My life continues to be consumed by the awesome world of active-source seismology.  Specifically, picking, picking, picking, and more picking ... I have now realized that no matter what field of science you're in at some point in your career you're going to end up picking something.  Maybe you're picking spectral peaks to identify organic molecules in soil samples, or different types of pollen scraped off of bees (I've done this), irregular heartbeats in a mouse's cardiogram, boulders in satellite images of the Moon (I've done this too), or just regular ol' seismic arrivals to locate earthquakes, no matter what field you're in you'll likely end up picking something!  

Example of picking boulders in satellite images of the Moon.  The image is ~ 500 m across and comes from the lunar region called Compton-Belkovich.  Image was taken as a part of the Lunar Reconnaissance Orbiter mission.  

Example of picking boulders in satellite images of the Moon.  The image is ~ 500 m across and comes from the lunar region called Compton-Belkovich.  Image was taken as a part of the Lunar Reconnaissance Orbiter mission.  

Now you may be asking "picking, what do you mean picking?" - I mentioned this phrase in my last post so if you need a refresher head back there because we're now going to dive deeper into how we actually do this.

I think it's fair to say that everyone has their own favorite computer program to use for refraction picking.  For quick and simple picks a great option is the Seismic Unix (SU) utility suxpicker.  This is a no frills, easy to use utility that can quickly be added into your pre-existing SU scripts.  It's a little finicky with mouse controls and given it's simplistic interface can cause you to easily overwrite old picks files so if you're doing anything slightly complicated I'd definitely suggest moving on to other programs.  

Another great option, if you have a Matlab license, is the utility upicker.  This is a Matlab package developed by William Wilcock in the early 2000's.  It's greatest quality is how seamlessly it let's you move back and forth between different display options.  The ability to switch between displaying traces in terms of offset, trace number, ... is essential when trying to tease apart noisy arrivals at far offsets.  Additionally, the package let's you quickly and easily modify old picks and their errors - another essential tool if you're like me and rarely pick things perfectly on the first try!  Until recently this has been my picking program of choice and continues to be heavily used by other graduate students in my department.  Now BIG DISCLAIMER, I don't believe this package is still available online.  We have a version floating around the department which is how I got my hands on it - but it looks like it has been replaced by a new package called tlPicker.  I haven't tried this version but seems like it could only be a better version so go out there and give it a try!

Example of picking in the Matlab utility upicker. Red lines show a few example picks.  

Example of picking in the Matlab utility upicker. Red lines show a few example picks.  

The last option and perhaps the most niche is the open source program OpendTect.  This program is actually intended solely for reflection data (both 2D and 3D) but turns out to be quite useful for refraction work.  We've moved to this software because of how easily it displays complicated 2D lines - let's stop here, complicated 2D lines?? In our experiment we collected data on several major, nearly perfectly straight 2D lines (the thick colored lines in the image below) but we also collected data in a racetrack style, organized-chaos of crazily curvy lines (dark grey lines in the below image).  Where the traditional picking utilities described above are perfect for our main lines, they wouldn't be as helpful for the majority of what I'll lovingly call "crazy lines".  And it's for these crazy lines that we've moved to OpendTect.

Map of all the lines we collected data on during our SEGMeNT experiment.  Thick, colored lines indicate the main transects we shot for the OBS data.  Thin gray lines are ALL the other data we collected for the both the reflection and the refraction work.

Map of all the lines we collected data on during our SEGMeNT experiment.  Thick, colored lines indicate the main transects we shot for the OBS data.  Thin gray lines are ALL the other data we collected for the both the reflection and the refraction work.

The wonders of OpendTect are best described in the image below - it allows you to view receiver gathers for all of your crazy lines in a single view.  This allows us to move from picking close, clear arrivals onto the far, noisy arrivals seamlessly.  

Example using OpendTect for refraction arrival picking.  Green lines show the map view projection of all the lines that we collected data along.  Yellow dot shows location of OBS 116 which is the instrument used for the receiver gather being shown.  

Example using OpendTect for refraction arrival picking.  Green lines show the map view projection of all the lines that we collected data along.  Yellow dot shows location of OBS 116 which is the instrument used for the receiver gather being shown.  

OpendTect definitely take's a little getting used to and you're going to have to force it to bow to your will and handle the refraction picking as you want it to.  Also because it's intended for reflection work, it will not keep track of pick errors - we're still struggling with how to handle this and for now rely on spreadsheets of shot numbers and associated errors. Despite this drawback though, it's definitely the route to go if you're picking 2D data with crazy, complex geometry.

We're madly picking away for now and maybe forever in OpendTect - just think 85,000+ shots recorded on each and every one of our 82 instruments! Ok, ok, we definitely will not see arrivals for all those shots, but still OpendTect is going to be my constant companion for a while.

Until next time (when I'll probably still be picking), get out there and get picking! You know I will.