Hi everyone,

After collecting 70 spectra with the Raman and changing temperature and pressure, I found that the change in enthalpy is relatively small and does not change largely based on pressure.  Temperature has a much larger affect on the enthalpy and hydrogen bonding index of water, which pairs nicely with previous work done by my mentors.  They even used some of my data in their 2018 budget proposal, indicating that 3000psi is nice to use in the lab, but measurements taken at even higher pressures (6000psi = ocean depth of 4000m) would be ideal and only accessible by deploying the ROV.  Now since this data set is wrapping up, I will most likely start a similar experiment with salt water and see if those changes in enthalpy look similar to the pure water graph.

Highlight of the week: seeing whales outside this morning!

-David

Happy Thursday!

I have been continuing to work on my machine learning script and making progress. In my post last week, I mentioned that I had written a script that extracts a specified chunk of data and automatically recognizes clusters in the data points. Since then, I have added a method which flips through a chunk of data at a specified time interval and identifies clusters in each interval (e.g. I can flip through a day's worth of data in smaller ten-second chunks and identify clusters in each ten-second interval). The reason for this is that we hope to one day be able to flip through the data and track specific data clusters over time.

My current task is to write another method which saves the clusters recognized by the machine learning algorithm back to the database where the data is stored. I have an error-free script but it doesn't seem to be doing exactly what I want it to do, so I will be working on that today and maybe through early next week. We'll see. Progress here seems to come in spurts.

Looking forward to seeing Liquid Robotics tomorrow - it should be pretty cool!

Cheers,
Rachel

On an island...

...well, not a literal island. We set sail yesterday on the Western Flyer after being delayed a day. Because the wind is blowing pretty hard (~23 knots last time I checked) we're staying in the bay and sampling around the M1 site. So far this trip has been great! We trawled last night and were up into the wee hours of the morning sorting through our catch and processing samples. Today, the goal is to fill the MRS system with Galiteuthis, drop off the MRS, then start collecting specimens on the "wish list." I'm looking forward to learning more midwater species and getting my hands dirty in more trawl catches! See you all on Tuesday!

Cheers
Megan

Good Morning all!
I'm in awe at how quickly this internship is going by! this week has been no acceptation..we began our Monday morning with an all interns kayaking trip at Elkhorn Slough estuary. While there, we saw purple stripped jellies, Kelp crab, harbor seals, and too many adorable otters! later on we held a root-beer float/cookie break and had a wonderful turnout
Tuesday was a full-work day that I used to catch up on project stuff and was able to organize, clean out, and annotate my Matlab files in preparation for our final research outputs and analysis.

On Wednesday, Diego, Emily (REU intern), and Myself joined Timothy Pennington and Marguerite Blum on their tri-weekly CTD cast, we were put to work with initial processing of water samples at 3-different stations, we all had a great and fun time!

Today and tomorrow the REU will be keeping us busy with a boating safety workshop that will culminate with us driving the boats, which I'm super excited for

Have a great weekend & see you all next week!

-Ash

Week 7:

I accomplished the Dictator aspect, where the SeeStar orders the Feather to wake up, record samples, and put its channels and itself to sleep. The Feather goes to standby mode and wakes up through a digital interrupt. I have 2 ports acting as sensors, a SeeStar port, and a port for producing dummy data. Essentially, there can be 3 ports for sensors and 1 for the SeeStar. Users can indicate which sensor(s) should take samples and write to the SD card before the Feather returns to sleep.

The next goal is to include a former student's code by the end of next week, and have it communicate with Danelle's SeeStar RPi system. One of the libraries, SoftwareSerial.h, however, does not carry well onto the Feather due to the pin-muxing.

It was a short-lived celebration on accomplishing one task, and there is a lot of pressure on integrating past work so that an adopted mentor can test their system by next week, but at least this is good hard deadline to keep me motivated on making a significant contribution to the project before the summer ends.      

Side note : I want a better method for finding how much current is drawn. 

Noooooo! I just wrote up a post and then accidentally refreshed. Dang it.

The short and sweet version is kayaking was great! I loved seeing the otters and I got to see one open up their food with a rock, so that was great!

For Kelly Benoit-Bird (KBB), I've written two drafts of different behind the scenes articles (one about the hydrophone on the Paragon and another about the common murre bird's acoustic signal) that are going to be edited/revised soon by KBB. I've also started working on her website a bit, but that's been a bit tricky. It's very well-written to start, so I might have wasted some time trying to rewrite it when it was already pretty good. Still, there's plenty of other stuff to be done, and my next task for that is to start working on adding a few pages about her instruments, acoustic identification of animals, and a few snapshot stories of the type of research she does.

For Jim Birch, I think we officially have a final skeleton for the website and I've started writing content now. It's going full speed ahead!

Finally, had a great morning seeing the whales outside! I got some blurry photos of some flukes, an OPEN MOUTH during a lunge feed (maybe ???), and some good spouts.

P.S. Guess the mystery pic! http://www.mbari.org/animal-plant-mineral-or-man-made-4/

Emma

Hey guys!

There are 4 whales breaching super close to the beach right now! I was just watching them from my mentors window. This week has been fun and productive! I really enjoyed kayaking in the Elkhorn Slough on Monday. It was nice to be out there again. I spent most of my time this week going through VARS images from Station M and doing individual counts. My goal is to get through a full pulse and compare the data to climate indices and oxygen data using a cross-correlation statistical method. I am excited to see how the results turn out! I was able to obtain a second computer monitor, which has made my work so much easier. I hope everyone is having a great week!

• Danielle Fabian

Morning!

So this all is going by way too fast and I only just feel like I'm really getting the hang of what I'm doing...and actually understanding the why/how part. I've been amplifying, running, confirming, PCRing, purifying, sequencing, and comparing DNA and RNA all week, and almost everything has gone really well, with my first "failed" sequencing reaction yesterday. But almost all of my sequenced DNA sequences have been matching up with the transcriptome data that I based my primers off of to originally amplify this DNA, which is really exciting because it means that our transcriptome sequences are reliable, and might actually be the gene that we're looking for.

Today I'm going to be troubleshooting and confirming that we actually have DNA transcript in the product that we used for our failed sequencing results, aligning the sequences that we do have to see if they match the transcript, and making some primers to bind in the middle of a long DNA sequence to see if we have get higher quality sequencing results. Then I'll be setting up a new sequencing reaction to see if I can get some better results. Lastly, I'll be doing some research on how to clone my sequence into a vector and maybe creating some primers for that.

In general, my next steps are to learn how to clone this sequence into a vector, and then into bacteria and try to express this protein, which would hopefully confirm the sequences we've found and isolated from both the transcriptome and my amplifications as the bioluminescent gene --> protein in vampire squid. Keep your fingers crossed for me!!

Cheers, 

Chloe

Hello everyone!

I spent two days this week on shipboard missions, and it felt great to go back to sea. My first trip was last Friday on the Paragon with Jeff Sevadjian and Chris Wahl to service the OA1 buoy. It didn't go quite as planned as there were some voltage issues created by the boat's battery that didn't let us bring the buoy onboard. We ended up servicing most of it from the side of the ship (quite the effort) and attempted to change the pH sensor underwater at the start of our dive. The old pH sensor didn't come out after half an hour of bobbing right below the surface, so we had to secure it again and proceed with our scheduled dive to check the chain and general state of the mooring. It was a really cool dive because there were sea nettles (the jellyfish) everywhere as we descended, and it added natural beauty to a routine work dive.

My other trip was yesterday on the Rachel Carson with Tim Pennington, Ash and Emily (REU intern). We helped draw and process water samples from the rosette and ate lots of food and took a couple of naps. I also saw a mola mola and many humpback whales on the way back, it was really cool. If you haven't been out I'd highly recommend it, unless you are extremely prone to seasickness--then it wouldn't be so much fun.

The other 3 days I spent working on the time series data from the three moorings (OA1 and M1, which I got to see this week on each boat trip, and OA2). I was able to finish all the analyses of the most recent time period and now I have two different time periods that I can compare (part of 2015 and some of 2016-17). I also finished the first section of my write up for this new time period, and I hope to finish the rest in the next few days. I still have some loose ends that I will tie together when the time comes and I figure out how to do it, but for now I'll keep working on the bulk of my project--it feels like it's starting to come together.

Anyways, I hope you all have a great long weekend!

Pura vida,

Diego

What is good, fam!

My mentor, John, is out for the rest of the rest of the summer. Since someone needed to water his three office plants once a week I thought it was appropriate to set-up shop in his office overlooking the Salinas River harbor and bridge to make sure the plants were doing OK. We also realized that his computer would do a much better/faster job of processing the spectrograms and training the CNN than my computer in my cubicle.

I presented the passive acoustic monitoring sound system to the science educators in the EARTH conference that MBARI is holding this week. I was able to show them samples of biophony and anthrophony and they really liked it. Teachers are awesome. I also let them know what I have been working on and a bit about my background. I want to thank the interns who were able to stop by and check it out. I really appreciate your guys' feedback.

I am just about ready to start training the CNN. As I type, John's computer is running the program that is creating filtered spectrograms that represent the whale calls I collected from Raven. Danelle is helping me with setting up the CNN model while I am generating these images. We are using a CNN model called Inception that uses transfer-learning which piggy-backs on other successful algorithms and starts with defined kernel weights. It's quick and easy. Given how much time is left in the summer, this is the best option for us at this point.

Progress has been made and I hope to be training the CNN Monday.

Have a great 3 day weekend!

Daniel De Leon

Week 6!

On Monday I fully wrapped up the Benchtop simulator, this involved writing up a user guide and on Tuesday I gave a small presentation to my mentor and Brent Jones, they're both interested in using the software. It makes me happy to know that I made something that MBARI will continue to use after the internship is finished. After the presentation I dove fully into project #2, the Inertial Current Profiler. I wrote a serial communication driver for the IMU using a protocol called SPI which I'm happy to finally get some experience doing. Communication protocols like SPI or I2C get used in almost every project that requires communication between embedded systems. Writing the driver was super educational and at times majorly frustrating. But today I finished it and what I'm working on now is a data logger that records all the sensor data on an SD card. There are some major limitations to what I can record if I want to get data quickly, right now it's looking like it'll be a file of straight binary code that I'll probably have to write some desktop software to parse into its relevant bytes before doing any data analysis on it.

Lots of whales visible from the beach today, spent some time out there watching.

Enjoy the long weekend
-Pierre

Hey everyone!

Project is wrapping up (ish)! I have finally closed my data pipeline loop. Now, someone can extract a bunch of images from a new video, use a neural network to create annotation predictions, manually correct the annotations, and then train an even better neural network with the additional data. This approach should allow any researcher to iteratively make a better neural network to solve any object detection and classification problem. The two big bottlenecks are the initial annotations to create the first neural network, and the manual corrections. The idea is that it takes less time to do corrections than annotations from scratch.

I also learned how to process videos at the video lab. I processed a video tape and will use it to gather metrics as to how well my neural networks is performing.

I've also started thinking about my presentation and how I'm going to introduce machine learning to a large and diverse audience.

Nathan

Aloha!!!!

Wow I am so late for this post!  I am so sorry. This week I finished my last interview at the Municipal wharf about the weekly water sampling.  I also participated in the Earth teacher workshop, creating a lesson sort of based on the things that have inspired me this summer working with CeNCOOS.  I was fortunate to work with two awesome teachers on creating this lesson and I think we put together something really good.  I am also almost done with updating the algal blooms timeline which has taken me forever! Whew I am all over the place and in mid thought and have a million projects going on right now.  I need to focus on wrapping everything up and working on my final paper. Oh ya, kayaking was awesome and it reminded me how much I miss kayak surfing. See you all on Monday!

Mahalo

Alia

Hi Everyone!

Super late for this post so sorry! This week I was able to continue with my data analysis. I finished up testing a linear regression and started with logistic regressions. After a bit of trouble with trying to use Excel for it, I've decided to use R. Today was mainly about testing out R and running some scripts to see if this is the best method for answering one of my questions (the relation between type of fecal path and POC flux). I've also continued with gathering data in VARs. T-shirt orders are all collected as well so super excited about that!

Hope everyone has a great weekend!
Victoria

Hi all!

Sorry this post is ridiculously late - it was a busy week! I've been continuing with daily sampling and analysis on my algae. I thought I was going to do more extensive photochemical analysis this week, but we decided to wait until next week to give the algae a bit more time to acclimate to the light levels. I did run the photochemical tests on the algae growing in the lowest light level on Thursday, and I'm excited to start looking at the data from that on Monday! Now that some of the algae are growing quickly I've been spending more time transferring them so they don't run out of nutrients in the media.
After a few setbacks, we're finally running the carbon dioxide and nitrate experiment on the chemostats. This week we gave them 1000ppm CO2 and the algae have been acclimating. The results haven't seemed quite as exciting as we were hoping yet, but we're giving them more time to respond.

Hope everyone's having a great weekend!