This week consisted of more hurdles and more victories. After creating 3-day composites of the HF radar derived u and v components of sea surface currents last week, I was tasked with getting them all into raster format. I spent the first two days monkeying around with writing the Matlab arrays into ArcGRIDs but they would not line up correctly with the CA coast. After hours spent troubleshooting I figured out that this particular grid has rectangular cells, which is a no-no for the ArcGRID format. So I had to reluctantly revert back to GeoTiff formats, which are often clunky in ArcMap. The GeoTiffs exported correctly and the layers lined up perfectly with the CA coast. Victory! I spent the remainder of yesterday creating a model to tune the layers so they behave the way i want in ArcMap. Now im off and running again! Victory! Today will be spent playing with the layers in ArcMap and figuring out the best way to display sea surface current.

Once these layers are finished my final dataset to convert will be COAMPS Wind data. This dataset will be my toughest challenge, according to the CeNCOOS Matlab guru.

The big highlight of the week was the tour of the Western Flyer! I've been on several ROV cruises, on much smaller vessels; including the fishing vessel Donna Kathleen (docked next to R/V Zephyr). Our control room fit 3 people (pilot, navigator, photo-taker), shoulder to shoulder, on a single bench a half meter away from the displays. The Dock Ricketts control room was everything I dreamed it would be and more! The pilots/copilots "thrones" were soooooo cool. And the controls for the robotic arms were awesome! Great set-up!

 Kayaking was a blast too. Michael is a paddling machine! I pretty much let him do all the paddling and I just steered, and the guide kept telling us to slow down.

It's been another fast week!

On Monday we all experienced the Elkhorn Slough on kayaks. Amazing! Our guides were very knowledgeable about the animals and history of the slough. My favorite experience of the week was having an otter and her pup pop up right next to my kayak. In addition to otters we also saw harbor seals, sea lions, white pelicans, cormorants, and other birds.

A couple other interns and I went to the documentary "Green Fire" hosted by the MBA on Tuesday. The film is about Aldo Leopold and his idea of a land ethic, which I found, can also be applied to the ocean.

As for work, the pace has really started to pick up as now I have just two weeks left. Most of this week has been spent working on a way to organize the data I've already collected. Using Google Refine and a script that Lonny made, I've been able to successfully get a sum of populations per size class for each type of coral, per dive. Now off to annotate more Davidson Seamount videos. 

 Happy Thufriday!

This week I worked on analyzing CTD data collected from the Fulmar cruise. I made a bunch of TS plots and compared it to those generated with the wave glider data. The comparison will allow me to get an idea of how the surface is mixing and also the source of the water masses. From the CTD data, I found that the upwelled water is coming from about 50 m.

Hello and goodbye Week 5! This past week I have really felt like there are not enough hours in the day to get everything done, but maybe part of that is anticipating next week's business.  Last Thursday I went to MLML and picked up books on programming, Matlab style, modeling, phytoplankton pigments, and correlation and spectral analysis for engineering.  Needless to say, I want to read through all of these right now, but have only really worked into two of these books.  It was a productive move to pick up those books, because I find myself getting caught up in the codes within my scripts and don't feel like I'm getting very far with them.  I also had the brilliant idea to look into chl/fluorescence analysis via satellite oceanography and remote sensing, mostly to see how those instruments work and take measurements and to see how they compare with our LISST instrument.  I had a business call with my father (meteorologist at Goddard) and he suggested what I had already considered - that exploring the satellite oceanography ocean color data analysis techniques is beneficial for accounting for the mutitiude of variables in an in-situ study.  The big component of satellite oceanography is accounting for the atmosphere, and to a lesser extent upwelling radiance.  I really got to thinking about considering the atmospheric physics and dynamics that can contribute to bloom events, though I don't know how much of that I will have time to include in my summer project, especially given that the AUVs sample at depth at not only in the surface waters.  Anyway, I have made some headway in my Matlab techniques, thanks in part to Dorota and Dongsik, but still feel like I have a long way to go to effectively describe which combination of LISST channels can serve as a surrogate for chl/fluor mesaurements.

Jason - who is the CeNCOOS Matlab guru?  Is it Jenn Patterson?  I'd like to meet this person, as I will be back to CeNCOOS work this fall.

Michael - I'd love to see the plots/script where you determined the origin of the upwelled water.

Diane - its Fred Bahr. He's been here the longest.

Michael - I'd like to see your upwelling info too!

This week has been a busy one for Brian, talking to a bunch of people in IS and working on the security for the external server we will be hosting the DSG on. It's been less busy for me. I've mostly been working on getting the code production ready and making sure everything works on the "external" dsg (it's exactly the same as what's going to be external once we make sure all errors are fixed and that the web and data base sides of things are functioning as expected). I'm very excited for it to truly be made external!

I feel like writing on confluence every day!!! That is freeky!
So here I finally start to find my way through the tones of Facebook metrics. I now know how I want to analyze them in a quantitative way. We decided what parameter to test and which one will will not.
We will measure the impact of 7 different kind of posts posted in the morning or in the afternoon which means 14 combos. We need at least n=6 for each combos so I will need about 16 weeks!! So I am staying here a bit longer (I wish). No but I will keep collecting data after the end of the internship and hopefully have the data I need to present my master in December. I will also do a qualitative analysis of the content of the comments.
I came up with a spread sheet that everyone in ITD can check to coordinate the post on Facebook and to make sure it stays relevant with new release on the website or major MBARI event. So this will hopefully help to have a better social media coordination at MBARI.
I am also working on define clear social media goals based on the MBARI strategic plan. Only with a clear goal we could evaluate if we reach our objectives or not.
We have been increasingly active on Twitter and I believe it has had an important impact but I need to find some ways to get our Twitter metrics and it is not an easy task...
And last but not least, tomorrow I will start my very first cruise!!! So excited, and also a bit afraid to be sea such but Brandon gave me a lot of advices and I will make sure I follow them.
Jason lent me a book about social media and I am taking it onboard, I am very looking forward to reading it!

Apart from that, very nice interns activities this week with kayaking and visit of the WF. My family is having a blast here and my daughter loved painting cups that will go under water. She is now asking us all kind of weird questions about "what happens if I send a table/a biscuit/my ukulele under water?" Life is good here!!

Let me start by saying that I hate that this is the end of week 5! This is going way too quickly for my liking. Does anyone know how to slow down time? That'd be helpful

Anyway, this week has been very productive for my project. We've put together the files to test for self consistency with the LISST-HOLO images, so that's one BIG task out of the way (no more staring my eyes seeing what kind of diatom is on the screen, now that'll be someone else's job ). Also, I'm finally going through the Nanomia videos and the Holothurian videos are being processed for a 3rd time to make sure they get EVERYTHING!
Things are going really well We're hoping that next week, by the latest on Wednesday, we'll be starting the testing of everything. So YAY!

I realized that this project is extremely ambitious.... Danelle kind of put it into perspective that the amount of data I've been processing is usually done in a span of a few months... and I've done most of it in a few weeks. Whaoo! Go me!

Other than my project, this week has been pretty decent. House sitting for Craig Dawes has been really fun. Taking care of these animals is easy as pie and I'm really enjoying the television (missed my fix of awful tv shows!) But I am definitely missing my roomies and living at CSUMB... not the same without them

Wow... There is a surplus of smilies in that ^^^^ oops.....

Whew, let's see. I didn't have any data collection to do this week and all of the preliminary processing of the spectra I had collected was done, so I've entered some unfamiliar territory.

I've been playing around with MATLAB to import and plot my data, deciding on the most effective way to present the findings. Since I have very little experience with the program, it's taken some time to get used to, but I've been able to make good progress, I think. I've also been helping one of my mentors write a code to convert a series from the dataset we have from CO2 pressure to dissolved CO2 concentration. This sounded simple enough to all of us at first - we thought we just needed to look up one constant and then plug everything into a simple formula, but it turns out that our solubility system is more complex than that. And, since we are working at very high pressures, a few of the equations need to be modified. So it's actually involved tracking down papers with a lot of math that I vaguely remember how to do in them, deciphering the equations and the meaning of their coefficients, and taking what we need to put into the program. We're slowly getting there.

And this morning, for a change of pace, I got to go observe/work in the machine shop to help make a part we need to attach to the Raman instrument for a new method of deep sea spectral data collection. Definitely out of my area of expertise, but it was fun to see the other side of the work my group does and to get some experience using the machinery.

Kayaking on Monday: lovely. So was the tour of the Flyer. I'm excited about all the activities coming up next week!

My 5th confluence! Reading my co-interns' confluence has been so much fun, and looks like everyone's work is well sorted out and progressing. My work is going smoothly too. Let me check where I left last week... Okay. This week, I successfully(???) trained a prediction model using EOFs I generated last week and did quality controls on the outputs. But, somehow the outputs looked not right, and my mentor and I agreed that the units of the water velocity in the data he used and my data are different. So, I need to change the codes so it can convert the water velocity unit in my data to agree with the one he used. Hope this solves the problem. To verify the prediction outputs, my mentor ran drifter simulations, but this time we are planning to do something different. I suggested running a station keeping marine vehicle this time, and my mentor liked the idea. Later, I am actually planning to integrate the dynamics of the glider into this simulation, but for now we will go simple with a ideal perfect vehicle. Hope this idea turns out cool later.

What??? Again confluence?? Ok, let's go ...

Funny activities

Kayaking -> Was very very fun, with all the otters, seals, ... and so close to the kayak.
Tour WF -> It's incredible that a ship has been built around a ROV. Now I understand how they could record the awesome video that we can watch at the lobby. I really enjoyed it a lot, and was one of the thing that I wanted to do during my time at MBARI.

Work
During this week Michael and me had a meeting with Francisco Chavez to explain our idea. He likes the idea to work together but the time it's running fast and we have too much to do, so I don't know how it will end. My mentor still is on Holiday and will come back the Open House day. Now I am working with Brian Schlining. He have a lot of experience with Matlab and developing open source project, so he gave me very good advices about how to organize my code and my project. Now I am organizing my code and starting to look for a way in Matlab to get the value of a point in the middle of a 3D Mesh.

Haha! What a week. All I can say is there is no sweeter sound that that of a relay clicking after many hours of frustrating silence. Any other programmer who is reading this knows what I mean...

It took quite a bit of blood, sweat, and tears getting to this point, though. On Monday, I was still having difficulty getting any SPI clock or data out of the PIC. Well, after many hours of probing the hardware and software for problems, Chad and I decided to call it a day. Not a very encouraging way to end a Monday, no less. However, Tuesday morning provided us with an important breakthrough. Despite having looked over my code probably a thousand times, I decided that I would check (again) to see if I had set up the re-programmable pins associated with the SPI2 bus correctly. Surprise! I had actually configured the re-programmable pins for the SPI1 module, instead of the SPI2 module (which is the one I was using). Finally, I was able to get clock pulses and data out of the MOSI (master out, slave in) line. Joy of all joys!

Unfortunately, that wasn't the last of my problems. The digital I/O expander, which is the device I was trying to talk to, wasn't responding when polled for data through the MISO (master in, slave out) line. This meant 1 of 2 things; either I wasn't "talking" to the I/O expander in the way it was expecting, or I still had issues with my SPI2 configuration. Yesterday, Chad and I (once again) went through my code, the hardware, and the I/O expander's datasheet... nothing. I even searched the Microchip forums for any discussion about this particular I/O expander, but was unable to glean anything useful... other than a link to some example code, written by Microchip engineers, for a demo board that used this part. This code did reaffirm that I was formatting my read/write commands properly (at least for normal reads of the GPIO registers, which control/tell me the state of all the I/Os of the device). However, when it came to setting up the various control registers, I was still lost, because the example code set up the chip to work in a different way from what I wanted. Also, the datasheet for the I/O expander just plain sucks!!! Whoever wrote it made it extremely unclear about the way in which we were supposed to set up control registers (not to mention the fact that they placed more focus on the I2C version of the chip instead of the SPI version). Sadface.

So, much of yesterday and today was spent trying to work out these last few kinks (thankfully it was only a few). One thing that the demo code did for me was figure out an issue in the way I configured the SPI2 module's clock. I had the CKE bit set low in the SPI2 control register, which means that the output data changed on transition from idle clock state to active clock state. I actually needed this to be the other way around. After setting this bit high, I was finally able to perform reads on the I/O expander! However, despite being able to read from various other registers, I still couldn't read from the GPIO registers (the most important ones).

I have to stop here just to give a shout out to engineer Brian Kieft. Since Chad is away on vacation, he has been the one helping me work out the problems in my code. He had the brilliant realization that helped me fix this last problem. This I/O expander chip has two different banks of I/Os – GPIOA, and GPIOB. However, by default these registers are mirrored together; in other words, they are logically "ORed" together. For my application, I need these two registers to not be mirrored, since each of them have different functions. However, when you have the registers mirrored, their addressing system changes from when they aren't mirrored. It turns out that I have to initially configure the chip using the mirrored addresses, instead of the non-mirrored addresses, even though I'm simultaneously telling the chip not to be mirrored (ok, that was a mouthful, but it really was that confusing). So, after many hours of frustration, I was FINALLY able to read the values on the GPIO registers! Huzzah!

So... that is why I waited so long to update confluence. I wanted to wait and see if I could give you all a positive update... which finally, I have.

This week was spent mostly doing more CFD. Due to XFLR5's limited capabilities in modeling double curved surfaces, I'm forced to use a more complete CFD package. This is a bit of a bummer since this literally takes 100-1000 times longer to run a simulation to convergence. I think I'm now up to speed on Autodesk's Simulation CFD software for single trials. I was able to get a couple test designs inputted and back out lift and drag coefficients. I'm not entirely sure if my boundary condition for the bottom wall of the simulation is correct for modeling ground effects. I need to do some more reading on this. I also still need to learn how to integrate my Solidworks CAD model with the CFD software to automatically optimize the configuration of key geometry values. If all goes well I should be able let the system chug away on a few batches while I'm on the boat next Tuesday. Lots to do in the next couple of weeks!

Also started working on open house stuff figuring out replacement shaft adapters for the BYO ROV booth.

Not enough time to do all the work that needs to be done, and the setbacks have been a bit frustrating...transferring copepods with a cat whisker is a painstaking task! I'll be depleting my copepod culture further as I filter them to use with water samples collected south(-east-ish) of M1.

However, kayaking was great with my personal bird ID-er Patti : ) and the tour of the Flyer was a definite highlight. I also visited cube-land today and had mint cookies at Poseidon's Fortress (??), thanks to Diane.

I have a lot of water to process, lots of pictures to analyze, and am getting ready to collect more water next week on Tuesday. With the intern BBQ Wednesday and Thursday's outreach, it will definitely be a busy week.

Hi Everyone!

It was another incredible week at MBARI, wasn't it?

Monday's kayaking trip on the Elkhorn Slough was fantastic! It was so beautiful out there, and so much wildlife to see! Audrey was an impressive, double-kayaker, who kept me in line! It was fun to be out there as a group!

The tour of the Western Flyer was another highlight of the week. It was so cool to see the control room, as I have been viewing and listening to a lot of audio/video of the ROV dives in the Gulf of California. It must have been so great on be on that cruise! I loved the way that Linda showed us the vessel as she told us about life aboard an oceanographic cruise. She is obviously optimistic about all of us experiencing such a voyage someday!
I want to wish a Bon Voyage to Geraldine, who will be going to sea tomorrow! Have a GREAT time out there! Fair winds and following seas!

I am really looking forward to the day cruise aboard the Fulmar on Tuesday! Thank you for arranging this, Linda and George!

This week, I also worked on a video loop for the Open House, and made some headway on my project, but time went waaaay too quickly!

I can't believe that we have reached the halfway point of our internship! I really can't, and to be perfectly honest, I am starting to feel little bit concerned about my project. I am going to come in on Sunday, and try to make some headway!

See you next week, everyone!

-Patti

So I am definitely last this time...

Week 5: Another short one. Kayaking on Monday was a blast! There is so much to appreciate. Tuesday I went out to sea, again! As of Sunday, the PowerBuoy had stopped producing the amount of energy it should be, given the wave conditions. When we retrieved the system we learned that the bridal attaching the submerged plate to the tether line had snapped. The plate was hanging in a vertical orientation, instead of being in the horizontal orientation it should be in, to provide adequate resistance and thus drive the piston and turn the generator. So we recovered the whole system, and brought it back to MBARI. Further tensile strength and fatigue tests to ensue. The rest of the week I spent working on SolidWorks and MATLAB, making and testing the new fill in pieces for the submerged plate.