In the last month I have been captivated by the new OBS data we received!  In case you haven't been paying attention (or more likely I have been lapse in my blog updates), we have been patiently waiting to receive the data from our enigmatic, unique, miraculous "lake" bottom seismometers that were bravely retrieved from the depths of Lake Malawi in November of 2015.  These represent some of the few seismometers that have been chucked overboard into lakes (usually these ultra-sensitive instruments are thrown into the dark blue abyss of the oceans).  More important than their unique deployment location/situation, this data represents the VERY last segment of data to be retrieved for the SEGMeNT (Study of Extension and maGmatism in Malawi aNd Tanzania) experiment.  Wahoo!

Overview of the SEGMeNT experiment.  I participated in the acquisition of the passive source (red triangles) and active source (yellow circles and orange lines) seismic data.  Background colors shows elevation from Etopo1.   

Overview of the SEGMeNT experiment.  I participated in the acquisition of the passive source (red triangles) and active source (yellow circles and orange lines) seismic data.  Background colors shows elevation from Etopo1.   

Seismic data collected in oceans (or in this case, lake) environments are notoriously noisy because the ocean currents push and shove the instruments around on the ocean floor.  Many years ago any attempts to "clean this data" (i.e. bring out the seismic signals) would be a hopeless case but lucky for us times have changed!  Currently there are well used methods to remove a portion of the noise that we want to remove.  Specifically, these techniques aim to remove the noise generated by the motion of the seafloor caused by pressure variations (termed compliance) due to traveling oceanic waves.  If you're really interested in this science you might want to start with Broadband seismology and noise under the oceans by Webb (1998) and Removing Noise from the Vertical Component Records of Ocean‐Bottom Seismometers: Results from Year One of the Cascadia Initiative by Bell et al. (2015).  

Needless to say these methods for cleaning OBS data are incredibly useful and thus I've been determined to apply these corrections to our data.  And what do you know it worked! Thanks to being at a large research institution I did not have to travel far to ask a fellow graduate student to share their compliance correction codes with me.  I've put an example of what these codes can do down below - it's practically magic! In the original trace there's only a noise hint of a an earthquake buried somewhere within the noise and then afterwards - voila!

Comparison of earthquake data recorded on lake bottom seismometers without correcting for compliance noise (right) and with the compliance correction.  Before the correction only a wisp of an earthquake is identifiable within the trace but afterwards the earthquake is easily observable.

Comparison of earthquake data recorded on lake bottom seismometers without correcting for compliance noise (right) and with the compliance correction.  Before the correction only a wisp of an earthquake is identifiable within the trace but afterwards the earthquake is easily observable.

Ok so this update has gotten longer than anticipated and obviously I've been consumed by other similarly exciting projects as well but more on that in the next update.  

Until next time, keep your references updated and your ocean noise thwarted!