I don’t usually do this; I don’t usually write for WHOI in the first person. Because what I have to say is not about me, it’s about the scientists and engineers, their work, and what they help us learn about the ocean. That’s what matters day-to-day and so that’s what I write about as objectively as possible. Occasionally I’ll feel brave and throw in a “we” here and there.

But today is different. Almost eleven years to the day after walking onto Atlantis for the first time to cover an Alvin cruise in the Gulf of Mexico, I made my first dive in the sub.

My bad luck with being selected to dive in the years since then had become sort of a running joke. Anytime someone asked me if I’d ever dived, I’d reply that I was the person the chief scientist would turn to at the end of an expedition, with the ship’s engines spooling up and the bow pointed toward our demobilization port, and say, “If we’d had one more dive, you would have gone down.” There’s more truth to that than not. But while I’ve born the disappointment with good humor (again, it’s not about me), not being able to dive always been a little bit of a sore spot that I’ve salved with other work. I wrote, I photographed, I interviewed, I put in late nights and early mornings. As a result, I have hundreds, if not thousands of images (some good, most mediocre) of everything that happens on the ship and in the water to get Alvin down to the seafloor and back each day. But the period in between deployment and recovery has always been a story for someone else tell.

No more.

Today, the choreographed intricacy and intense focus of launch and recovery were for others to watch and for me to experience from inside the sub. Today, when the pilot completed his checks and called up to the Launch Controller to load the sub, I made the climb to the top of the stairs leading to Alvin‘s sail. And I watched from inside the personnel sphere as the hatched closed above me.

Dozens of launches told me what to expect: the hoist and boom out over the water, the preparations by our swimmers to clear lines and confirm our vent valves, the call for final checks and request for permission to dive. But there was something I was completely unprepared for. For nearly 11 years, I’ve seen Alvin as a hulking, 43,000-pound machine. Don’t get me wrong, it’s a wonderful machine, but it’s also something so massive that it requires a 274-foot ship with 24 crew and a skilled team of engineers to move it around the world and to accomplish whatever science might call on it to do. I wasn’t prepared for this sumo wrestler to transform into a ballerina the instant it submerged.

As soon as there was water over its sail, it was as if Alvin breathed a sigh of relief, shook off its bulk and danced. It actually danced. And when we arrived on the seafloor, yes I was amazed to see the view in three dimensions as everyone told me I would be, but at some point I sat back from the porthole and laughed out loud, struck with the joy that Alvin radiated to be home again doing what it was meant to do.

This is a blog post, not a New Yorker essay, so I’ll spare you more prose and go back to being a correspondent for a moment. Bruce, Nick, and I were fortunate to be on Alvin‘s deepest dive ever. We settled on the seafloor at the end of the day and the depth readout clicked over to 5338 meters. We didn’t see much because we were on a mission to get Alvin deep-5000 meters at least, deeper if possible-so we flew high over the seafloor and, at 1 knot, relatively fast for Alvin. We surpassed the former and met the latter goal, so the dive was objectively a success. But at the surface, as you may have read elsewhere, the team found a problem with Alvin‘s syntactic foam that will put further dives on hold while the team sorts things out.

However, this fact does not detract from what is a remarkable achievement. Alvin is in an experimental phase, diving places it has never gone before, and it did so with ease and grace. It also underscores that what this team does is inherently difficult. They may make Alvin dives look easy and successful dives to the deep seafloor a matter of course, but as I once wrote during another expedition, the ocean does not give up its secrets easily. It’s a challenge-a solvable one, but a monumental challenge nonetheless.

What comes next will require long hours and considerable time and effort on the part of a talented and (temporarily) dispirited team. But I have no doubt they will make it right because that is what they do-they solve problems and make difficult tasks appear routine. I’m already seeing signs that they are turning to the challenge at hand, proposing solutions, and working through alternatives. I also have no doubt that Alvin will soon be back dancing across the seafloor because that, too, is what it was meant to do.

-Ken Kostel

After completing its 5,075th dive to a record 5338 meters (17,513 feet), post-dive visual inspection of Alvin revealed unexpected damage to several attachment points of the specialized syntactic foam used to provide buoyancy to the 43,000-pound human-occupied research submersible. As a result, further test dives have been postponed so that the sub can be thoroughly inspected, the damage assessed and repaired, and, if necessary, modifications made to ensure its continued safe operation.

Syntactic foam is made of tiny glass spheres embedded in an epoxy resin and is highly resistant to changing shape under high pressure. At no time were occupants of the sub, which carries three people at a time, in any danger as a result of the issue that were found.

Alvin is owned by the U.S. Navy, funded primarily by the National Science Foundation (NSF), and operated by WHOI on behalf of the scientific community as part of the National Deep Submergence Facility (NDSF). Alvin has been in service for 57 years, although it has gone through repeated upgrades over the years to the point that nothing of the original vehicle remains. The most recent upgrades began in March 2020 with the goal of making the sub capable of reaching a new maximum depth of 6500 meters (21,325 feet or about 4 miles), 2000 meters (6,562 feet) beyond its previous maximum of 4500 meters (14,764 feet).

Dive 5075 near San Juan, Puerto Rico, occurred as part of a series of increasingly deeper dives intended to test Alvin‘s systems and to eventually receive certification from the Naval Sea Systems Command (NAVSEA) to operate to 6500 meters. Alvin and its support ship, the research vessel Atlantis, are still in Puerto Rico, so that the sub’s engineers can better evaluate the sub and consider next steps. It is currently unknown when testing for 6500-meter dives will resume.

“Alvin‘s success over the past 50 years, as well as the success of the team that operates and maintains the sub, can be measured by the fact that its dives have come to seem routine,” said NDSF Director Andy Bowen. “But working at extreme depth remains an engineering feat of remarkable complexity. The ocean is a harsh environment, but that will not stop us from working as hard as possible to learn as much as we can about such a critical part of our planet.”

The process of certifying Alvin to dive to 6500 meters doesn’t mean the sub simply dives to its new maximum depth and that’s it. We are required to move in a gradual and stepwise fashion from 500 meters to 6500, stopping at every 500-meter increment to do specific checks of the vehicle and log our data. Originally, that meant diving to each interval individually, but we later received approval to visit several depths in succession on a single dive.

We did that on November 5, with a dive that started at 500 meters and followed the bottom down to a depth of 1500 meters in a single day. The trick in planning a dive like that is to find an area of seafloor that is steep, but not too steep, that is free of obstacles (like seafloor cables), and that is within easy reach of the ship between Bermuda and Puerto Rico. Fortunately, Mona Canyon and the upper reaches of the Puerto Rico Trench (as its name implies) are near our eventual end point of San Juan, P.R.

Our dive in Mona Canyon began with the next interval after our last dive in Bermuda: 2000 meters. From there, pilot Mike Skowronski pointed the nose of the sub down about as far as he could and drove forward, staying about 100 meters above the seafloor for most of the dive. They didn’t see much, but these dives are all about the sub’s performance from an engineer’s perspective-its science capabilities and finding interesting places to explore will come later. By the end of the day, Skowronski and his team made it to 3500 meters, setting up the next dive to take the sub below its previous maximum depth of 4500 meters for the first time-and possibly even deeper.

On Tuesday, we got word that we have a location to get back on the road to 6500 meters: the east wall of Mona Canyon northwest of Puerto Rico. And after nearly three days in transit, we expect to arrive in the early morning hours of Thursday to do some seafloor surveys with the ship’s multibeam sonar.

The canyon slopes down to the north, from the shallow Mona Passage separating Puerto Rico and the island of Hispaniola to the much deeper water of the Puerto Rico Trench. We didn’t have much information about it when it appeared as a likely dive site, but a little digging revealed that we will not be the first to dive there. Our friends on the NOAA ship Okeanos Explorer took the remotely operated vehicle Deep Discoverer in 2015 to almost the same location as our dive today. That expedition included some scientists who have dived in Alvin before, and a few quick emails to shore brought back a wealth of information about the conditions they encountered: No, currents were not a problem. Water clarity was excellent. Check the dive site for seafloor cables. Be on the lookout for lots of good geology.

Although this isn’t a science mission, it’s difficult not to dig into details about the canyon. It was formed, and continues to grow, as a result of extension occurring along the edge of the Caribbean tectonic plate where it meets the North American Plate. This is causing Puerto Rico and Hispaniola to separate at a rate of about 3-5mm/year, making it a seismically active area, as the people of Haiti can attest after the number of earthquakes that have affected the country in recent years.

Moana Canyon also bears the scars of several submarine landslides, one of which is probably the cause of a tsunami that hit western Puerto Rico in 1918 and killed about 100 people. After the 2005 Indian Ocean tsunami, this region received renewed interest as one of the few places in the Atlantic that could threaten the U.S. East Coast with a tsunami. Uri ten Brink from the USGS office back home in Woods Hole was co-author on a 2008 paper about the 1918 tsunami.

It also turns out that ours will not be the first visit to Mona Canyon by Alvin. In 1976 and 1977, trips led by Columbia University geologist Bruce Heezen-of the famous Tharp-Heezen map of the seafloor-explored parts of the canyon. In fact, when we compared our dive coordinates with those of past Alvin dives, it turned out that Dive 707 began almost exactly where we hope to end up.

Alvin will start at 1900 meters and head down to its first set of NAVSEA checks 100 meters lower. From there, the team inside will pick one of two paths down the side of the canyon, stopping every 500 meters to conduct more checks and, if successful, finish the day at 3500 or maybe even 4000 meters. That will set up the first 4500 meter-plus dive by Alvin on Friday and put us on a good course toward 6500 meters on Sunday or Monday. There is still a lot to do before then, but the weather looks good down here for diving, the sub is ready and the team is rested and ready to go.

We spent a rainy weekend tied up to the pier in Bermuda tending to some maintenance issues on the sub and ship. This morning (Monday), as we were departing between rain squalls, we got the word that today’s dive would be scrubbed. The rainy, windy conditions we’d been experiencing the past couple of days are apparently just the start of deteriorating weather over the day and into tomorrow, so what were already marginal conditions would only get worse.

It’s disappointing, especially after the sub, Alvin team, and ship’s crew performed so well on Friday. After a delayed start to the dive, Alvin dropped to 500 meters with Danik Forsman in the pilot seat, who spent the next four hours navigating along the sloping seafloor down to 1500 meters. Then, with the clock ticking on our arrival time to pick up the pilot into St. George’s Harbor, we made a near-flawless recovery and arrived at the sea buoy right on time.

The downtime in Bermuda allowed everyone to focus on little things. But those items tagged as “nice-to-haves” at this stage of the game, like preparing the 4K cameras and fine-tuning the sub’s navigation system, will be important later when a science team comes on board and wants to see not just a functioning submersible, but one that will allow them to focus on their research.

And now it appears we’ll have a little more time to finish off some of those tasks. Instead of waiting for the weather to clear, we’ve decided to begin our transit toward Puerto Rico and look for progressively deeper sites along the way, with our first good prospect coming on Thursday. At this point, it’s all about making the most of our remaining time out here. We know we need to be in San Juan the morning of November 18, to pick up the team on the science verification expedition, and that we need to progress in a stepwise fashion to our eventual certification dive, which has to be at least 6200 meters with one of our NAVSEA representatives in an observer seat.

True-to-form, however, the ocean is not letting even that simple scenario play out unimpeded, as we are currently working our way through 15-20 foot seas and 30-plus knot winds. Thankfully, the forecast is for smoother sailing in coming days.

The first two open-water dives are in the book, with the sub descending to 100 meters Wednesday and Thursday to complete Phase 2 of Sea Trials. As much as the Alvin operations group needs to exercise the sub and its systems, they and the ship’s crew also need to refresh themselves on the detailed choreography of launch and recovery.

Each dive day, the entire ship becomes focused on one goal: getting the sub and its occupants safely into and out of the water so that it can complete its mission. There are teams on deck, on the bridge, in the ship’s Dog House and Top Lab, in the engine room, in the small boat, and of course in Alvin, each with their own specific tasks to complete, often in a specific order, and sometimes requiring intricate timing. The photos here represent just a few of the people and steps that go into the critical moments leading up to launch and culminating with it touching back on deck, a cycle that has happened more than 5,000 times on three different ships (Lulu, Atlantis II and Atlantis) over Alvin‘s lifetime.

And on top of it all, the members of the Alvin Team who rotate through each of the positions on the aft deck are the same people who maintain the sub and, in the case of a few, also who pilot it. Bunk space is at a premium on the ship, so everyone has to be able to do several tasks. This trip is giving the old hands a chance to refresh their muscle memory and giving the new members (Rose Wall, Nick Ellis, Kaitlyn Beardshear, and John Dymek) an opportunity to see everything play out for the first time and begin their first steps in training toward becoming the Launch Coordinator in control of operations, or even a pilot. Even Randy Holt, the team’s Expedition Leader, is new to the team and, although he’s no stranger to submersible operations, is still learning how Alvin launches and recoveries play out.

Today’s dive to 500 meters-and deeper-is underway. It needs to cut short, however, so we can make a 5:30 p.m. rendezvous with the St. George’s Harbor pilot. We need to go back to port briefly the other day because, while lowering the ship’s ultra-short baseline tracking pole, which allows us to mark the sub’s location underwater, it got jammed inside its housing. The Chief Engineer, J.T. Walsh, had the brilliant idea of using ice to shrink the metal pole and its bearing just enough to free it, but not before a replacement unit was already on its way to Bermuda from WHOI. So, we’ll head in to pick it up and to have on hand as a back-up. We’ll also take the opportunity to do some quick repairs to the ship before heading back to sea to resume diving, probably on Sunday.

Alvin passed another critical milestone yesterday with its first untethered post-overhaul dives. The goals of the 12-meter dives were to test the sub’s electronics, communications, and navigation systems while conducting a series of maneuvers at the surface and on the seafloor. St. George’s Harbor put out the welcome mat for us with clear skies, light winds and relatively little boat traffic. And when Alvin finally swung over the stern of Atlantis, it was with a small audience of Bermuda media recording the event for print, broadcast, and online outlets.

As far as depth goes, it was a baby step toward 6500 meters. But in many ways, diving a 43,000-pound sub designed to drop four miles through the water at a rate of more than 30 meters per minute is more difficult than a deep, open-ocean dive. Alvin builds up a lot of momentum, and as soon as the sub had left the surface, it was time for pilot Mike Skowronski to put on the brakes, so to speak, and make it neutrally buoyant.

Once down, he had a long checklist to work through, item by item. His co-pilot, Nick Osadcia, and observer, Stefano Suman, were busy, as well. Osadcia is a mechanical engineer who is also a pilot-in-training, so he had to re-acquaint himself with the sub’s controls after nearly two years since his last dive, as well as help Skowronski manage his tasks.

At the same time, Suman continued improving his simulation of Alvin by matching it to an actual dive and also worked on troubleshooting bugs in the sub’s software. Prior to the overhaul that began in 2011, all of the sub’s controls were mechanical switches. Afterward, touchscreens and digital controls made their first appearance in the personnel sphere, but critical systems have retained mechanical back-ups, so that the sub could still be operated in the event of a computer failure. Although that has never happened, Suman was able to make small fixes that further improved the software running Alvin.

But the real star of the show yesterday was the sub’s Doppler velocity log (DVL), an acoustic sensor that keeps track of position by measuring the sub’s velocity relative to the seafloor. Skowronski marked the sub’s start point when it released from the main lift line and, after a full day of maneuvering over the bottom of the harbor, when Alvin returned back to the same position to be recovered, the DVL showed a difference of less than six feet from its actual location, which pleased everyone.

After a successful first off-tether dive, today we’ve moved south of Bermuda and the sub is in 100 meters of water. No one is calling this a deep dive, either, but it means we are well and truly on our way to 6500 meters.

Yesterday (Monday), Alvin successfully completed its Navy-mandated inclining test that is used to verify a ship’s stability. The test involved moving known weights between the ends of a bar bolted to the back on the sub and measuring the resulting roll angle, both submerged and at the surface. It made for a long day, but conditions were ideal. Nevertheless, it was dark when Alvin came back on deck and released three very tired divers to applause from everyone. It’s a major step in the sea trials process, and one that has allowed us to proceed today to a shallow harbor dive. If all goes well, we’ll be leaving Bermuda tomorrow for the first open-ocean dives northeast of the island and eventually to its deep dive on the edge of the Puerto Rico Trench.

Alvin hasn’t competed a deep dive since it’s overhaul, but in a computer in a corner of one of the labs on Atlantis, it dives all the time.

WHOI software engineer Stefano Suman is on board to ensure that the sub’s programming is ready to take scientists in Alvin deeper than they’ve gone before by using a technique he and other designers working on robotic vehicles like the autonomous underwater vehicle (AUV) Sentry and the hybrid remotely operated vehicle (HROV) Nereid Under Ice have used for years. He’s running simulations of dives to identify bugs and make the complex software operating throughout the sub more efficient and more resilient during a dive.

“A simulation can be 90-95% accurate,” said Suman. “With a couple of actual dives, we can get that to maybe 98% accuracy. We say simulations are doomed to succeed.”

The computer in the Hydro Lab near Alvin’s hangar contains computer code that simulates all of the sub’s sensors and other systems. With data from a previous dive feeding into it from another computer, Suman can watch his virtual Alvin respond to conditions it might encounter. When he finds problems, he fixes them without having to abort the dive and bring the actual sub to the surface, thereby losing valuable time at sea.

The only problem is, he sometimes has to work late into the night, as he did recently. That’s because for Suman to do his work, he needs a period of time with the sub powered up. But the mechanical and electrical engineers who are troubleshooting their systems at the same time, occasionally have to power down the sub in order to work.

So, after the sub came out of the water from its first official, post-overhaul dive on Thursday, Suman went to bed, got up at 1:00 a.m., and used the off-hours to work uninterrupted through the night on the sub’s navigation system. Yesterday, during the inclining test, he rode along for the entire dive to compare his simulation to the sub’s actual responses.

Unfortunately, just as he’s always pinched for time in the competing demands of mechanical and electrical engineers, Suman is pinched for time in another way, as well. He won’t be able to see all of Alvin‘s open-ocean dives because he’s committed to another expedition in the Pacific with NDSF Chief Scientist Anna Michel leaving from San Diego in a week with Sentry and the remotely operated vehicle (ROV) Jason. So, after Alvin‘s first few dives near Bermuda, he’ll have to take a small boat back to shore, then head to the West Coast for his next assignment bridging the real and virtual worlds.

Alvin‘s cameras have produced some of the most stunning and iconic views of the deep ocean ever seen. But when the sub came in for its most recent overhaul, its imaging systems were facing two major challenges. First, it would be getting new 4K video cameras made by Deepsea Power & Light, which means file sizes would nearly quadruple. In addition, the controls that scientists use to adjust camera settings and log events during a dive were spread across two separate and very different interfaces.

Before the overhaul, at the end of a dive, members of the Alvin team and shipboard technicians were hard-pressed to assemble a package of data from the dive that scientists could use to plan the next day’s dive. Processing video from the sub’s four high-definition cameras into a usable format, with metadata attached, could take six to 10 hours, meaning scientists often wouldn’t get their information until midnight or later. And with deeper dives possible after the overhaul likely to push return times even later, something needed to change.

Over the waning months of the upgrade, WHOI engineer and Alvin pilot (and sometime Sentry Team member) Mike Skowronski built a system that will process all of the sub’s video feeds as they come off the cameras in real time. That means as soon as Alvin is in its hangar, images from the day’s dive will be ready to download and copy to hard drives for the scientists-with metadata embedded-within an hour or two. In addition to making the video files more readily available, the hope is that the new system, which embeds meta data into the video files, will make them more useful to future users who are planning dives in places Alvin has already visited.

The second problem is both a technical challenge and one of human behavior. The controls for Alvin‘s old cameras were based on a video game controller, with a joystick and rocker switches with a limited set of features. To that, designers later added a separate iPad interface known as Sea Log that allows scientists to make note of their observations during a dive. Rather than switch from one control to another, however, sub pilots noticed scientists tended to pick one or the other controller and use it almost exclusively. As a result, they would either keep the camera’s zoom, exposure and position settings alone and focus on using the tablet to log events, or manipulate the camera settings to change what the camera was recording at the expense of the functions offered by Sea Log on the tablet.

To fix this, Skowronski chose to abandon the handheld units and instead expanded the tablet-based app to include a set of electronic camera controls. To merge them into one super-controller that allows scientists to manage their cameras (each person has two cameras at his/her disposal) and to log data on a single screen, Skowronski enlisted the help of WHOI software engineers Ethan Andrews during the three-month project. He’s also being joined by another WHOI software engineer, Ryan Govostes, to fine-tune the user interface and the way the app communicates with the cameras.

The hope is, with a streamlined system in place once Alvin goes back into service, scientists will not only be able to make better use of the sub’s video system, but that they will also have even more spectacular images to share with the world.

Yesterday, Alvin finished the in-water tests of its electrical and mechanical systems. It also dove briefly so the team could check the operation of the variable ballast system used to make it neutrally buoyant at the seafloor. All of that occurred while the sub remained tethered to the ship’s A-frame-and there’s still one more test before it can dive and maneuver on its own: the inclining test.

The inclining test is used to determine a vessel’s stability by moving a set of weights between two set positions on either side of the sub and measuring the resulting angle of roll. Once that’s completed successfully, Alvin can move to shallow harbor dives before we begin open-water dives of increasing depth each day until we reach 6500 meters.