Colleagues, Please see below the details on a NSF- (EAR/GI-2148990) and SERC- supported workshop (free of charge) that we are running for 2 hours and 2 days on Feb 18 and March 4, 2026 to discuss changes needed to update our community’s publication best practices related to Experimental Determination of Trace Element Partitioning. Roger Nielsen (South Dakota School of Mines & Technology, Rapid City, SD, United States) Gokce Ustunisik (South Dakota School of Mines & Technology, Rapid City, SD, United States and American Museum of Natural History, New York, NY, United States) As many of you know, EarthChem is part of an ecosystem of related projects governed by IEDA2<https://www.iedadata.org/> and includes a system for Earth Sample registration which is designed to advance Open and FAIR Samples in the Earth, Environmental, and Planetary Sciences, as well as The Library of Experimental Phase Relations (LEPR)/traceDs<https://lepr.earthchem.org/>, which compiles and provides access to experimental petrology data. We have been working on the experimental trace element partitioning database for several years (traceDs). We continue to add experiments as they are published and provide access through the EarthChem portal, as well as running new phase equilibria and trace element partitioning experiments. https://earthchem.org/communities/experimental-petrology#about-lepr-trace-ds The workshop is designed to be forward looking with respect to partitioning experiments. Specifically, our goal is to update the best practices related to the publication of experimental data. Over the past few years, we have discovered that there are aspects of the experimental partitioning database that may result in systematic bias in models based on the existing published data. The breadth of these problems has only become clear as we have been able to examine entire mineral/melt datasets. Unfortunately, many of the issues cannot be resolved using the published results, at least in the current form. The most important of those issues (but by no means the only one) is the inclusion of contaminated analyses in the data used to calculate averages of the analyses for each phase in each experiment. As per Betts et al., 2025 (https://doi.org/10.3389/fgeoc.2025.1660826) Cung et al., 2023 (https://doi.org/10.1029/2023GC010876); Nielsen et al., 2017 (https://doi.org/10.1002/2017GC007080) Nielsen et al., 2020 (https://doi.org/10.1029/2019GC008638) the inclusion of contaminated analyses will impact the averaged results in proportion to the partition coefficient of the trace element being analyzed. If we apply standard regression analysis to the data in an attempt to create predictive models of trace element partitioning, those regressions will include a bias that will result in consistently erroneous model results. This systematic bias is not related to the specific model being applied – but is an inherent characteristic of the data. This is not a small problem – inclusion of contaminated analyses can result in errors in the calculated partition coefficients of greater than an order of magnitude for highly incompatible elements. Correction of the existing data requires access to all the data used to calculate the averages of each phase in each experiment. However, such data is rarely included in the published results – and even if a few of the published papers did include all the analyses, that would not allow us to correct the entire dataset because the impact of contamination is dependent on the diameter of the analytical volume and the size of the crystals in each experiment. The reason we have not published all the individual analyses is related to space limitations in journals in the past. That reason is no longer valid due to the advent of supplementary files. We as a community of experimentalists have not made the changes made possible by that possibility. One possible solution is for everyone to unilaterally change the way they write their data for publication. It is important to note that this problem cannot be solved by a subset of experimentalists. To be able to move forward, we need to have access to essentially ALL the data as we go forward. Obviously, the issue raised by contamination is not the only one that could be addressed by an increase in the “granularity” of our published data. No change will happen unless there is evidence that the community agrees to such changes (experimentalists, reviewers, program managers, editors, etc.). It is important to note here that such changes will not require additional documentation of our experiments – we have all collected such contextual information – we just have not had the mechanism to provide access to it within our publication protocols. What we propose is that we begin the process with two workshops in the next couple of months to discuss the magnitude of the problem and what “must” be included (individual analyses from which the averages were calculated) in datasets submitted for publication, and what “should” (analyses of secondary standards) or “might” be included (e.g., images). Towards that end, we will be collaborating with SERC to organize a virtual workshop<https://serc.carleton.edu/ieda/events/expet_workshop/index.html> with a 2-hour session on Feb 18 at noon-2 PM MST with a second session on March 4 at the same time. We will begin the workshop with a short (15 minute) presentation on the nature and magnitude of the problem and where it is most critical. Following the workshops, we can ask a smaller working group to come up with a draft of a white paper that would lay out recommendations for how to process and publish trace element data. We can then have a final meeting at Goldschmidt in July to bring in any other perspectives and inform the community as to what we are planning to do. That statement of best practices would/could then be published in an appropriate format/journal. We will not require an application; we just need participants to "sign in" or register somehow so that we can keep track of those who want to keep informed and/or meet at Goldschmidt. A short survey at the end of each of the two sessions will help us to document the "sense of the group." If you feel you have information relevant to the topic, please contact us, and we will provide you time to address the group in one of the workshops. The current title of the workshops is: Experimentally determined trace element partition coefficients: A review of recommended practices for design, analysis and data presentation and publication * Workshop website<https://serc.carleton.edu/ieda/events/expet_workshop/index.html> * Registration<https://serc.carleton.edu/ieda/events/expet_workshop/registration.html> If you have questions regarding participating, please send us an email at roger.nielsen@sdsmt.edu<mailto:roger.nielsen@sdsmt.edu> or gokce.ustunisik@sdsmt.edu<mailto:gokce.ustunisik@sdsmt.edu>. If you can join us complete the registration<https://serc.carleton.edu/ieda/events/expet_workshop/registration.html>, and we will make sure you get the information necessary to connect with the workshop. If you know someone who would be interested – please forward this announcement to them. Our goal is to be inclusive – and involve as many early/mid-career colleagues as possible since you are the ones who will be dealing with the outcomes longest. Program Wednesday February 18, 2026 (all times are approximate) 2:00-2:20 PM (EST) Introduction and statement of the problem 2:20–2:40 PM (EST) Development of questions for breakout groups (max of 5 participants in each group). 2:40-3:20 PM (EST) Breakout groups – goal is to develop and prioritize our recommendations for best practices related to what data should be included in publication of the results of trace element partitioning experiments 3:20-4:00 PM (EST) Reconvene and discuss recommendations from each breakout group In the 2 weeks between Feb 18 and Mar 4, we ask that attendees gather feedback from their network of collaborators and students. Think about whether there critical perspectives not captured from the breakout groups in the first workshop session. Consider which of the recommendations from the first workshop are: * What ideas/recommendations are viewed as most significant? * Which represent “a bridge too far” at this time * Are there recommendations that are viewed positively but may be “aspirational”? Wednesday March 4, 2026 2:00-2:20 PM (EST) Discussion of recommendations from the breakout groups from the Feb 18 session 2:20–2:40 PM (EST)                      Development of questions for breakout 2:40-3:20 PM (EST) Breakout groups – mixed membership compared to the first sessions breakout groups 3:20-3:50 PM (EST) Reconvene and discuss recommendations from each breakout group. 3:50 PM (EST) Decide on next steps – those below are not necessarily in chronological order Do we: * Write up a short recommendations paper for Chemical Geology; G3; American Mineralogy? * Meet at Goldschmidt 2026 to iron out/amend the list of recommendations? * How does the group advocate for the implementation of our recommendations Dr Roger L Nielsen Academic Policy Coordinator Research Scientist South Dakota School of Mines and Technology Rapid City South Dakota 541-231-7943