Elements: OpenGHM: Enhancing reproducibility in global hydrologic modeling

University Of New Hampshire, Durham NH

Investigators

Abstract

Water is central to many global challenges and opportunities, including food production, drinking water supplies, energy production, and transportation. Researchers often use global hydrologic models to study these large-scale water questions because these models are designed to simulate both natural land surface water resources and a wide range of human interactions with them. To better protect water supplies, food and energy production, and infrastructure, it is in the national interest to provide local, state, regional, and national stakeholders with reliable information needed to make informed decisions. The provision of such information requires global hydrologic models that run quickly on cutting edge computational platforms, make use of emerging large datasets, and are fully transparent in how results are obtained. The water resource research community currently uses Global Hydrologic Models (GHMs) by downloading large datasets, running model simulations on a local computer system, then uploading results to public repositories. This project aims to accelerate innovation in water resources research by removing the bottlenecks caused by downloading and uploading large amounts of data and by increasing training and collaboration opportunities so scientists can more rapidly test and develop hydrologic models that can be used and trusted by other research teams. These goals are achieved by building the OpenGHM platform, a computational system that leverages existing NSF-supported computing resources to enable scientists to directly access data, models, and surrounding workflows on efficient cloud-based storage systems. This project creates an innovative cyberinfrastructure (CI), called OpenGHM. The objective of this work is to solve challenges to accelerating and reproducing global hydrologic research, including studies in the fields of water provisioning, food and energy production, and transportation through river systems. This is achieved by overcoming the resource and technical roadblocks faced by scientists and educators who wish to use Global Hydrologic Models (GHMs), an important tool in these fields. While recent advances have made hydrologic science data more findable and accessible, there remain gaps in fully transparent, accessible, and reproducible integrated data-model workflows. This project deploys an integrated data-model CI composed of four parts, each of which is a contribution to the national CI on its own, and each of which leverages the advances of the other components. A web-based science gateway enables access for anyone with an internet connection, leveraging national high performance computing resources. The utility of these tools is tested with two novel science applications and one education use case; each of these is also paired with a reproducibility study. This work represents an important early stage of CI innovation required to achieve a longer-term goal of GHM community science that is reproducible with tools available to all scientists regardless of institutional computational resources. This award by the Office of Advanced Cyberinfrastructure is jointly supported by the Division of Earth Sciences and the Division of Research, Innovation, Synergies and Education in the Directorate for 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.