Q. How do I share my publications?

A. There are several ways to provide public access to journal articles, and the best option depends on how and where an article has been published. These include publishing an article as Open Access in a reputable journal or submitting an accepted manuscript to PubSpace. Please see ‘How to Share Publications’ in the SMD Open-Source Science Guidance for more information. Technical reports should be uploaded directly to the NASA STI Repository.

Q. How can researchers share their publications immediately at the time of publication if they are behind a publisher’s paywall?

A. Researchers can meet the public access requirements for publications through multiple methods. For journal articles in which the final published article has associated subscription or access fees (i.e., those that are not published Open Access), authors may comply with the requirement by submitting their copy of the final accepted manuscript to the NASA STI Repository via the PubSpace submission page.

Q. How should researchers budget for Open Access fees in proposal budgets, including the potential for some articles to be published after the award period has ended?

A. SMD encourages publishing articles as Open Access in reputable journals, and reasonable costs for Open Access publishing (including Gold Open Access) may be included in the budgets of proposals for SMD funding. Article processing charges vary widely across publishers, and SMD encourages proposers to budget for publishing in reputable journals that are most appropriate for their scientific discipline. Publication fees included in the proposal budget should correspond to the plan described for making publications publicly available included in the proposal’s OSDMP.

Note that Open Access publishing, which can be associated with high article processing charges, is not required to meet the publication sharing requirements of SPD-41a. For articles not published as Open Access, investigators may satisfy public access requirements by submitting the author’s copy of an accepted manuscript to the NASA STI Repository via the PubSpace submission page. This option may be useful for situations in which an article is published after an award period has ended. See ‘How to Share Publications’ in the SMD Open-Source Science Guidance for more information.

Q. It is common to present scientific results that are preliminary or unpublished at conferences. Does SPD-41a require these presentations to be made publicly available?

A. SPD-41a does not require the sharing of preliminary results. If preliminary results are included in a conference presentation, they may be redacted prior to sharing other parts of the presentation more widely. There are significant benefits to researchers to appropriately sharing their preliminary results publicly: it makes the scientific process more inclusive, transparent, and collaborative at all stages.

See ‘Sharing Materials from Science Events’ in the SMD Open-Source Science Guidance for more information.

Q. What Mission publications need to be made publicly accessible?

A. All SMD-funded publications describing Missions must be made publicly accessible via a NASA-designated repository, including technical reports, peer-reviewed publications, conference proceedings, dissertations, and books. This does not include any documentation that would be covered by exceptions such as export control or other sensitive information.

Examples of Mission publications that must be made publicly accessible include peer-reviewed manuscripts describing the detectors used in the mission, unrestricted design documents from mission reviews, and conference presentations on the mission status. This also includes the output from conferences or workshops that are primarily sponsored by a Mission. 

Research created from sub-awards or grants from Missions would follow the sections in SPD-41a related to research (Sections III and VII). For example, funding supplied for research from successful observations of the Hubble Space Telescope would follow the Research policy for access to publications once it was integrated into the relevant solicitations.

Q. How should Missions share publications? 

A. As it is important for Mission data to become publicly available as soon as possible, the information necessary to understand the Mission data should also be made available immediately upon publication. As such, we encourage Missions to make publications from the team that describe the Mission, data, or aspects of the Mission as open as possible.

For peer-reviewed journal articles, the methods described under ‘How to Share Publications’ in the SMD Open-Source Science Guidance apply to Mission publications. Other types of Mission publications, such as technical reports or design documents from mission reviews, should be submitted to the NASA STI Repository through STRIVES.

Q. SMD divisions provide their own templates for data management plans (DMPs) and Open Science and Data Management Plants (OSDMPs). These templates vary in aspects such as required metadata, which creates challenges for multidisciplinary data management and data interoperability. Does NASA plan to unify requirements for DMPs / OSDMPs?

A. The SMD Open-Source Science Guidance includes a general OSDMP template for SMD, which may be appropriate for cross-divisional research. SMD will continue to review OSDMP templates and work to ensure that requirements for OSDMPs do not hinder multidisciplinary research.

Q. Does NASA provide a repository to archive data generated by SMD-funded research?

A. NASA maintains numerous science data repositories to archive data arising from SMD-funded scientific activities. SMD-funded researchers should follow any guidance on selecting a data repository in their solicitation or relevant SMD division policies, which may require the use of a particular NASA archive. If the solicitation or SMD division do not require the use of a specific repository, data should be shared in the most appropriate repository for the scientific discipline, which may be a NASA-supported repository or a public, generalist repository. For more information on selecting a repository to archive and share data from SMD-funded research, see ‘Where to Share Data’ in the SMD Open-Source Science Guidance.

Q. If a SMD-funded researcher has already made their data publicly available via another federal agency’s repository, would this satisfy the requirement in SPD-41a to make data publicly available?

A. If the data have been shared via an appropriate repository that has been approved by another federal agency, this would satisfy SMD’s requirements for data sharing.

Q. Not all data are worth archiving or sharing with the public. Why doesn’t SMD  allow scientists to decide what data will be useful for the future and should therefore be archived and shared?

A. At a minimum, SMD requires that the scientific data underlying publications be archived and shared openly to allow for the validation and reproducibility of scientific results. The additional requirement for sharing other scientifically useful data at the end of the award period of performance supports the accessibility of SMD-funded research and enables future research activities to accelerate the pace of scientific progress. SMD recognizes that the definition of what data are ‘scientifically useful’ varies across disciplines and gives SMD divisions the responsibility to provide further guidance to their communities. SMD-funded researchers are also responsible for describing the scientifically useful data that will be generated by their research and plans for archiving and sharing these data in their OSDMPs.

Q. What are the requirements for sharing model and simulation data?

A. SPD-41a requires model and simulation data needed to validate scientific conclusions to be shared at the time of publication of a peer-reviewed manuscript that describes the scientific results. Proposers to ROSES must describe what types of data will be shared and how they will be shared as part of the OSDMP.

Further guidance on what types of model and simulation data should be shared and how these data should be shared is forthcoming and will be developed based on community input. SMD divisions may provide specific guidance on sharing model and simulation data that is relevant to specific disciplines.

Q. Validating the scientific conclusions in a publication can require a very large volume of simulation data, which require substantial storage space and are often archived in locations that are not publicly accessible, such as High-End Computing Systems. Will NASA fund the purchase of storage and provision access to these simulation data?

A. Proposers should include plans for sharing model and simulation data as part of the OSDMP. This should include a description of the model or simulation data would be scientifically useful to share.  SMD is currently working toward solutions to make the outputs of high-end computing and cloud computing more accessible, please see the Data and Computing Architecture Study for more information.  For ROSES23, proposers do not need to include the cost of storing the scientifically useful data beyond the period of performance of the award.

Q. Are datasets used to train AI/ML models within scope of SPD-41a?

A. Data used for training AI/ML models (or training datasets) used in SMD-funded scientific activities are within scope of the policy and fall under the same requirements for data archiving and sharing as other scientific data. Training data that would be necessary to validate the results of a scientific finding that is based on an AI/ML model are important to research reproducibility and must be shared at the time of publication of an associated manuscript or at the end of the period of performance of a research award. However, training datasets that are based on commercial datasets that prohibit public sharing and evaluation sets of training datasets required for independent objective validation of new models may be exempted from the policy.

Considerations and best practices for sharing training data will be developed and added to the SMD Open-Source Science Guidance. In the meantime, please consult your Program Officer or SMD division representative for further guidance on sharing training data.

Q. What does this policy mean for existing missions that have a period of exclusive use for targeted observations? 

A. SPD-41a applies to new missions; exclusive use periods on existing missions (e.g., Hubble, Chandra, JWST) will be discussed on a case-by-case basis. SPD-41a does not take precedence over existing international agreements for missions like the James Webb Space Telescope that set a period of exclusive use for targeted observations, and any changes to those agreements will take place in consultation with SMD’s partners. While SMD’s policy is that mission data will have no period of exclusive access, there may be a period, not longer than six months, to allow for calibration and validation of the data. This provides researchers time to validate the scientific quality of their targeted observations and experiments. SPD-41a also includes a process for policy variances when a longer period of time for calibration and validation is warranted for data from missions for exceptional reasons. 

Q. What license should I use for my data?

A. SMD scientific data should be released with a Creative Commons Zero license if there are no other restrictions on the data. This provides a clear license for the user that the scientific data is in the worldwide public domain and that they may use it freely. In some cases, there might be existing restrictions on releasing the data due to intellectual property rights, contract restrictions, underlying licenses, or other issues. If unsure, contact your counsel that can help with intellectual property rights or ask for clarifications at HQ-SMD-SPD41@mail.nasa.gov.

Q. SPD-41a requires that SMD-funded data be made available in non-proprietary, modifiable, and open formats. Will SMD continue to allow data to be made available in closed, industry-standard formats if they are also provided in open formats?

A. SMD-funded data must be made available in open formats as a minimum requirement. However, this requirement does not prevent data from being shared in additional, closed formats, such as industry-standard or proprietary formats. SMD encourages the sharing of data in multiple formats if this will increase the accessibility or reusability of the data.

Q. What are SMD’s expectations for interoperability of data?

A. The FAIR principles provide a general expectation that:

1.  Data and metadata use a formal, accessible, shared, and broadly applicable language for knowledge representation.
2.  Data and metadata use vocabularies that follow FAIR principles.
3.  Data and metadata include qualified references to other data.

Additional recommendations on interoperability will be developed within the SMD Open-Source Science Guidance.

Q. What steps is NASA taking to make data analysis environments, including cloud computing environments, more accessible to data users?

A. The SMD Data and Computing Architecture Study is an ongoing effort to identify the scientific data and computing capabilities and architectures needed to enable open science across SMD. The study is soliciting input from a broad set of stakeholders, and feedback obtained through this process will be used to identify opportunities for the long-term evolution and sustainability of SMD-wide data and computing infrastructure to enable open-source science.

Q. Why should I share my software?

A. Sharing scientific software helps support two main pillars of open science: reproducibility and building on the work of others. Publications can be incomplete in their description of the underlying algorithm. While this is rarely intentional, this can still be problematic in trying to reproduce the results. Accessibility to the software used can help clarify what was written and what was done. In sharing software, we also give the next generation of scientists a clear foundation on which to build. This saves time through re-use of the software or by building on top of existing software. Sharing your software also allows it to be cited by future users, which will increase the impact of your work.

Q. What should be included in a software management plan?

A. Please see ‘Software Management Plans’ in the SMD Open-Source Science Guidance for recommendations on the components of a software plan for SMD-funded research. Software management plans should reflect the practices of specific research communities, so please follow any additional guidance on format or content for software management plans provided by SMD divisions or funding solicitations.

Q. Will NASA provide a location to host software generated by NASA proposals?

A. Solicitations or SMD divisions may provide specific guidance on where to host software. If no specific guidance is provided, options include publishing source code in a software-specific journal, including code in the supplemental material of a corresponding publication, or using a community-approved public repository. Please see ‘Where to Share Software’ in the SMD Open-Source Science Guidance for additional recommendations on where to host software generated by SMD-funded research.

Q. If I share my source code openly on GitHub, does this meet the SPD-41a requirements for archiving and sharing software?

A. Sharing research software on a version-controlled platform, such as GitHub, is encouraged. However, GitHub, alone, does not satisfy SPD-41a requirements for archiving software in a publicly accessible repository or making the software citable with a persistent identifier. As one solution, GitHub has an integration with Zenodo to create an archived copy of a GitHub repository and a DOI for the archive. SMD-funded researchers are encouraged to archive their software in an appropriate repository to the best of their ability, though methods for archiving software will not be viewed as a weakness for proposals submitted to ROSES-2023.

Q. I’ve developed some software specifically for a paper, how should I share this? 

A. Single-use software that is written for specific instances, such as creating a plot or running analysis for a scientific paper, should be shared as part of the supplemental materials of the corresponding publication or in a repository that supplies a persistent identifier that can be cited in publication.

Q. If I archive my software in a public repository, does SMD have a requirement for how long the archived software must persist in the repository?

A. Repositories used to archive SMD-funded scientific information must be consistent with the NSTC Desirable Characteristics of Data Repositories for Federally Funded Research. Accordingly, the repository should provide documentation on policies for information retention and a plan for long-term organizational sustainability.

Q. After making my research software publicly available, am I required to provide ongoing maintenance, documentation, or responses to inquiries from future users of the software?  

A. There is no requirement for maintenance or documentation of research software after it is made publicly available. The policy requires that scientific software developed with SMD funding as part of an award be made openly available at the time of the peer-reviewed publication it supports or at the end of the period of performance of the research award. While there is benefit to maintaining and documenting software, no requirement was included in the policy due to the range of different types of software and to minimize the burden on researchers. Similarly, there is no requirement for the developers of the software to respond to inquiries from future users of the software.

SMD-funded software projects that are made publicly available must include a code of conduct and guidelines for how to make contributions. If a software project is complete and not taking any further contributions, this should be indicated in the guidelines for how to make contributions.

Q. The process to make software written by NASA civil servants open is extremely onerous and is generally antithetical to open science principles. For example, it can take months or years to release research code to GitHub, and the process must be repeated if the code is updated. What is SMD doing to fix this problem?

A. NASA is in the process of updating its policy on the Release of NASA Software (NPR 2210.1C). SMD contributed to this process to ensure that the updated policy will support the development of open-source software by NASA civil servants. The Chief Science Data Office will continue to work across NASA to improve the processes related to the release of open-source software.

Q. Some SMD-funded researchers use proprietary programming languages, such as IDL or MATLAB, to conduct their research because they were trained in these languages or because their work builds upon a body of existing scientific software that is not open source. Is this use of proprietary software languages still allowed under SPD-41a?

A. SMD-funded researchers may continue to use proprietary or commercial programming languages under SPD-41a, though the use of open-source software is encouraged. SMD-funded research software developed in a commercial language (a programming language that requires a license to compile or run software) must be shared in a similar manner as other types of software developed in open languages. Sharing source code is important for scientific reproducibility and reuse, even if the code is written in a commercial language. See the SMD Open-Source Science Guidance for recommendations on sharing software.

Q. I work on software that has great scientific utility within SMD, but it is written in a proprietary language. Is there support available to convert this software to open source?

A. The ROSES element F.8 Supplemental Open Source Software Awards provides supplementary funding to convert legacy software into modern code to be released under a generally accepted, open source license (e.g., Apache-2, BSD-2-clause, MIT). To be eligible for the F.8 supplement, proposers must have an existing SMD research award with at least 15 months remaining in its period of performance at the time of submission to F.8. If there is no existing award to supplement, the F.15 High Priority Open-Source Science program element provides additional support for the development of scientific open-source software.

Q. My software project includes code that I wrote, code written years ago by my former collaborators, and code from an online source that is no longer maintained. My software will only be useful to future users if I include all the code gathered from these various sources, but I’m not sure about permissions and control over the code written by others. How should I share my software?

A. Under SPD-41a, SMD encourages researchers to share software that was developed previously and not open source if enhancements were made as part of the SMD-funded work (see Section VII-C-ii). This was left as a recommendation, rather than a requirement, in SPD-41a to acknowledge that situations such as the one described here can complicate software sharing. Do the best that you can to share your software in a manner that is reproducible, which may include citing the original sources of code written by others (even if that code is from a source that is proprietary or no longer maintained).

Q. Another scientist previously developed code in a proprietary language for commercial use, and they shared that code with me to modify and use in a non-commercial, scientific application on the condition that I would not share the original source code. Now I am using the enhanced software in my SMD-funded research, which will correspond to a peer-reviewed publication. What software must I share to comply with the policy?

A. As in the previous question, SPD-41a does not require SMD-funded researchers to share enhancements to previously developed software (see Section VII-C-ii), though researchers are encouraged to share software projects to the extent that they are able.

Q. Do I need to archive input and configuration files?

A. If input and configuration files are needed to reproduce or validate any scientific results, they must be archived in a manner consistent with other scientific information. This can include in a repository that provides a persistent identifier or as supplementary material of the publication.