In the spring of 2009, the Waitt Institute conducted a massive underwater search for Amelia Earhart’s lost Electra in the remote South Pacific. Dubbed the CATALYST 2 expedition, this search was the largest, continuous deep-water seafloor mapping effort ever undertaken, and introduced two new hi-tech Autonomous Underwater Vehicles (AUVs) to the ocean science community.

This series of films serves as the expedition’s video log. Follow along with the Waitt Institute’s CATALYST 2 expedition team as they embark on a mission to find Amelia Earhart’s lost Electra. Narrated by Waitt Institute Director of Operations, Michael Dessner.

View Video Logs #1-10
View Video Logs #11-20
View Video Logs #20-26

Individual Video List
2.01 - Pago Pago
2.02 - One Last Test
2.03 - Setting Sail
2.04 - Home Sweet Home
2.05 - A New Approach
2.06 - A Tangible Search Grid
2.07 - Trawl Net Test
2.08 - Farming Dots
2.09 - Launching Ginger
2.10 - At-Depth
2.11 - Recovering Mary Ann
2.12 - A Promising Target
2.13 - Gone Fishing
2.14 - New Search Boxes
2.15 - The Daily Grind
2.16 - Disaster Strikes
2.17 - Trawl Net Survivor
2.18 - Secret Weapon
2.19 - Mary Ann Hits Bottom
2.20 - Down For the Count?
2.21 - Mary Ann Phone Home
2.22 - Cataloging Critters
2.23 - Back on Track
2.24 - Plane-Shaped Rocks
2.25 - Too Much Time at Sea?
2.26 - Homeward Bound

For the CATALYST 2 expedition, the Waitt Institute based operations out of the port of Pago Pago in American Samoa. The deep, naturally sheltered harbor of Pago Pago was the closest port to the search site that offered the amenities and facilities that were needed to start our journey. While preparing for the expedition, the CATALYST team enjoyed a traditional Umu feast put on by the Pago Pago Yacht Club. The Umu usually consists of local foods like breadfruit and kalua-style pork cooked in an earthen oven. The CAT 2 team had a great time and it was a much-appreciated homemade meal before we set out to sea.

Before putting out to sea, the vehicle engineers from Woods Hole Oceanographic Institution decided to run some last-minute diagnostic tests on the AUVs, so they prepped the vehicle nick-named Mary Ann, checked her components and installed the electronic still camera. When she was ready to go, they rolled her out onto the Launch and Recovery System (LARS), and lowered her into Pago Pago harbor. The team ran Mary Ann continuously while they monitored her systems and tweaked her for maximum performance. Twelve hours later, the AUVs got the thumbs up from our engineers, and we were ready to set sail.

On the morning of February 10th, 2009, we officially began the search for Amelia. Up on the bridge, Captain George followed the guidance of the Harbor Pilot to clear us out of port, while out on deck, the expedition team watched the shores of American Samoa disappear, knowing that we may not set foot on land again for 7 more weeks. From here, our 4-day transit would take us 1000 miles northward into one of the most remote regions on earth. Medical assistance and rescue would be days away, and anything but the most severe injury would have to be taken care of at sea. Everyone on board was anxious to get to work, and also hoping for a quick discovery, so we could head back to the safety of shore before dwindling fuel and food would force our return some 45 days later.

This is the Research Vessel Seward Johnson – our home-away-from-home for two consecutive 45-day stretches at sea. At a little over 200 feet long, she was just the right size for our 30-person expedition team. The forward dry lab was set aside for the AUV team. The aft wet lab was set up for the marine biologists who would be organizing catches from their net trawls. And the third, and largest, lab was set aside for sonar analysis and became the nerve center of our expedition. This is where we had team meetings, discussed the search plan, and spent countless hours hunting through data for the lost Electra. The entire aft deck was customized to support our two underwater search vehicles and their continuous 24/7 operations. On the bridge and in command, Captain George Gunther kept the ship running smoothly, making sure we were where we needed to be, when we needed to be there. Overall, the Seward Johnson was a good ride and a great platform from which to work.

The Waitt Institute took a new approach to the search for Amelia by working with aviation investigators who reconstructed her last flight as they would a modern aviation accident. As part of their research, they inspected a vintage Lockheed Electra that had been modeled closely to Earhart’s 10E. They also took thousands of photographs of the plane to have on file to compare with any debris we might find on the seafloor during the mission. Our flight re-construction team also tapped their own perspective and experience as pilots to try and predict the steps that Earhart and Noonan may have taken during the last hours of their flight. This direct experience combined with time spent in modern flight simulators and pouring over thousands of historical documents, helped crystallize their theories concerning the events that took place in July of 1937, and resulted in a 2500-square-mile search area.

During our 4-day transit to the search site, the CAT 2 operations team used the time to refine the search area provided by our aviation investigators into a workable survey grid. In order to cover the 2500-square-mile area as efficiently as possible, we had to determine the best strategy for laying out Deep Ocean Transponders and leapfrogging the AUVs through the search area. The team had to take into account the time needed for launching and recovering AUVs, setting transponders, and efficiently running the ship back and forth between these operations. In the end, with all of these factors carefully weighed, the search area was broken down into 88 mission boxes, each covering roughly 24 square miles and representing a single 23-hour AUV sortie. After over a year of preparation and research, we had our final search grid laid out and were steaming full speed ahead towards our first AUV launch at the site.

In addition to our search for Amelia, the CAT 2 team put together a plan to carry out biological sampling while the vehicles were busy surveying. During our first transit to the site, Harbor Branch’s Project Manager, Lee Frey, led the team in fine-tuning the deployment of a trawl net that the science team would use for catching specimens.

After a long transit from American Samoa, the team arrived at the south end of the 2500 square-mile search area. Their first task on site was deploying an array of four Deep Ocean Transponders (DOTs). The DOTs work in concert at depth with the AUV’s navigation system to give the vehicle a precise fix on its location at all times using sonar pings. The DOT configuration consists of two bright yellow flotation spheres attached by a long line to the transponder itself, which is then followed by another stretch of line and a weight to anchor the entire set-up on the seafloor. Later, after the AUVs have finished surveying a particular area and no target has been found, a command is sent to the DOT from the surface that tells it to release the weight. The transponder and flotation spheres float to the surface, where they are recovered alongside the ship by team members using a boathook. The transponders are then re-configured and dropped farther up the line in the search grid, where they await the next AUV sortie.

Launching a 2000-pound AUV is a lot simpler than you might think. The AUV operations team first performs a diagnostic checkout on the vehicle’s electronics and instruments to insure that all systems are ready for deployment and functioning properly. Once the pre-dive checklist is complete, the AUV is winched out onto the back deck via a set of removable rails onto the LARS – or Launch and Recovery System. The LARS is a portable cradle and A-frame designed specifically for these AUVs and can be mounted on nearly any research vessel. Once the AUV is sitting on the LARS, it is secured into the docking head. The LARS frame then hydraulically lifts the vehicle up and out over the stern of the ship where it can be safely lowered into the water. After being released from the LARS, the AUV is towed behind the ship until we reach our launch point. At that time, a descent weight is sent overboard and the AUV begins its journey of more than 3 miles down to the seafloor. After launch, the operations team has limited communications ability with the vehicle, but they are able to track the vehicle’s progress via a customized software program that shows the vehicle’s depth and current direction. Team members rotate throughout each 23-hour sortie to maintain constant observation of the AUVs at depth.

The Waitt Institute’s CATALYST AUVs took roughly 3 hours to dive to a depth of nearly 4 miles below the ocean’s surface. The AUV would then use side-scan sonar to map the ocean floor in long over-lapping lanes, referred to as “mowing the lawn.” At the end of a sortie, the AUV would then drop a small weight and begin its climb to the surface.

Recovering an AUV after a mission can actually be riskier than launching one, but the operation itself is relatively simple. When the AUV team receives confirmation that the vehicle has reached the surface, the ship pulls up alongside the AUV, where a command is sent to release a float from the AUV’s nose. This float drifts away from the vehicle extending a recovery line from the float back to the vehicle. An air-powered cannon is then used to fire a grappling hook across the recovery line and the deck team uses those lines to attach the vehicle to the LARS. The LARS winches the AUV out of the water and into the LARS docking head. The LARS A-frame lowers the AUV back into its cradle on the back deck, where it’s locked down and washed with fresh water. It’s then rolled onto the removable rails back into it’s van, where the team will download the new data, change the batteries, and get it ready for the next mission.

With each new AUV mission completed, the sonar analysts sit down with the new data and look for potential targets. They look for anomalies on the seafloor with the right size and shape, as well as material properties, to match that of Earhart’s Electra. We found just such a contact in Mary Ann’s first survey box and decided to send her back down to re-image the target. Twenty-four hours later, the team gathered in the analysis lab to see the results. The re-imaging of a promising target involves using higher frequency sonar, which gives us a higher resolution image, and taking thousands of high-resolution still photographs. This particular target turned out to be geology – not an uncommon result. Regardless of a negative find this time, with each new re-imaging survey, we became more confident that our analysis of contacts was accurate, and we knew that when the time came for the next promising target, we’d be ready.

Once the team dialed in the trawl net operations, we had a series of successful tows that kept the science crew very busy. The opportunity to collect mid-water samples in this remote location for such an extended period of time was an unprecedented one and was truly a rare opportunity for our researchers onboard. Everyone on the team was excited to welcome the strange creatures aboard and were excited to see what the next net trawl might bring up to the surface.

Following the Waitt Institute’s helicopter flight towards Howland Island, the expedition leader, Ted Waitt, brought back with him a new perspective on Earhart’s last hours of flight, and, as a result, the team wanted to add more boxes to the search grid to reflect this new point-of-view. The experience of being in the air, so close to the island, and having difficulty picking it out against the shadows on the water, brought questions into play about how close Earhart may have needed to get to Howland, before being able to positively identify it. With this new insight in hand, the operations team discussed options for adding a small search area between Howland and Baker Islands, and finally decided on two new boxes to add to the search.

Part of what made CATALYST 2 a ground-breaking survey was our ability to run two 6000-meter Autonomous Underwater Vehicles, or AUVs, simultaneously. The custom launch and recovery configuration on the back deck of the Seward Johnson allowed us to survey the seafloor 24 hours a day, 7 days a week. With our techniques and turnarounds dialed in, the CATALYST AUV team was able to leap-frog Mary Ann and Ginger across 2000 square miles of underwater terrain. We were definitely testing the limits of this gear, and they proved out to be one of the best survey tools in use today.

About a week into the survey, we suffered a severe setback. During a night recovery, Ginger slipped under the ship’s stern and the ship’s screw made contact with the AUV’s tail section. After taking a good look at the damage, the AUV team determined the thruster housing was beyond repair and that a new one would need to be acquired. It would set us back several days, if not weeks. While we worked through solutions, the team leads went over recovery scenarios in order to prevent having the same type of incident in the future. We only had one AUV left, and we had to keep her moving through the search boxes as best we could.

University of Queensland biologist, Adrian Quinn, discusses one of the trawl net’s amazing catches – a live Anglerfish. This video includes rare footage of the Anglerfish, which the CAT 2 science team managed to keep alive for several hours after bringing her up to the surface.

After nearly losing one of our vehicles to the ship’s propeller, the CATALYST 2 team re-defined AUV recovery protocols to make the entire process safer. Even then, we were sometimes forced to deploy our secret weapon – Joe Lepore.

Even though the AUVs have a titanium frame and can withstand the pressure of 6000 meters of seawater, they are not invincible. While going through images that Mary Ann took during a camera test dive, we discovered just how close she came to not only hitting the bottom, but also potentially lodging herself in the rugged terrain. It’s a difficult task to fly a 2000-pound vehicle just 60 meters above the seafloor, especially when the seafloor isn’t flat. To counter some of the inherent problems, the AUVs have a built-in collision avoidance system. A pencil-beam sonar pings out in front of the vehicle and lets it know if needs to adjust its altitude to fly over approaching obstacles. In some cases, though, the vehicle can only move out of the way so fast. Thankfully, on this expedition, we managed to walk away with only a few minor scrapes.

Sometimes with a little bit of luck, a little expertise, and a whole lot of elbow grease, what could be a major setback is reduced to a small blip in the schedule. Following a mission abort by Mary Ann, the team discovered a failure of the motor in her tail thruster. They took the thruster to the AUV lab and tore it down, did varied diagnostic tests, and, in the end, declared it unfixable. With Ginger already sitting on the sidelines waiting for her own tail assembly fix, Mary Ann was the expedition’s only operating search vehicle, and she needed to get back in the water. Harbor Branch project manager and engineer, Lee Frey, refused give up. After endless hours of testing and troubleshooting, Lee had the motor working again. What we thought was going to be our second major disaster was narrowly averted.

When an AUV returns to the surface, its acoustic tracking system relays to the team where it’s located. However, if we’re out of range when it hits the surface, we’re unable to get a good fix on her. When that happens, the AUV is capable of taking a GPS reading on its location and then making a satellite phone call to give us those coordinates. Even then, we have to account for the time it takes us to get there and the drift of the vehicle. Once on site, it turns out the best way to search for a lost vehicle is with your own two eyes.

During the expedition, researchers from the University of Queensland carried out dozens of net trawls with the help of the AUV team. Every day, more specimens came aboard and that meant that the science team had to spend hours in the wet lab organizing, identifying, preserving, and packing hundreds of organisms. There were dozens of species of animals - and most of them were creatures none of us had ever seen before. The live catches proved to be one of the highlights of the expedition. The results of these efforts will benefit organizations and research initiatives around the world, including the Census of Marine Life and scientists at more than a dozen universities. In the end, the science program was one of the most far-reaching successes of our journey.

Ginger’s collision with the ship’s propeller earlier in the week had left her sitting on deck, unable to work. While we ran survey operations with one vehicle, a replacement thruster assembly was working its way to us and finally arrived on site. The team immediately began working on getting Ginger back up and running, and by the following morning, we were fully operational again. We had two vehicles surveying nearly four miles below and we were back on track.

After more than a month of non-stop surveying, it’s hard not to get excited about a good-looking target. This particular target was incredibly promising. It was the right size and had the right shape and reflectivity for what we were looking for. After 24 hours of waiting for a more detailed picture of the bottom, we flipped through the new sonar data and a new set of photos, while the whole room held its breath.