Quantifying gender gaps in seismology authorship

Abstract. According to 2018 demographic data of the American Geophysical Union Fall Meeting, seismology is among the Geoscience fields with the lowest female representation. To understand whether this reflects seismology more generally, we investigate female authorship of peer-reviewed publications, a key factor in career advancement. Building upon open-source tools for web-scraping, we create a database of bibliographic information for seismological articles published in 14 international journals from 2010 to 2020. We use the probabilities of author names being either male or female-gendered to analyse the representation of female authors in terms of author position and subsequently per journal, year, and publication productivity. The results indicate that: 1) The overall probability of the first (last) author being female is 0.28 (0.19); 2) With the calculated rate of increase from 2010 to 2020, equal probabilities of female and male authorship would be reached towards the end of the century; 3) Compared to the overall probability of male authorship (0.76), single-authored papers in our database are disproportionately published by male authors (with probability 0.83); 4) Female representation decreases among highly productive authors; 5) Rather than being random, the composition of authorship appears to be influenced by gender: Firstly, all-male author teams are more common than what would be expected if teams were composed randomly. Secondly, the probability that first or co-authors are female increases when the last author is female, but first female authors have a low probability of working with female co-authors.