ALBERT

Description: Author Posting. © The Oceanography Society, 2018. This article is posted here by permission of The Oceanography Society for personal use, not for redistribution. The definitive version was published in Topics and trends in NSF ocean sciences awards. Oceanography 31(4), (2018): 164-170. doi:10.5670/oceanog.2018.404.

Description: The National Science Foundation Ocean Sciences Division (NSF-OCE) provides the majority of the support for ocean research in the United States. Knowledge of the trends in research and funding for NSF-OCE awards is important to investigators, academic institutions, policy analysts, and advocacy organizations. Here, we apply topic modeling to NSF-OCE award abstracts to uncover underlying research topics, examine the interrelationships between awards, and identify research and funding trends. The 20 topics identified by the model capture NSF-OCE’s 10 largest programs (~90% of awards) remarkably well and provide better resolution into research subjects. The distribution of awards in topic space shows how the different topics relate to each other based on their similarity and how awards transition from one topic to another. Awards have become more interdisciplinary over time, with increasing trends in 13 of the 20 topics (65%). Seven topics show a growing fraction of the number of awards while six topics have a declining share. Both the annual inflation-adjusted amount of money awarded and the fraction of the annual funding have been increasing over time in four of the 20 topics. Three other topics show a decline in both the annual amount awarded and the fraction of total annual funding. The identified topics can be grouped into three major themes: infrastructure, education, and science. After 2011, increases in the mean annual cost per project result in a relatively constant fraction of annual funding for infrastructure, despite a significant decline in the infrastructure fraction of awards. The information presented on research and funding trends is useful to scientists and academic institutions in planning and decision-making, while the metrics we employed can be used by NSF to quantify the effects of policy decisions.

Description: We thank T. Horner, B. Peucker-Ehrenbrink, K. Buesseler and M. Kurz for discussions and comments, the Woods Hole Oceanographic Institution Department of Marine Chemistry and Geochemistry for support, and A. Mix and three anonymous reviewers for their comments and suggestions. NSF deserves special credit for making its data publicly available.