as CyberWater2.

CyberWater2 will tackle pressing scientific issues using Earth system models made from complex component models—from land surfaces to rivers to coastal regions to oceans to atmosphere—where each component model is coupled with one another.

“The goal of CyberWater2 is to lower the hurdle for researchers to couple their own models with other component models of Earth system models to investigate challenging issues via two-way model couplings. These problems are complex and multidisciplinary,” Liang said. “Researchers and practitioners from diverse fields must work together to find solutions.” 

CyberWater2 is aimed at developing a new open-source cyberinfrastructure framework, where two-way model coupling is shifted from the current code-coupling approach to a new information-coupling approach. The information-coupling approach can be configured without writing glue code—code used solely to integrate incompatible model codes together—to minimize the need to access and modify each participating model’s original code. This new approach allows for large-scale cross-institutional collaborations and scientific investigations across disciplines, computing platforms and geographic boundaries.

Because of the ability to collaborate across disciplines, CyberWater2 will be able to facilitate a comprehensive and in-depth analysis and prediction of extreme weather events, like floods and droughts, and their impacts on inland and coastal flooding, storm surges, and landslide risk post wildfire.

For users, CyberWater2 will be able to automate complex model parameter calibrations and facilitate data assimilation to various models, support task-based and in-situ hybrid workflows, provide a web service framework, enable sustainable data access from diverse sources by automatically adapting to changes made by external data sources providers, and enable automated resource planning with intelligent site recommendation for High Performance Computing (HPC)/Cloud access on demand to maximize users’ benefits. 

CyberWater2, the new system, will significantly boost the original CyberWater system into a new level of capability and effectiveness for broad collaborations. A $1.3 million NSF collaborative project which Liang has also led since 2019, Cyberwater was developed to significantly simplify complex data and modeling integration with little to no coding, support HPC on demand, support model system’s reproducibility, easily integrate other systems with CyberWater, and provide tools to facilitate data and model parameter files pre-processing.

“We want to build a collaboration infrastructure to make it easier to conduct large-scale scientific collaborations across heterogeneous computing platforms, disciplines, and organizations in order to solve complex problems,” Liang said. “This improved cyberinfrastructure will make it possible with efficiency and accuracy.” 

The project will combine expertise from a large, diverse team. Alongside Liang and Lin and their hydrology/engineering student researchers from Pitt, the team includes atmospheric scientists from the Pacific Northwest National Laboratory of the Department of Energy, computer scientists and cyber-infrastructure experts from Indiana University-Purdue University Indianapolis and Indiana University, as well as software engineers from the Ball State University and education and outreach experts from CUAHSI.