this abstract from AGU was interesting, though:

Slow Earthquakes Triggered by Typhoons
C Liu, A T Linde, I S Sacks

Taiwan experiences very high deformation rates, particularly along its eastern margins. To investigate this region, we have started (in 2003) to install several small networks of Sacks-Evertson strainmeters. The initial data from all sites show characteristics of good quality: tidal signals with very high signal to noise ratio and large (~10,000 counts on 24 bit ADC system) amplitudes; strains trending into contraction with rates that decrease exponentially with time and earthquakes clearly recorded. Additionally the instruments have recorded a number of slow strain changes with durations ranging from about an hour up to a few days; we interpret these signals in terms of slow earthquakes.

All of the slow events identified to date occur at the times of typhoons passing over or very close to the study area, but not all typhoons are associated with slow strain events (9 typhoons in 2004 were accompanied by 5 slow events). Seismicity for the area delineates a roughly north-south striking steeply dipping (to the west) zone with reverse slip; the shallowest extent of the zone is just inland. We look for source solutions consistent with that tectonic setting. The slow events exhibit a considerable range of amplitude and complexity; small, short amplitude events have a quite simple and smooth waveform; the longest (2 days) and largest (100 to 350 nanostrain at 3 sites) has waveforms with a lot of structure. The similarity among the stations (located in an ~isosceles triangular array with spacing ~10 km and 4 km) is indicative of rupture propagation of a slow slip source (equivalent magnitude about 5). We are able to match the essential character of the data with a very simple model of a downward propagating line source with uniform slip; the largest slow event appears to be comprised of 3 sub-events all starting at a depth of ~3 km with the final sub-event propagating to a depth ~10 km. Typhoon activity produces a large increase in short period (~sec) energy so it is not possible to determine whether these slow events are accompanied by non-volcanic tremor, as has been reported for the Nankai subduction and Cascadia slow events.

We hypothesize that the slow earthquakes are triggered by the typhoon activity due to the resultant low air pressure over land reducing the locking force on the fault zone. Such repeated slow events may explain why this area of high deformation does not experience very large earthquakes.