Effective and efficient data management throughout the research process is essential for maximizing data utility both during and beyond the research project. In data-intensive research, reducing barriers to access for data and associated documentation increases the efficiency of research and the impact of research data. The CI-Data (Data Management Platform and Adapter Development Component of the WC-WAVE project) has made significant progress in lowering barriers through the development of a loosely-coupled data access and ingest system that automates the process of acquiring the data to execute a watershed-scale model (iSNOBAL), running the model, and transferring the model outputs to a web-accessible data management, discovery, and access platform -- the Virtual Watershed Platform (VWP) developed by the project team.
The VWP provides a robust shared system for data management, discovery, and access. It can be used by any authorized computer that supports interaction through standard web protocols and open standards. By using a custom adapter that communicates with the VWP's web services, any model may obtain data from the VWP to execute a model run and push model data and associated documentation back into the VWP for near-real-rime doscover and access by other models, visualization systems (such as those being developed by the WC-WAVE CI-Vis Development team), and external data network (such as the DataONE and Data.gov systems).
By minimizing the effort required from researches, the data management loop is shortened to the least time possible. This yields more efficient and impactful data-intensive research.
Automating the data acquisition, uploading, and documentation process of the Virtual Watershed Platform (VWP) minimizes the data management effort of modelers. As a result, the time it takes for the products of a given model to be available for future research (subsequent model runs, visualization of results, sharing with collaborators) can be reduced from days or weeks to minutes, limited only by the time it takes for the data to be transferred between the VWP and the modeling system.