It's not clear from the article, and I'm curious as to why it took them so long to figure out? Seismographs (which, as I understand, are able to "hear" the hum as well) are not a new technology.
The newspaper article (not the journal one) got essentially everything wrong...the journal article does not talk about this hum people claim to hear at all.
The paper describes a model for seismic noise generated at periods between 3 and 300 seconds (not Hz!) and I strongly doubt any human can hear that (frequency AND amplitude wise).
Seismometers did indeed record the noise for quite a while but it was mostly considered an annoyance. Now in the last decade or so we learned that we can extract useful information from the noise and this caused all kinds of new studies, some (like this one here) focused on determining the regions and mechanisms of the origin of the noise in the seismic wavefield.
