Data Management Plan

Types of Data

This project will compile, document and archive CyberTraining modules data with relevant associated tutorials, workflows, environmental datasets, metadata, and contextual datasets. Development of the required data storage and integration techniques, and processing methods/algorithms for enabling access to and reuse of the data by interested users is part of the intellectual contribution of this project. Development of training materials used in this project will be devoted to meeting these goals. The following are the types of materials and data that will be assembled:

Data and Metadata Standards

We plan to organize and share the data we assemble within the HydroShare data repository (see plans for archival below). HydroShare makes full use of existing and emerging standards for sharing environmental datasets. All HydroShare resources (i.e, datasets, models, scripts, etc.) are described with metadata that conform to the Dublin Core metadata standard (DCMI, 2012), conform to a data model that is an implementation of the Open Archives Initiative’s Object Reuse and Exchange (OAI-ORE) standard (Lagoze et al., 2008), and are stored on disk and packaged for download using the BagIt hierarchical file packaging specification (Boyko et al., 2012). These standards are well known within the library, information science, and digital archiving communities. HydroShare has also adopted standard file formats for the content files of known resource/data types. For example, HydroShare uses Version 2 of the Observations Data Model (ODM2) for time series data (Horsburgh et al., 2016), the Network Common Data Form (NetCDF) for multidimensional space/time datasets, ESRI shapefiles for vector geospatial data, and the GeoTIFF format for raster datasets. Part of the contribution of this project will be to extend HydroShare’s capabilities of storing data derived from environmental samples using ODM2 for archival in HydroShare. This combination of standard data formats, standardized metadata description, and standard packaging means that HydroShare resources are publishable and fully archivable.

Policies for Data and Research Products

The goal of assembling training materials and data in HydroShare is to promote collaboration around, sharing, and reuse of these data. As such, all of the data we assemble as a direct result of this project will be publicly, and freely available in HydroShare using a Creative Commons License, where applicable. HydroShare is a data repository and collaboration environment. We anticipate that by curating and sharing materials and data in HydroShare we will enable users to not only discover and access the original data, but also to create derivative products through collaborative analyses and research that can also be shared within HydroShare.

When creating derivative products, HydroShare provides users with the choice to create public or private resources and public or private collaboration groups, accessible only to selected users, within which these activities can take place. Groups of researchers may wish to share data, model instances, or simulation results derived from the data within their group before they are published externally. To support this, authentication and access control have been fully integrated within HydroShare, and users can already choose how to share resources with other users or the larger community. Our experience has been that collaborative activities may result in multiple intermediate research products, only some of which may be considered publishable by the researchers. As such, HydroShare has functionality for creating formal, tracked versions of resources. Users can choose the Creative Commons License under which their resources are shared, and HydroShare has already established the facilities required to formally publish data, models, and simulation results, enabling individual researchers to select and publish their results as they see fit. Formally published resources are made immutable and receive a citable digital object identifier (DOI). Access to private resources and private research groups is at the discretion of resource and group owners. Final research results can be made freely and publicly available when they are deemed publication ready by the author. These collaboration capabilities make HydroShare an ideal location for publishing CyberTraining materials and data in a way that they can be cited, reused, and formally linked to derivative products.

All HydroShare resources have a landing page the displays the resource’s metadata and contents, including attribution information (i.e., authors and contributors, funding agency credits, etc.) and a formal citation. HydroShare users must agree to a formal publication agreement prior to formally publishing a resource. This agreement was developed in collaboration with the Consortium of Universities for the Advancement of Hydrologic Science, Inc. (CUAHSI) and specifies the terms and conditions under which users can publish resources in HydroShare.

Plans for Archiving Data

HydroShare will serve as the archival system for results from this project. Curated research products (e.g., datasets, models, etc.) published in HydroShare are citable for use in peer-reviewed journal articles, conference presentations and proceedings, and other formal publications using a formal digital object identifier (DOI). HydroShare and all of its attendant systems are hosted on fault-tolerant, enterprise-class servers dedicated to the HydroShare system and housed in the Renaissance Computing Institute’s (RENCI’s) managed, climate controlled, UPS-backed information technology facility, ensuring the reliability of the HydroShare system. All source code developed as part of this project will be openly shared in GitHub repositories associated with the HydroShare project.

Protecting personal application data Personally identifiable information about applicants will be maintained on a protected server in password protected files and only accessible to researchers on the grant and the selection panel. Application information will be assigned a unique ID for each person. Data for selection and analysis will be de-identified before shared with analysts, the selection panel or others.