Greg Brenn

August 7th, 2012

Hi everyone,

After finishing up time at Boise State University last week,  I have shifted gears in finishing my abstract (due tomorrow) as well as my poster.  I hope to finish as much as possible before heading off to Australia on the 25th of August!

I've taken a little bit of time to reflect on my oustanding experience out in Boise.  I look back at the beginning of the internship at orientation in Socorro, thinking about how on earth I will complete a research project in a geology discipline that I have had no previous experience with at Union.  However, I realized that this opportunity with IRIS would be a great introduction to geophysics with complete immersion in a seismology-related project, and I just tried to absorb as much information as possible, both in orientation and at Boise State.

I really could not have asked for a better project (though the other interns probably disagree!), with the incorporation of seismology concepts with induced seismicity from a geothermal power plant.  The combination of field work, analysis of a large set of data, and the incorporation of computer programs that were new to what I was used to, definitely sparked my interest in pursuing more geophysics work after my undergrad in geology finishes up.

The AGU meeting this fall will only be a great conclusion to this experience, being able to present results from the past 9 weeks in an environment that will most likely be extremely overwhelming but rewarding!  I also look forward to see what the other interns were involved with this past summer, along with seeing what other research projects have occuring in the geophysical sciences.  

With my term abroad trip to Australia and New Zealand this coming fall,  I'll be taking a tiny hiatus from the geophysics and immersing myself in marine biology and terrestrial ecology!

I will probably do one more update on how everything is going, but I look forward to reuniting with all of the other IRIS interns in San Francisco this December!

July 30th, 2012

Hey All,

So this is my final week out in Boise, and I have to say, it seems like just last week that I got into Boise!  ~9 weeks later, it's about time to begin wrapping up my research out here.  I've begun writing my AGu abstract, which I'm not too stressed about.  Even though it's somewhat difficult to summarize 2 months of research into <200 words, I think I can handle it and will definitely get it in by the 8th of August.  

This past week has been pretty busy for me.  During this time I've gone through all of the passive seismic data recorded at Neal Hot Springs since last May, over 400 days of data.  As I was working through the ample amounts of data, it seemed like on most days during a specific time interval, there was definitely characteristic fluid flow seismic waveforms that were being generated on the 10 seismic stations I am analyzing.  Identifying the fluid flow events was the somewhat easy part (aside from deciding what is actually fluid flow and what is truck/man made noise).  The difficult portion of looking through this data set, which took close to 5 days to analyze in depth,  was trying to see if I can identify local events.  It was like searching for a needle in a haystack, except there was no needle (so we think).  Many of the events in the system were already catalogued through the NEIC database, but I was trying to find local events with frequencies greater than about 10 Hz, that were not catalogued in the event.  Once I find these events, I can go in and look at P-wave arrival times relative to each seismic station so I can estimate the direction the waves are coming from.  Unfortunately, only two very local events were found in the database, and these events appear to be arrving aprooximately 20-50 km outside of the Neal Hot Springs area, which suggests that these events were not caused by the fluid flow. 

In a way this conclusion is great news for US Geothermal, in that their geothermal power plant is not creating any significant seismicity that could ultimately affect production.  At first I looked at my finding and thought, "Wow all that time wasted, searching for something that does not exist!"  However, after review and looking at the larger picture of this project, going forward if further monitoring of these stations shows that there is in fact an increased in seismicity located very close to the production zone of the hot springs, then it can be assumed that a recent event that the geothermal company began performing must have caused the seismicity. 

Below you will see a map and a P-wave arrival screenshot.  These waveforms come from a non-catalogued event that most likely originated to the northeast of Passive Seismic Station 4, not in the area of geothermal production. 

I'll do another blog post later this week to truly wrap things up!

July 24th, 2012

Hey everyone,

During past week and these past couple of days, my advisors and I have decided to take a direct our attention away from receiver functions and begin to look at induced seismicity at Neal Hot Springs over the last year.  One of my advisors is still working on the receiver functions, and I think he has discovered the problem that we were having, which lied in correctly calculating the back azimuth.  Since we just figured this problem out and I have less than two weeks left out at Boise State University until I return home, I figured in order to produce semi-quantitative data for my poster, it would be best to go back and explore induced seismicity.  So far this change of gears has been successful, and by the end of these two weeks I hope to come out with some nice figures and results.

This part of my project will included quantitatively looking at fluid flow events and local earthuakes caused by the geothermal fluid flow in the seismic records going back until last June.  One method of locating and logging these events has been to use a spectrogram in the frequency domain to help identify/ classify the type of seismic event that is occurring.  For example, a fluid flow event will have a much different spectrogram image than a teleseismic event, which will have a different spectrogram image than a local event.  Typically, the teleseismic events will be characterized by their lower freuencies (<1 Hz) while the local events will have higher frequencies, ranging from 1 to 50 Hz.  These can be identified using a spectrogram. 

So my goal for the rest of my time out in Boise is to generate histrograms for these events.  I have many variables that can be generated, such as the timing of these events during the day, the timing of these events over the various months, and hopefully I can use the trends I see and match them up with the geothermal fluid logs taken from the lead geologist at Neal Hot Springs.

Oh, and I will also have to begin working on that AGU Abstract!  It shouldn't be too extensive, but it will be beneficial to get something written so I have something to work with when choosing figures and designing my poster.

Only a couple more updates to the blog, it's going well!

Here's an example of spectograms for 9 Passive Seismic stations, over an 8 hour period.  As you can see at the 1:00 mark, there is an increase in amplitude where this event occurs with a frequency ~>1 Hz to ~20Hz.  This event most likely is a local event possibly caused by geothermal fluid flow!

July 17th, 2012

With three weeks left in my internship, I think it's time to begin reflecting on the goals I made for myself at the beginning of this internship.

Overall, I would like to develop a greater understanding of geophysical concepts and how these concepts can be applied to solve problems.

-  I hope to gain experience with Unix as well as Antelope and understand the codes and commands that will help me interpret the large amounts of data given to me.

This is definitely a goal that I have accomplished.  I have become familiar with navigating through large amounts of data using Unix and Antelope, and I definitely hope that I can apply this computer programming experience to future ventures in geology/geophysics.

- Develop an understanding of geophysical techniques used in the field, specifically at the Neal Hot Springs geothermal plant.

At Neal Hot Springs, I have gained the experience of accessing and servicing many seismic stations around the geothermal plant.  I pretty much have the service sheet process memorized!  I have also taken part in acquiring paleomagnetic and magnetic data using different geophysical tools that are extremely useful in understanding the local earth structure. 

- I hope to analyze the data from the field to come up with valid interpretations about the processes that may be occurring, in order to prepare an abstract and poster for the AGU conference in December.

Because my project took a slightly different direction than what was originally planned (microseismicity to receiver functions), my new project involved working with a process that was even new to my advisors!  Although I have read and learned about the receiver function process, how it can be applied in the field, as well as the hoped-for results, with three weeks left it has been frustrating to not be able to produce sufficient receiver functions*.  Although presenting a poster at AGU was my end goal, and it still is, it's somewhat frustrating to think that I may not be able to acquire valid data to come up with any sort of interpretation.  

- Lastly I hope to explore all of what Boise has to offer, from the ample coffee shops to the bike trails along the Boise River!

I have definitely tried many of the coffee shops in the Boise area, as well as haven taken advantage of the Boise Green Belt, a 20+ mile bike/footpath along the Boise River.

* Last Friday, it became apparent that the gain settings on our seismometers out at Neal for each of the components, N,E, and Z were all mixed up.  Some gains were at a factor of 32, while other gains were at a factor of 1.  This creates problems when analyzing unnormalized data for receiver functions, because for some stations, the vertical-component amplitude was at a much higher gain setting than the two horizontal component amplitudes.  Rotating this unnormalized data proved to be difficult when having to use different gain settings.

3 of us went out into the field yesterday to fix these gain settings for the future data sets, but for the data sets that I hoped to use, the gain settings obviously could not be adjusted in the field.  Hopefully we can adjust these settings using code, but the question of whether these gain settings actually affects the receiver function application also arises.  Because in the deconvolution process, the Impulse response of the seismometer is cancelled out of the equation when isolating only the earth response.  This week, we'll troubleshoot these receiver functions, almost as a last-ditch effort to see if we can produce anything.  What's even more frustrating is that even when we shipped our data off to another receiver function resource to see what they produce, similar, bad receiver functions result, as shown in the image below.

Hopefully we can produce something significant, but filling that AGU poster with no interpretation of good data is really starting to haunt me...

July 9th, 2012

Today was a somewhat frustrating day.  Simply put, the receiver function code did not seem to create a decent model, as seen in the attached figure.  After changing the water-level parameter as well as add in the Gaussian Filter, I still could not clean up the Signal-to-Noise ratio, that would allow me to see the P-S conversions. 

On the bright side, my advisor and I have a working code that we can build off of and stare at for the next couple of days to see what we can do that may help fix our problem.  This deconvolution method is known as the water-level deconvolution method, working in the frequency domain, so tomorrow/rest of this week if we do not figure out what is wrong with our present code, I will investigate new receiver function deconvolution methods, such as working in the time domain or using an iterative deconvolution method.  I have been doing a lot of reading on numerous journal articles, and I am definitely understanding the process of creating receiver functions, but it's just applying that process to a new computer program that is becoming very difficult and frustrating.  I guess this is all part of this IRIS learning experience: finding ways to troubleshoot through problems to hopefully come up with a solution.

More updates as this issue is hopefully resolved.

Greg

(NOT A RECEIVER FUNCTION I WANT)

July 9th, 2012

Hi everyone, 

I'd like to update you all on this past week of progress made regarding receiver function analysis.  Kasper, my advisor, and I have been working with this Python program, Obspy, new to the both of us, but we finally finished the first step and created a figure that contains the vertical, radial, and transversal components of a seismogram (See figure below).  This is a great step forward because now I can use these three components in the next process, known as source equalization.  To put it simply (and to practice writing my research for the layman to understand), three components are needed to isolate the receiver response from other effects that create seismic waveforms.  Now, the next step is to use deconvolution, a somewhat complicated mathematic equation using Fourier Transforms, to create the receiver function seismogram.  Deconvolution takes a signal and removes these effects in the signal, to leave only the pieces of the signal we want.  Now all I need to do is add in some filters (both water level and Gaussian), as also mentioned in Erin's blog!  

Next week I hope to get as far as I can on the process for one receiver function, so I can hopefully copy this process for more than seismogram.  The process is going a little slower than I was expecting, but I guess this is what happens when one is working with a new code and new program.  More to come next week!

June 28th, 2012

Hi!

My Third week at Boise State has included both collection of data as well as becoming more familiar with what I plan to do through July.  The seismic data we collected from NHS last Wednesday is finally downloaded and turned into the proper file format for Antelope.  This data set will be extremely important to see if there has been seismicity related to fluid flow/injection in the geothermal system.  I've only briefly looked at this data, but next week I plan on attempting to find events/ characterisitic waveforms that correspond to these injections.  There are a couple of steps that need to be accomplished before identifying these events:

1.  Can I flag and catalog events during the time of data collection, and match these events up with the events in the NEIC database?  If so, these events should be removed from my data set to make the data more manageable in finding local events that were not catalogued in the NEIC database.

2.  Once these larger, regional/global events are removed from Antelope's Datascope, I will need to figure out what exactly I'm attempting to look for!  I will need to answer the question, what kind of characteristic waveforms are generated from fluid injections in a geothermal system?  I am going to search through previous research to see if there is any information on what these waveforms may look like.

3.  Lastly, I will need to figure out which filters are suitable to pick out the clearest waveforms. 

Additionally, I have been reading up on receiver functions concepts, Fourier Transforms, deconvolution, etc., to attempt to understand the theory behind how receiver function imaging is generated.  One of my advisors, Kasper, has been spending this week trying to figure out how we can use receiver functions in Python (ObsPy).  Though both Kasper and I have not had much experience with this program, we are playing around with sample data sets to see what this seismic python application can create.  Additionally, we have received code from a receiver function guru that visited during the IRIS workshop, and we hope to try and decipher this code and use it with our data.  The main step in this receiver function project is to use a code that picks teleseismic events from Datascope that are greater than a magnitude of 'X', and events at X-Y degree epicentral distances.  I'll be sure to update everyone on how this code is progressing next week.

I plan on relaxing this weekend in Portland, Oregon, visiting grandparents and cousins, and will be right back to work next week on the pieces of this projects!

PS Katie, if you have any resources on receiver functions that you think may be useful for a neophyte like I am, I would greatly appreciate it!!

Greg

June 20th, 2012

Hi everybody!

I'd just like to make a post on my blog before I delve into more research this week!  These past couple of days have been extremely busy.  On Monday, because I hope to use receiver functions later in this internship, I began to find and organize teleseismic events in the Neal Hot Springs data set, which contains data from last June until this May.  Once finding these larger events (greater than magnitude 5), I hope to look at the difference in P-wave and S-Wave arrival times, which will then be used to tell the relative response of the earth structure near the seismometer.  That's just a basic overview of what I'm doing with receiver functions, but understanding the conversion of P to S waves and how to use computer programs to image the receiver functions is a work in process.  

In the meantime, on Tuesday and Wednesday, I got the chance to join my roommate and masters student, Clint, to help him with collecting paleomagnetic data in rock surrounding the NHS geothermal plant.  From what I've learned so far, paleomagnetism is the study of the earth's magnetic field in rocks, and because the rock is mostly igneous out at Neal, when the rock is cooled, the magnetic info. relating to the magnetic field is stored.  The time of the formation of the rock can then be found. So, our job was to drill about 4 inch cores into various outcrops, using a diamond-tipped drill bit along a steel tube that was attached to a drill.  For some reason, it was frustratingly difficult to drill a decent core!  We must have drilled 30+ cores, and for some reason when we were about 2 inches in to the rock, the core would break!  After about 8 hours of this, Clint and I only gathered about 4 semi-decent cores, when we should have gotten about 30.  Although we did not receive the results we wanted, it was still an outstanding experience in the field.  Things don't go as planned, equipment malfunctions, but hopefully if Clint and I go out again, we can learn from these mistakes and obtain decent cores.  

On Wednesday, my advisor Lee and project partner Dan drove out to Neal to meet Clint and me.  I joined up with them and gathered the data from the 10 passive seismic stations around Neal, taking the flash drives and replacing them with clean drives- not difficult but definitely interesting to see the process in gathering the data.  I also had the chance to see Clint hook up his magnetometer and collect a magnetic survey across an area around Neal.  This data will hopefully be used to locate the boundaries between different geologic units, providing magnetic contrasts between different rock.  Later on in my geology career, I may have the chance to use these geophysical imaging techniques, so this was an outstanding opportunity to become exposed to what method can be used.

For the rest of this week, I hope to continue finding teleseismic events in the data, as well as begin to look at the newly acquired data.  I also hope to develop a schedule of tasks to get done in these coming weeks with my advisor, Lee.  

This next week will be packed with new, exciting but challenging data and computer programming, but I look forward to every bit of it.

June 16th, 2012

Hey everyone,

this first week at Boise State University has been an outstanding introduction for what to expect in the coming weeks.  I'm gradually becoming comfortable with the layout of the campus and city, and I'm really getting used to the lack of humidity out here!  Luckily for me, the IRIS workshop was held in downtown Boise this year, so I was able to attend speeches and poster sessions on topics ranging from tremors off the coast of Washington state to the future of ocean bottom seismology.  I've also had the chance to look at the application of receiver functions and how they are useful in understanding the structure of the upper mantle, a concept that I hope to pursue this summer.  There are so many applications of geophysics, and it's very exciting to think about how future technological advancements will contribute to our understanding of the forces driving our planet.    

One aspect of seismology that I had not realized until attending this workshop was the need for more efficient and less costly seismometers that can acquire as accurate data as the larger broadband seismometers, for example.  MEMS (micro-electro-mechanical systems) accelerators for example, are low-cost sensors that are gradually making their appearance around the world.  It will be interesting to see the advancement in this aspect of seismology.

After reflecting on this past week, I have come up with a couple of goals that I hope to attain throughout the course of this internship:

 Overall, I would like to develop a greater understanding of geophysical concepts and how these concepts can be applied to solve problems.

- By achieving this goal, I hope to gain  experience with Unix as well as Antelope and understand the codes and commands that will help me interpret the large amounts of data given to me.

- Develop an understanding of geophysical techniques used in the field, specifically at the Neal Hot Springs geothermal plant.

- I hope to analyze the data from the field to come up with valid interpretations about the processes that may be occurring, in order to prepare an abstract and poster for the AGU conference in December.  

- Lastly I hope to explore all of what Boise has to offer, from the ample coffee shops to the bike trails along the Boise River!

I leave you with a photo of latte art, taken from the Flying M Coffeehouse, downtown Boise. Additionally, I've attached an example of a photo of the data set I have been becoming familiar with, using Unix.

.

Best,

Greg

June 12th, 2012

Hey everyone, I arrived safely in Boise on Sunday at 12 pm, with Dr. Lee Liberty, my advisor, and Clint Colwell, my grad student roommate picking me up from the airport!  We then grabbed lunch in downtown Boise and then headed up to the Boise Foothills to look at some of the regional geology of Boise since it was such a nice day- 80 degrees, sunny and dry heat!  The city and its surroundings are beautiful!  We then took a tour of the campus, which is also very nice.  Coming from a school with just over 2000 undergrads, it's somewhat overwhelming to be at a university with 19,000+!  The Environmental Research building, where I will be spending most of my time, dwarfs the geology building back at Union!  I guess that's one of the tradeoffs of going to a larger institution....  However, the campus still is very easy to travel around, both by bike and on foot, and it's very convenient that the apartment that I'm staying at is literally a 30 second walk to the ERB.  

Today was my first day in the lab, getting situated with where I will be working, who I will be working with, and what I will be working on.  There's other grads and undergrads working here over the summer, and everyone seems very friendly!  I met my other advisor, Prof. Kasper van Wijk, and we discussed the project that I will be working on over the course of these 8-9 weeks.  Basically, there's seismic data that has already been acquired, as well as data that is currently being acquired at Neal Hot Springs at different seismic stations around the geothermal plant.  Currently the geothermal plant out at Neal is beginning its first injection phase, and it will be very interesting to see if there are any seismic events/induced seismicity associated with the beginning of the injection.  However, that data still needs to be gathered from the stations, which will most likely be done in the next week or two.  That will be give me an opportunity to gain some field experience as well as understand the geologic and geopraphic environments from which the data is acquired.

We also talked a little bit about receiver functions, time series acquired using three-component seismometers, which will be useful in understanding the earth structure around the geothermal plant.  Receiver functions are still extremely new to me as I have only read one packet about how they work, but I hope to fully understand them once all is said and done!  Additionally, myself and another undergraduate student named Dan began dabbling with Unix today, which had its ups and downs, as expected.  I gradually got the hang of the basic commands and the navigation around the different files and data sets, but there were rough patches, such as trying to work around bugs that we came across while interpreting some of the seismic data.  It's all a learning process, and it was only the first day so all of this programming will eventually make much more sense, as exposure and practice with the different programming languages is the best way to become comfortable.  

I will keep you all updated on how the rest of this week is going, but so far everything is outstanding!  I plan to explore some of the bike paths later this week, as well as possibly hit the rock climbing gym in the Rec Center.  As a side not, I'm gradually learning how to become self-sufficient with cooking my meals! The all-you-can-eat dining halls are unfortunately a thing of the past...

Cheers, 

Greg