To the Editor
From the editorial 'Globalizing quake information' (Nature Geosci. 1, 803; 2008), I could not tell whether the Global Earthquake Model (GEM) requires participating nations to open their data sets. If it does not, then I hope GEM will reconsider and call for open data, to allow independent investigators to test or replicate the GEM analyses. If the output reports are always open, as GEM promises, then specialists could probably infer the broad contours of the input data from these reports, even if the input data are not open themselves. Hence, nations unwilling to make their data public may be unwilling to participate in the project even with closed data. If so, then the GEM needs to make the argument for open data as part of its case for worldwide participation.