Building Capacity in MorphoSource through state-of-art, flexible data storage protocols for broader and more sustainable adoption by museums and other mass-data producers.

The ability to acquire high-quality data is a central challenge to advancing scientific knowledge about biodiversity and natural history. Without such data it is impossible to run analyses that generate meaningful results from which to draw conclusions. MorphoSource, a repository founded by researchers at Duke University in 2013 has begun to revolutionize how researchers and the public can discover, access, interact with and learn from the museum specimens constituting the world’s primary evidence of (or data on) natural history and biodiversity.

Cited for data access or archiving in 1,500 papers at the end of 2022, MorphoSource is unique and unprecedented as a publicly accessible web data repository where museum curators, researchers, and scholars can upload, archive, and share 3D, 2D, and audiovisual image data representing physical specimens from museum or lab collections. Currently, MorphoSource hosts over 170,000 datasets regularly used for research, education, and general interest purposes.

MorphoSource currently relies on finite commitments to operate, a model that puts constraints on how much data can be hosted and does not adequately ensure the long-term availability of funds for continuing its operations. Therefore, as one part of a larger plan for long-term sustainability, the current project will build features that significantly reduce data storage costs, enable the use of more flexible storage technologies, and recover some costs associated with data deposition, curation, and storage.

By its core mission, MorphoSource is an open access repository and all of the data it hosts must always remain accessible to anyone without charge. However, it must ultimately balance demand for data deposition, curation, and storage by its contributors (the users who upload data) with available resources. This project will begin that process by developing site tools that charge contributors for access to certain advanced deposit/curation features and for data deposits based on file size.

As an important broader impact, this project will create innovative technologies for storing data in affordable ways that will be usable not just by MorphoSource but will also benefit other repository platforms and storage infrastructures. Further, results of this work will directly contribute to the continued availability of the 170,000 datasets presently in MorphoSource and future data to come, which will, in turn, provide benefits to individuals involved in research, education, art, and general personal edification where these data are essential.

To accomplish the goals of this project, MorphoSource’s current file-and-disk storage architecture will be upgraded to support object storage, a more scalable and efficient storage technology used by many cloud storage providers. As part of this upgrade, MorphoSource will be integrated with Duke University’s NSF-supported “CC*Storage Fast” platform (OAC-2232810) that supports storing and moving data between disk, object, and cold cloud storage tiers.

Repository data will be moved from disk storage to more affordable object storage. Using the FAST platform, predictive algorithms and storage tiering tools will be implemented to automatically place less frequently used data into much cheaper cold cloud storage. Recovery of operating costs from users will be accomplished through implementation of eCommerce features to improve and scale an existing practice of charging contributors a deposit fee for uploading data.

This will result in shift towards contributor-sponsored storage, while still allowing MorphoSource to subsidize certain datasets based on need. MorphoSource will also distill the feature support for contributor users into two kinds of accounts: a basic account, and more powerfully featured, advanced account. While all critical components of depositing and curating a dataset will be possible with the free, basic account, the advanced account will include optional site features useful for frequent site users or user groups who contribute large amounts of data.

The advanced contributor account will be associated with a modest annual fee. Advanced accounts will also be made available to organizational management teams without requiring payment by individuals on the team. Intended outcomes by the project end include significant reductions in MorphoSource storage costs, larger scales of data mobilized by contributors due to lower storage costs, more data becoming public over time, and increased interest from various institutional entities in joining a community consortium to help sustain its vision and promise.

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.