Snow Avalanches at Mount St. Helens - Again

With the warm storm systems going through Washington in the past few days generating lots of snow up high, again Mount St. Helens has shed some avalanches big enough to be well recorded by the seismic monitoring system.  Wes Thelen, a seismologist at the Cascades Volcano Observatory (CVO) called my attention to the seismic records that clearly show long lasting surface events.  It is clear based on the timing and amplitudes that these events took place on the east side of the volcano, probably starting up high near the crater rim and traveling down and off the steep part of the mountain and perhaps as far as tree line.  

 

There are many seismic stations monitoring Mount St. Helens so it is interesting to look at the shaking at each of these stations all together.  The following two figures show the seismograms for 7 stations for the two big avalanches.  They are auto scaled to fit in their frame and instrument type and gain are different for different instruments so their relative amplitudes can not be compared on this figure.  Note that the two figures have different time scales as well.  

 

There are several things I find interesting about these seismograms. Each starts rather abruptly and then rapidly builds up to their maximum signal strengths within about 30 seconds. The high amplitude signals only last for 10 to 20 seconds and then quickly decay to a lower level that remains fairly constant for many minutes, perhaps slightly decreasing over time.  However for both events the end of the signals are rather abrupt, perhaps preceded by slight increases in amplitude.  These long duration signals imply a continuous source of shaking over this period, which means that either there were multiple back-to-back, merging together avalanches or once the major mass of material has fallen down the steep slope the avalanche changes to a comparatively slow moving flow that slowly moves across the lower, gentle slopes finally stopping with a jerk.

 

One can look at these same data using what are called spectrograms; as shown below.  Here the time axis is the same as above but the vertical axis is frequency content going from 0 to 20 Hz and the colors are proportional to signal strength, red for strong and blue for weak signals.  Some details I find interesting (but not sure I understand) are the observations that the early strong signals have much of their energy in the lower frequencies (2 - 4 Hz) while the long lasting later and weaker signals have energy throughout the band from 1 - 12 Hz).  Again the "stopping" signal is quite obvious in these spectrograms.