# What condition(s) need to be met to be able to determine the hypocenter of a microseism?

Today, this XKCD popped up in a chatroom:

comic CC BY-NC 2.5 linked to the source

User Rob mentioned that (as usual with XKCD) this is based on the findings in a research paper.

In Source location of the 26 sec microseism from cross‐correlations of ambient seismic noise by N. M. Shapiro, M. H. Ritzwoller and G. D. Bensen the method to determine the location of microseism with a period of 26 seconds is described. While the hypocenter is indicated in latitude and longitude I couldn't find in that paper at what depth the microseism takes place.

Living in a country known for its seismic activity I'm used to pinpoint accuracy:

map from KNMI, image links to live map

To take two examples from the above map, Weert and Farmsum I learn the hypocenter of one is tectonic and one is induced. Both have a precise location, different magnitudes (1.1 vs 0.2) and a depth (12 vs 3 km).

What condition(s) need to be met to be able to determine the depth of a hypocenter of a microseism?

I'm not an Earth scientist and although I'm not scared of scientific details I appreciate if an answer comes with subtitles in layman terms.

• Determining the depth of a seismic event is tough. You have data in units of time, and you want a point (x,y,z) in space! What you need to link the two is 1) a model of expected wave velocities, 2) a lot of data to pinpoint one solution from all possible fitting points (x,y,z) as a_donda writes. KNMI has both, so can do a good job! In the paper you link, the available data are surface waves (propagating at the Earth's surface!). With this low-frequency signal, small number of recordings, poorly known African-plate velocities, and only surface waves available, it's pretty hard to do any better! – Erik Aug 11 at 19:38
• NB -- if the 26s microseism really is caused by oceanic waves crashing into the solid Earth, the depth is of course constrained to the sea floor at the (x,y) location found by the authors. That would be fairly well constrained! – Erik Aug 11 at 19:42