Archive for July 17th, 2008

History 103, Where did RU17 come from?

Thursday, July 17th, 2008

The voyage across the ocean highlights the ability of the gliders extending from our backyard offshore New Jersey to now feeling confident to send them to extreme ocean environments which historically are difficult to work in.  During my graduate school years I was lucky to join a host of labs (Drs. Barbara Prezelin, Raymond Smith, Langdon Quetin, Robin Ross) working in the Antarctica along the West Antarctic Peninsula.  This region of Earth is characterized by major zones of sea ice, icebergs, and extremely large storms that produce epic waves.   It was descibed by early explorers as the land of gods, and it is.  This region is also changing. As the world warms, ecosystems are changing. Understanding the ecological consequences is critical as the magnitude and pace of the predicted warming will produce novel climate conditions. The Western Antarctic Peninsula (WAP) is undergoing the most dramatic climate change on Earth .

The WAP region has experienced a significant winter warming during the past half century (5.4 times the global average). The figure below shows the air temperature changes at two different stations. over the last 50 years.   This warming has shortened the sea ice season and perennial sea ice is gone. The maritime system of the northern WAP is expanding southward, displacing the continental, polar system of the southern WAP while 87% of the glaciers are in retreat. Associated with this warming has been a climate induced migration in the WAP ecosystem spanning the plankton all the way up to the penguin populations. We do not understand what is driving  the temperature changes or the most dramatic factors driving the ecosystem changes.  We know the decline of ice is driving one of the largest and most rapid climate induced shifts in a marine ecosystem on Earth . A project, the Long Term Ecosystem Research (LTER) has been conducted in the WAP for over 18 years.  The project was just starting as I was finishing graduate school, and was ongoing when I eventually joined Rutgers in 1995.  The history of the next period was highlighted in the blog "History Lesson 102".  The Palmer LTER is dedicated to study the processes underlying these changes and its impact on the ecosystem. In summer 2006, i was asked to attend a conference in Washington DC a steamy August.  I was lucky to joined by some old friends, Mark "Bull Dog" Moline (an old Antarctic buddy and part of the immediate COOL family) from Cal-Poly, Doug "Bam Bam" Martinson from LDEO, Gary "Nice Guy" Kirkpatrick, and Clayton "Sloccum" Jones.  After hanging with a bunch of smart people, when we were waiting for cabs to the airport in the hotel lobby we were encouraged to submit a small exploratory grant to test Glider tecnology in WAP.  It successful it might provide a unique tool to study the WAP climate change.

We submitted a grant to the National Science Foundation, and were very lucky to be funded.  By December 2006, two gliders were shipped to Antartic, traveling by frieght from New Jersey, to California (the Navy base in Port Hueneme), to Punte Arenas in Chile where they placed aboard the Research vessel Gould, and shipped to Antarctic.  The  person drafted for the field deployment was Clayton Jones from Webb Research.  Arriving in Palmer just after the new year in 2007.  While he was preparing the glider for the deplyment, Scott and Oscar was joining a team to provide a series of talks to the University of Tianjen and the Port of Tainjen in China.  On the 7th of January, a glider was launched from Palmer Station by Clayton, working Josh Kohut in the COOL room in New Jersey. A picture of him contorlling the glider of Palmer is shown below.

 The glider data was processed in near real time, and Scott and I were shipped data over the web.  This data collected in the Southern ocean was less than an hour old, when it was placed into our talks for the officals from Tianjen.  The idea that a data was being transported across the world was truly mind boggling for us old crusty oceanographers.  The glider was also recieving press in the Philadephia Enquier, the Asbury Park Press and other papers.  Unbeknowst at the time, a benefactor read these articles about Josh, a former Rutgers graduate student conducting a global climate change robotic experiment, and with great kindness contacted us on what would be a great next journey.  Scott, Oscar, and Josh had already recieved the pep talk and challenge from Dr. Spinrad from NOAA, and we suggested a glider to cross oceans to inspire the next generation of the scientist and engineers.  This was sufficient to gain the confidence of this donor, and we were graciously given the funds to acquire a glider, which is ru17 which now is deep in the Atlantic alone this evening,,, not even a remora for company.

For those interested in the Southern ocean, how did it work out?  The glider was the first robot to cross Antarctic circle, it collected over 1200 vertical profiles (the LTER had despite heroic efforts has collected just over 2400 profiles in 18 years using traditional techniques), and has found interesting information related to the why penguins feed where they do offshore Palmer Station.  This coming year the COOL team joins the LTER and we will focus on suing gliders to provide a more sustained presence in the Southern Ocean.  The long duration journey and new battery technologies in Ru17 will be critical to allowing the gliders to achieve the scientific potential we hope for.  Given that, the we toast ru17 tonight, its history of where it came, and the future it will usher in in the coming years.

The Scarlet Knight meets T.S. Bertha – 2 pm Saturday

Thursday, July 17th, 2008

5 pm update from the Hurricane Center.  Tropical Storm crosses 40 N about 2 pm Saturday.  Thats when we'll be seeing some strong winds.  We'll be making ready Friday morning, 24 hours in advance.






New SST Imagery – NLOM was Right!

Thursday, July 17th, 2008

Check it out! New Sea Surface temperature data just in, and the cold ring is clearly visible! Justin, you know what to do.  Get us into the northward currents.  We have a good plan for the night.

 Meanwhile, we are back at a small, unmarked yellow warehouse somewhere near Pearl City.  Chip recovered RU05 yesterday, and is replacing the attitude sensor.  The problem child is the small board Chip is holding.  We had an intermittent on it, and we can't really fly efficiently if the attitude sensor is flipping on and off on its own.

The radio in the background sitting on the wall beam is playing essential songs of the 80's for the guys working the forklifts in the warehouse.  But we are getting a lot of Peaceful Easy Feelings coming across.  We're working to get RU05 fixed up and back in the water tomorrow. Bad weather for small boats is coming on Saturday, so we need to get back out tomorrow before the wind kicks up. 

Meanwhile we set up a couple of laptops on top of an overnight shipment of Viking life saving equipment.  My laptop is checking on RU17 as we transfer all the mission logs from RU05 back to Rutgers.  All from a simple phone modem connection. Chip is linked to RU05 via a radio modem to check out his fix.



ru15 analysis

Thursday, July 17th, 2008

This is the initial analysis of the data taken collected by ru15 on its trip to Halifax.

Plot of Ballast Pump as a function of time

This is a plot of ballast pump (cubic centimeters) as a function of time. Note that the pump is measuring past it's specified position of + 230 cc. This could be electrical noise in the system while the glider is at the surface. Or could the ballast pump be actually at those positions?

This is a plot of m_battery as a function of time. You can see the voltage increase when we get into the gulf stream.

This a plot of m_depth versus time. The glider took some time heading out over the shelf in the beginning and then was in 100+ meters for the rest of the deployment.

This a plot of temperature measured by the ctd for the deployment. Most of the time in water < 10 deg C.

This a plot of temperature measured by the sensor inside the glider. The spike at the end is when the glider was sitting at the surface waiting to get picked up by the Halifax crew.

This is a plot of sci_water_temp in blue and m_vehicle_temp in red. The vehicle is a little warmer than the surrounding water.

This is a histogram of the pitch angle on climbs. The glider was at the optimal pitch angle the majority of the time.

This is a histogram of the pitch angle on dives. The glider was at the optimal pitch angle the majority of the time.

Tropical Storm Bertha – Sooner than expected

Thursday, July 17th, 2008

This morning's NOAA forecast has Bertha increasing its speed to the northeast.  It now looks like it will arrive near RU17, about 40 N, 45 W, sometime between 8 am Saturday and 8 am Sunday.


5-day forecast cone has it continuing Northeast between England and Iceland.

Tropical storm winds below still have the worst winds heading right at us.

Waves in near that 20 ft high mark in Bertha, but are starting to pick up in the vicinity of RU17.  We are up to 10 ft significant waves.  We also have clouds.

The main precaution we will take with RU17 will be to change its dive behavier.  We were going to do that on saturday, but now it looks like we'll do it earlier, likely friday morning.  We don't want to wait too long in case the waves get high and we have problems communicating.  Then it too late.  The modification is simple.  This is not our Hurricane Hunter glider, RU10, so we have no real need to go all the way to the surface unless we want to communicate.  Instead of rising to a depth of 5 m for each undulation, we will turn RU17 at a deeper depth, likely 25 m.  This will let the worst of the storm blow over out top.  We will also shorten up the bottom of each undulation.  No need to take a chance on the extra pressure forcing from waves.  Instead of stretching ourselves to our 100 m limit to stretch our power, we will only dive to about 90 m, giving us a 10 m margin of safety at the bottom.


Heading to 40 N on the NLOM highway.

Thursday, July 17th, 2008

Below is the satellite sea surface temperature data from the last 2 days.  White is clouds. Colors are ocean. There are lots of clouds.  Currents at location of RU17 are to the norteast.  We are looking for the northward currents associated with the cold ring.

Satellite altimetry says we are on the outer edge of the cold ring (the blue low in the sea surface height to our west). We should be seing increasing currents to north.  The idea is to ride this current north till we see the currents heading east around 40 N.


Below is the Navy NLOM model, with current speed plotted.  Strong currents are lighter colors.  In the model the cold eddy is farther west, and we are not quite in it yet.  Either way, we are moving the next waypoint for RU17 to more aggressively seek these stronger currents.  Yep, that means a turn to the northwest.  I know how much everyone hates to hear us talk about heading west to go east.


Below we plot the north/south component of the current from the Navy NLOM model.  The model has strong currents to the north shown as the bright green just to our west.  This is our highway to 40 N.

Why 40 N?  The Navy model also has strong currents to the east up there.  Here is a plot of the east-west components of the navy model velocities.  At 39 N. currents are heading west.  Along 40 N, there is a long band of bright reds indicating currents to the east.  That is where we are headed.

History 102, 2003-2008

Thursday, July 17th, 2008

The previous blog entry on our glider history (History 101) focused on the NOPP/ONR years (1998-2002) and the Coastal Predictive Skill Experiments we conducted down at Tuckerton, New Jersey in a local 30 km x 30 km box. This was before we had global communications operational on our gliders, so we used line-of-sight radio modems. This was the time period during which Tom Curtin at ONR said our mission at Rutgers was to work with Webb Research and get gliders into the operational Navy.  We’ve been working that mission ever since.


So here we pick up our story in the early summer of 2003. Through ONR we purchased our first set of 3 Slocum gliders, and they were beginning to arrive at Rutgers. We were on our own and we had a lot to learn. The first thing to learn was how to communicate with these gliders globally.  We all knew there was just not that many more years we could spend in the line-of-sight shadow of the Tuckerton Meteorological Tower before we ran out of sponsors.  And there were a lot of interesting research problems out there in the world beyond Tuckerton summer upwelling eddies. Leaving our 30 km x 30 km footprint required global satellite communications, and we chose Iridium.


The problem in 2003 was that the Iridium phones, like your cell phone, could drop calls right in the middle of a conversation with your robot. You would loose all the work you had completed, and have to try to reestablish the connection.  If you couldn’t get a good connection, and get the required commands across, you would risk loosing your glider.  Back then the Iridium cell phones we used were real phones that were disassembled and installed in the glider.  RU01, our oldest glider, still has the original Iridium cell phone with the key pad still in place. We had to make them work.


Steve Ackleson from ONR visited us in May, and I felt pretty bad presenting the news.  I had to tell our sponsor we had a problem, that it was critical to our development of a global glider fleet, and that I was not even close to a solution. Steve asked how long it will take to fix.  I told him it would take a month, not because I knew what to do, but because I knew we had a good group we could apply to the problem.  There was also one thing I learned at MIT that would help.  It was simply an approach we used everyday in our work back in school.  A simple three step process. Define the problem.  Devise the solution. Implement the solution.  Surely not rocket science.  But my office-mate Steve Koch repeated it often. Maybe the most important lesson I learned at MIT.  We formed a tiger team and went after the problem, meeting early every morning to define the daytime tests and late every afternoon to define the overnight tests.


But one barricade to progress was the cost of the Iridium calls. We had hours of testing over days and days planned, and we were running up quite an Iridium bill. Iridium satellite phone time runs about $1.50 a minute.  At the time, Ocean.US was running a program called the Iridium Pioneers.  The program provided free Iridium phone cards for researchers developing new communication systems for ocean platforms.  We went to Dave Martin, Director of Ocean.US, and explained our situation.  Gliders were still a very unproven platform at the time, but somehow we managed to get Dave to take a chance on us.  He managed to find one free Iridium card, but that’s all we needed.  I don’t know if Dave even knows what a critical role he played in the early days of glider development.


Back at Rutgers we managed to set up a communication test system that monitored every character we sent to the glider as it passed by, and also monitored every confirmation of that character that came back.  By testing various communication protocols over days and days, we finally concluded that you have to send information 1 character at a time, and until you see that character return, you have to keep resending it.  One discovery led to another, and we finally established a way to talk to the gliders, and a Z-modem program that Tom Campbell from Dinkum Software rewrote for the gliders.  Combining all these pieces resulted in a communication protocol that checked its transmissions, and if the phone connection was dropped, picked up the transmission where it left off when the connection was reestablished.  In a month we were good to go.


We followed this with as series of short nearshore tests in the water with ropes attached to the glider, and everything seemed to work.  By October we were ready to go. On that day we decided we were going to attempt something never attempted before.  We were going to take our glider and fly it from Tuckerton all the way across the shelf to the shelf break, turn it around and fly it back.  A round trip journey of 250 km. We also decided we would not just do this once.  We would do it every opportunity we had, starting something that we would end up calling the Tuckerton Endurance Line. We have been flying that line for 5 years now.   Data from the Endurance line can be found in papers with time scales ranging from storm events to seasonal climatologies. We continue to work with Steve on ocean optics research problems around the world.


In December of 2004, at the annual American Geophysical Union meeting in San Francisco, Terri Paluszkiewicz took Oscar and me aside to talk about our glider future.  Terri wondered if we wanted to get involved with the Navy flying gliders in deepwater for assimilation in models.  Even though the water was deep, gliders in the upper 200 m possibly could provide assimilation data for the more rapidly changing upper layer of the ocean that was influenced by the atmosphere.  Of course we agreed.  We needed the work. The catch was that SHAROM 149, as it was called, was taking place in the Pacific in January, and we would have to work over Christmas break to prepare.  The other catch is the Navy did not want just a single glider, they wanted a fleet.  Even on Christmas day I found myself on the phone with Clayton Jones from Webb Research, working on the details of assembling the fleet. 


But then on December 26, the Tsunami hit. What had happened to our colleagues in Indonesia that just spent the fall semester with us and had recently returned home? We still don’t know for sure.  I remember talking to Clayton on New Years Eve.  We were beginning to realize that every available Navy ship in the Pacific had been diverted to Indonesia for Tsunami relief. Our ships were gone, and without ships, we figured our exercise would be cancelled.  Within a few days everyone’s thoughts had been officially confirmed, and the mission was scrubbed. But after seeing the events on TV in Indonesia, missing a bit of Christmas did not seem like such a big deal anymore.


That put us on tap for SHAREM 150, which was to take place in the spring in the Sargasso Sea. Our mission was to assemble a fleet of four gliders, deploy them for the first time from the big Navy research ships from the four corners of a large box, and fly them into the center from over the horizon, rendezvousing in the middle and clustering for pick up. We needed four gliders to work, so as usual, we sent 5. The extra glider is there in case one is lost, for whatever reason, and is also the simplest way to ship a complete set of spare parts in case they are needed.


The mission was a game changer for the Navy. The models that assimilated the data performed better than expected.  A severe storm occurred that caused the surface ships to divert, suspending their mission to leave the area, while the 4 gliders continued flying and sampling the ocean, sending their data back to Navy for assimilation in the models.  They kept us there longer than expected, not allowing the Navy white ship return to port before the exercise was over. At one point we even transferred control to NAVO at Stennis Space Center, letting their glider pilots send one of the gliders whipping around an eddy they could see in the altimetry to speed its progress towards a central recovery point.


Now that we had the Navy’s interest, we had to make the gliders more robust, more rugged. The Navy envisioned fleets 10 times the size of our simple fleet of 4, and that would require robust gliders that requires minimum maintenance to deploy, pilot and recover.  In response, Terri started the ONR Glider Consortium, made up of the top groups working with the Slocum, Seaglider and Spray gliders. Many of the improvements that were made and tested as part of the Glider Consortium are now installed on RU17.  Probably one of the most significant, and certainly the most noticeable improvement, is the new short Digifin tail on the gliders. We had several failures based on the old design that have now been eliminated with the new fin.  And the fin is no longer a delicate failure point.  In one Digifin test, we appear to have been dragged by a fishing vessel near the shelf break.  The dragging snapped off one of the wings, but the fin remained intact.  The glider flew back to shore with 1 wing and a tail fin to compensate. On recovery we found scratches over the entire body, but a perfectly intact Digifin. Our electrical engineer, Dave Aragon, said the digifin is the most important improvement that has been made to the Slocum Glider to enable long-duration flights.


The NSF Coastal Ocean Processed (CoOP) program then came along and funded the Lagrangian Transport and Transformation Experiment (LaTTE) in the Hudson River Plume.  Our experiment was to inject dye in the plume, and follow it downstream, observing the transformations of the biology and the chemical properties of the water as it flowed through a highly instrumented coastal ocean observatory.  In each of the spring field seasons of 2004, 2005 and 2006, Oscar deployed a glider from the biology boat that we then controlled from shore in the COOLroom. In the COOLroom we used the satellite, CODAR, and Glider data to provide a 3-D context for the two ships at sea, coordinating their activities through daily briefings and 24/7 access to data.  Oscar, Josh and I talked about the coastal observatory and the LaTTE program at the U.S.- E.U. Baltic Oceanography meeting in Lithuania in May 2006, 1 week after we were out of the water.


Rick Spinrad, one of the Assistant Administrators at NOAA, saw our talks, and new exactly what buttons to push to get us going.  He sat us down, and said he has something for us to do for the good of our country. He first wanted us to read the Gathering Storm document prepared by our nations economic leaders about the need to entrain more students in Science, Technology, Engineering and Math (STEM) careers.  He then gave us three grand challenges.  First was to demonstrated flying gliders in coordinated fleets.  Second was to fly the gliders into hurricanes to gather data in the interface between ocean and atmosphere in severe conditions.  The third was a long-distance flight of inspiration.  Rick asked us to take one of our gliders, modify it, and fly it across the Atlantic on a mission that would inspire the public and entrain students in STEM careers.  Wow.  We had a lot of work to do.


We checked the first box in an ONR project, the Shallow Water 2006 Joint Experiment to be conducted on the NJ coast in the summer of 2006.  Terri managed to increase the size of our glider fleet with 9 new ONR owned gliders.  We flew a formation of 6 gliders along parallel cross-shelf lines where we would assemble then sweep across the front to the other side, providing data for assimilation in the models and information for the daily reports to coordinate up to 6 ships that were operating in the area. The experiment was 2 months long, plus spin up and spin down.  This experiment required us to switch from our LaTTE model of 24/7 coverage in the COOLroom for 1 week to distributed coverage from anywhere in the country that could be sustained for 3 months.  In trying to solve this problem, Hugh Roarty came up with the brilliant idea of letting McDonalds provide the internet infrastructure.  McDonalds and Starbucks all over the country provide access to WiFi. All we needed to do was provide the capability to browse the observatory datasets from a laptop, and we had the problem solved. We managed to coordinate the whole experiment from my laptop, running the glider fleet and sending out the daily morning reports from whatever McDonalds I could find.


SW06, shorthand for the Shallow Water 2006 Joint Experiment, also provided us the opportunity to check off the second item.  We had a single glider that was on the Endurance line get hit by Hurricanes and Tropical Storms before, but we never did a purposeful deployment. During SW06, Hurricane Ernesto propagated up the U.S. east coast, was downgraded to a Tropical Strom, and then in the Mid Atlantic, transitioned into a sever extratropical storm. We saw it coming, and when the surface ships diverted, we turned the glider fleet and flew it straight into the storm.  The dataset was amazing, especially related to the turbulent mixing.  With ONR funding, we are now working on a Hurricane Hunter glider with Oregon State University.  The glider will use the stretch payload bay, have turbulence sensors, acoustic profiling current meters, accelerometers for waves, and optical sensors for sediment transport.  We assemble the system every summer and sit it on the New Jersey shore, waiting for a hurricane to come by.


That brings us down to Rick Spinrad’s final grand challenge.  To fly a glider across the Atlantic in a way that inspires young people to take up math and science. This was a bit more difficult, because no glider actually existed that was capable of this.  The other issue was funding.  The Glider Fleet was funded by ONR.  The Storm Glider was funded by ONR. But what agency would fund something as risky as an attempt to cross the North Atlantic with underwater robot?  Turns out no one would do this with public money.  But private money, that’s different. Through the generous support of Rutgers Alumni, we were able to acquire sufficient funds to purchase a new glider to dedicate, and potentially loose, to this mission.  NOAA also needed the Lithium batteries tested to support one of Reg Beach’s NOPP projects to put some high powered biological sensors in the extended glider payload bays. The new sensors likely will require the high power density of Lithiums to meet the sampling requirements. We needed to test the batteries, and Reg said we should do something exciting with the battery test.  So now we had a vehicle, provided by private donations, and a need to test batteries, provided by NOAA. The long-duration mission was a go.  It would be enabled by the experience gained through numerous short-duration coastal missions in foreign waters including the Mediterranean, the Baltic, the Irish Sea, the Sargasso Sea, off Australia, and finally, off Antarctica. Oscar and Josh will tell those stories. Now it was time to build, test and deploy RU17.