Open Science in Practice Software Sharing and Citation in Earth Science and Astronomy Journals Sam Teplitzky Open Science Librarian and Liaison to Astronomy Department, UC Berkeley
[email protected]
Outline ● Establishing Data + Code Sharing behavior in a new journal ● Measuring Progress ● Comparing disciplines ● Reproducing Results
Establishing Data + Code Sharing behavior in a new journal
What is Seismica? ●A community-driven, open-access journal for peer-reviewed research in seismology & earthquake science ●Free to publish, free to read Since 2022! What does Seismica publish? ●Research articles ●Reports: ○Fast reports (4-week decision) ○Null results ○Instrument Deployment / Field campaign reports ○Software reports ●Opinion articles and reviews ACCESSIBLE TRANSPARENT RESPECTFUL CREDIBLE PROGRESSIVE
Data & code availability and reproducibility Authors are expected to act in the spirit of Open Science and make their data/codes available when publishing with Seismica. ●Seismica Zenodo community ●Figshare ●Dryad ●GFZ Data Services ●ISC Dataset Repository Institutional services, personal websites, and GitHub typically do not meet the standards for long-term, DOI citable archiving. https://seismica.library.mcgill.ca/policies Data & Code Policies
Data & Code Practicalities Positives ● General acceptance of data sharing ● Many examples of data re-use and citation ● Heavy use of: ○ Zenodo ○ GitHub ○ Disciplinary repositories + seismic networks ● Open source software packages - ObsPy, PyRaysum Challenges ● Limited Use of Seismica Zenodo Community ● Different practices around the world ● Early career/novice concerns related to code sharing ● All the things! ○ Narrative data statements ○ Citations ○ Acknowledgments ○ Need for RORs
Measuring Data + Code Sharing Progress
Data and Code Availability Statement (DCAS) Features Importance Feature Essential Was a Data and Code Availability Statement included? Essential Were data generated in the course of the study? (Must be noted when present) Essential Were data cited or acknowledged within DCAS? (Must be noted when present) Essential Were code or software cited or acknowledged within DCAS? (Must be noted when present) Optional If a DOI-generating repository was cited, was it a: -General repository -Disciplinary repository -Government repository or site -Institutional repository Optional Was a seismic network or station cited? Discouraged Are data or code cited with URL but not persistent id? Discouraged Does author request to be contacted for access? Discouraged Are third party restrictions noted? Discouraged Is there an embargo on data? Noted Are data or code published in a supplement? Noted Are there privacy or ethics related restrictions? Teplitzky 2025 10.26443/seismica.v4i1.1561
Teplitzky 2025 10.26443/seismica.v4i1.1561 Adherence to Data and Code Policy features
2020 Study results using AAS XML Bouquin, D. R., Chivvis, D. A., Henneken, E., Lockhart, K., Muench, A., & Koch, J. (2020). Credit lost: Two decades of software citation in astronomy. The Astrophysical Journal Supplement Series, 249(1), 8. DOI 10.3847/1538-4365/ab7be6
Software mentions in AAS Journals 1995-2025
Astropy mentions by journal 2018-2025 (*using ADS API script adapted from 2020)
Using Astrophysics Source Code Library - ASCL.net There are 3817 codes in the ASCL, and 110 submitted. There are 18699 citations to ASCL entries in ADS. 3657 (95.81%) ASCL codes are in ADS. 1511 (41.32%) have ADS citations Most viewed codes: (but not most cited) kungifu: Calibration and reduction of fiber-fed IFU astronomical spectroscopy (ascl:1407.011): 43550 views COSMICS: Cosmological initial conditions and microwave anisotropy codes (ascl:9910.004): 42720 views McScatter: Three-Body Scattering with Stellar Evolution (ascl:1201.001): 27516 views BHSKY: Visual distortions near a black hole (ascl:9910.006): 25410 views Turbospectrum: Code for spectral synthesis (ascl:1205.004): 23677 views PyEphem: Astronomical Ephemeris for Python (ascl:1112.014): 19061 views NEMO: A Stellar Dynamics Toolbox (ascl:1010.051): 18760 views Astropy: Community Python library for astronomy (ascl:1304.002): 17786 views emcee: The MCMC Hammer (ascl:1303.002): 17744 views Starlink: Multi-purpose Astronomy Software (ascl:1110.012): 17422 views https://ascl.net/dashboard (October 22, 2025)
Comparison of mentions of astronomy software in ADS via API script, ASCL bibcode or acknowledgments search
Conclusions ●Establishing and monitoring software citation practices in a new journal requires culture change and infrastructure support. ●APIs enable fast access to long-term citation data, but results can be inconsistent. ● The software “alias” challenge persists - even when bibcodes or preferred citations are used, variation arises due to evolving practices and new software versions. ● Retrospective analyses are easier to conduct than tracking the introduction and growth of new or co-occurring software packages. ● Understanding software citation trends supports both the reproducibility of research results and the proper acknowledgement of key software.
References Teplitzky. Seismica’s Data and Code Policy in Action, doi: 10.26443/seismica.v4i1.1561 Bouquin, D. R., Chivvis, D. A., Henneken, E., Lockhart, K., Muench, A., & Koch, J. (2020). Credit lost: Two decades of software citation in astronomy. The Astrophysical Journal Supplement Series, 249(1), 8. DOI 10.3847/1538-4365/ab7be6 Chen, Tracy X., et al. "Best practices for data publication in the astronomical literature." The Astrophysical Journal Supplement Series 260.1 (2022): 5. ADS API: https://ui.adsabs.harvard.edu/
