Winter Seismograms at Uccle (February 2007) Seismogram 1. Time series showing the ground movements at Uccle from Sunday 4 February to Saturday 10 February 2007 (vertical component, ground velocity in µm/s, universal time = local time + 01h00). The raw unfiltered data are dominated by the microseismic noise of which the periods dominate at 6 and 16 seconds. This noise is well known since the development of early seismometers during the XIXth century, and is mostly caused by ocean waves. From 7 February this microseismic noise increases clearly, due to the arrival of a low-pressure zone over the Atlantic Ocean. A way to reduce the microseismic noise is to apply an appropriate filter, as shown below. Seismogram 2. Same time series as above, after 5 Hz high-pass filtering. The microseismic noise disappeared such that other phenomena can be observed in the residual signal, mainly due to human activity : public transportations are at rest between 01h00 and 05h00 (local time), the general noise level is lower on Sunday, and at night heavy freight trains (a) on the neighbouring line 26. More unusual is the decrease of the noise level on Thursday 08/02/2007 from 08h00 to 11h00 (local time), when snow fall caused traffic to stop during the rush hour (b). After 0.1 Hz low-pass filtering, long-period seismic waves appear clearly, e.g. on Sunday 4 February caused by two earthquakes, one in Cuba (20h57 U.T., Mw = 6.2), the other one on the southern East Pacific Rise (21h18 U.T., Mw = 6.1) (c). These were the strongest earthquakes between 4 and 10 February; however, filtering is not necessary to see closer events (off Strait of Gibraltar, 12 February 2007 10h35 U.T., Mw = 5.9) or stronger ones (Pacific Ocean, Macquarie Island, 30 January 2007 04h54 U.T., Mw = 6.8 or Mariana Islands, 21h37 U.T., Mw = 6.6).