The Northwest Corner (NWC) is another member in the pantheon of unique features of ocean currents. It is where the North Atlantic Current (NAC) flowing northeast along the Newfoundland/Labrador escarpment turns east at about 51°N. No other western boundary current reaches more than about 35° latitude before turning east. The reason is that the NAC is a link in the Atlantic meridional overturning circulation (AMOC); it is not a wind-driven current. But why does it turn east just there?

Just imagine for a moment that you could put a tag, no more than an inch or two in size, on a salmon when it leaves its river and heads out into the ocean coming back to the same river to spawn, a year or two later, say. The tag is designed to listen for timed acoustic transmissions from stationary ocean sound sources, the arrival times indicating its distance from the sources. When the salmon is caught and the tag retrieved, one can reconstruct where the salmon went and learn about its diving patterns. The tag is, in effect, a tiny little RAFOS receiver (see my post Aug. 13, 2023). My colleague, Prof. Godi Fischer, developed such a tag and we were able to show that it works in a first test in the Gulf of Mexico (see reference below). That is the good news.

In my previous post I describe how the Mann Eddy is driven by the North Atlantic Current (NAC). It was one of the things we learned thanks to a major RAFOS float program we started in 1993 to examine the structure and coherence of the entire NAC as a western boundary current - from where it branches off from the Gulf Stream in the south to the Northwest Corner (NWC) in the north where it turns sharply to the east. We knew it was a meandering current, but I’m not sure we appreciated how steep and stationary the meanders were.

The Mann Eddy is a body of recirculating water driven by a branch of the Gulf Stream that turns north around the SE Newfoundland Rise (SENR). Known as the North Atlantic Current (NAC), it flows at first NW curving to the north before the blocking effect of the Newfoundland Seamounts forces it to curve clockwise to the east. This rapid current whose path is defined by bathymetry to the south, west, and north, sets up what appears to be a permanent anticyclonic circulation. It is not a free-standing body of circulating water as we usually envision coherent eddies, but one trapped and defined by the moving sides, the NAC, to the SW, W, and N. See the trajectory of float #260 below.

Yesterday Shafer Smith at NYU gave a very interesting lecture about how the new SWOT satellite can resolve submesoscale features in such detail that one can estimate vorticity, strain and divergence – all constructed from the horizontal velocity derivatives ∂u/∂x, ∂u/∂y and ∂v/∂x, ∂v/∂y at O(10) km scales. This is truly amazing. Those of us who have worked in the Gulf Stream know that on occasion the North Wall at the surface has the character of a velocity discontinuity, see photo below with apologies for the poor quality, I can’t find the original. We knew were about the cross the North Wall so I positioned myself about as high as I could on the R/V Knorr so I could get a good view the shearing of the ship’s wake by the surface velocity discontinuity.

In many ways my life in oceanography began in Bermuda. It was here, in January 1966, that I first met Doug Webb. This became the start of our journey to develop the SOFAR float to explore and map ocean currents. But that isn’t the reason I’m writing this. Instead, in becoming a frequent guest at the Bermuda Biological Station, I soon became aware of the time series of bi-weekly hydrographic casts, by then 10 years old and running (and thankfully still running today). My knowledge of oceanography at that time was limited; I knew little about hydrography and water mass analysis.

This is a brief rather technical comment. It is based on a conversation with my colleague Peter Cornillon yesterday evening. The standard procedure for ADCP processing is to collect all good pings within a prescribed time window, say 5 minutes, to construct an average velocity profile. Typically, one requires at least 50% of the raw data to be acceptable, otherwise no output. Of course, one can lower the threshold to 30% or even less. Although the resulting profile will be noisier, it will be based on good data. Here I’d like to explore another approach.

For decades oceanographers have used passive upwelling light to estimate phytoplankton distributions across the world ocean. Elaborate algorithms have been developed to estimate in situ quantities based upon the spectral properties of the up-welled light. But passive upwelled light says nothing about the vertical distribution of the plankton, and of course it depends upon a clear sky to see the surface.

A year ago (Sept. 18, 2023) I mentioned how tagged lobsters recorded some curious temperature variations on the continental slope. The lobsters had been released roughly 20 km east of Block Island. Of the 32 tags released 8 were recovered; their displacement vectors show how far they wandered; three of them 120 km to edge of the shelf and beyond. Sadly, the tags weren’t tracked as planned, but the fishermen logged where the tagged lobsters were caught.

Yesterday morning I read about a recent British expedition to the Clarion-Clipperton Zone, a huge abyssal plain located between Hawaii and Mexico. They found all kinds of marine life, unknown to us until now. For example, glass sponges, which are thought to live as long as 15,000 years! The ‘pink barbie pig’ must be one of the most bizarre creatures I’ve ever seen, google it. But there are plenty of others to pick from. The point being that there is much we do not know about the deep ocean. This includes the circulation in and around these abyssal plains. In fact, I suspect we haven’t a clue besides that it is probably very, very quiet.

This curious phrase refers to an oceanic region south of Greenland whose temperature has remained rather constant over the years compared to the warming taking place just about everywhere else. It is not a warming hole; it is a hole in the warming. It has gained prominence among scientists who think this lack of warming might reflect a slowing of the Atlantic meridional overturning circulation (AMOC). This sounds backwards; you would think the cooling was a good sign?! While this region does participate in the AMOC, it produces water of intermediate density, what might be called North Atlantic Intermediate Water (NAIW), not North Atlantic Deep Water (NADW). The distinction is important.

The volume of the world ocean according to Google equals 1.3x10^9 m3 or 43 thousand Sverdrup-years. The Atlantic meridional overturning circulation (AMOC) rate is 20 Sv. Thus, for the AMOC ‘to fill the ocean’ would take something like 2000 years. This is enormously interesting, because of its similarity to the roughly 1500-year timing of the Dansgaard-Oeschger (D-O) oscillations. What are these?

As part of the WOCE Atlantic Climate Change Experiment (ACCE) we and our colleagues at WHOI deployed over 100 isopycnal floats on the 27.5 kgm-3 surface, some west of the mid-Atantic ridge (we) and others east of the ridge (WHOI). The objective was to map out the circulation across the subpolar North Atlantic in this mid-pycnocline surface. Interestingly, this surface also separates the north- and south-flowing AMOC, i.e., it is the maximum in the AMOC stream function in density coordinates. This means that somewhere somehow water must pass from the upper to the lower layer. That is exactly what some of our floats revealed, how do we know that?

Imagine a point source that releases particles that are carried away by the currents in various directions. You do this over a period of time to get a representative ensemble of particle trajectories from the point of release. The area that has been reached by the particles as a function of time can be expressed in terms of an evolving or expanding domain of occupation. This is a powerful Lagrangian tool for charting how stuff spreads out in the ocean.

POGO was a modern variant of the dropsonde technique Richardson and Schmitz used to measure currents and transport in the Florida St. At various sites across the current they would release a free-falling instrument that would sink either to the bottom or to a predetermined depth, drop ballast and return to the surface for recovery. Its horizontal displacement between release and resurface is an accurate measure of transport at that site and time. They used an extremely accurate local navigation system called Decca Hi-Fix to determine its displacement. This is in the 1960s when the art and skill of measuring ocean currents was still in its infancy (and in some ways still is).

Cold core rings shed from Gulf Stream meanders dominate much of the eddy activity to its south. Rings can break away almost anywhere along the stream east of roughly 70°W. Once formed they will drift west toward Cape Hatteras near where they will eventually, in weakened form, rejoin the stream. Thanks to the Oleander’s weekly roundtrips between Bermuda and New Jersey we have an extensive archive of ADCP velocity transects of numerous rings, some recently formed, others older and from farther east. These enable us to see how they form and age. It’s a bit like stroboscopic sampling.

This may sound crazy, but I wonder if we couldn’t take advantage of the planar structure of modern ADCPs and mount the whole thing external to the hull rather than have to machine and weld in an expensive sea chest and coffer dam to house the ADCP as we do today? It would be less costly and potentially easier to service if necessary.

I’m thinking specifically of the 38 kHz ADCP, which is roughly 1 m in diameter and weighs roughly 210 kg in water. That’s a lot for sure, but if outfitted with a suitable flotation collar it could be made slightly buoyant for handling in water. Divers could then guide it to its location along the hull. Once in place the collar can be removed (or not depending on design approach?). Even better, since the ADCP will not experience great pressure perhaps the body could be 3-D printed or made of carbon fiber epoxy instead of marine brass; this would reduce its weight enormously. It could be attached magnetically or bolted to the hull. Power and communications would go through a wet-mateable plug in the hull. The massively parallel communications cable used today would have to be serialized to reduce the number of pins. Just a thought.

To justify such a redesign would of course require an identifiable market = perhaps that is something Science RoCS can help develop? In any event this suggestion is one of many possible approaches to improved coverage of ocean currents from merchant marine vessels. See my January 15 blog.

As I wrote a few blogs ago, it was a rude awakening Charlie Flagg and I got when we put to sea on the Norröna in early January 2006. The ADCP was producing no useful data. It quickly became evident what the problem was, and we knew then and there it was going to be a major challenge: bubbles from the huge bow thruster openings blanketed the hull wiping out the acoustics. As I also wrote, Charlie saved the project with his plan to build a fairing that would deflect the bubbles around the ADCP aperture. Later on I contracted with a student to explore the possibility of creating a bubble-free spot with the help of chines. See Charlie’s website (listed below) for a detailed discussion of all those activities. These include the construction of a permanent fairing, using chines to create upwelling ahead of the ADCP, and the use of magnetically attached cameras to get a better grasp of the bubble issue.

In the fall of 2022 I was a guest at the marine robotics group at the Royal Institute of Technology in Stockholm. My discussions of the bubble issue led me to propose a capstone project for two senior students. The first one was to carry forward Charlie’s camera idea by making it more streamlined and programmable, and readily deployable to explore and identify areas free of bubbles. The other project was to continue the numerical modeling studies about fairings and chines, the objective being to evaluate the pros and cons of both, i.e., how to create bubble-free windows or spots in the hull with as little drag as possible. These proposals didn’t fly then, but much has happened since then, and both projects are well underway now. This is welcome news about an incredibly important issue for ocean observing science. Why do I say this?

The answer is simple. The day will come when the oceanographic community begins to realize what an incredible asset the merchant marine fleet is for ocean observation. It will ask how we can make these platforms as effective as possible. The single most important issue facing those of us interested in the water column is how to create bubble-free spots in the hull for acoustic and optical instrumentation, and for future telemetry and communication needs. That is what the first of the two proposals was about: cameras to delineate possible bubble-free areas in the hull, most likely up near the bow. Such a camera has now been developed and deployed on several vessels. This is a wonderful development. We are going to learn a lot with this new tool. The second project seeks to evaluate the effectiveness of two approaches for creating bubble-free spots in the hull. One is the use of fairings, hydrodynamically smooth structures that extend a distance under the hull so that instruments in it sit below the bubbles streaming along the hull. The other is to use chines - vertical rails that gradually increase in separation forcing a divergent flow along the hull and so brings up clear water from below. In short, fairings extend our reach below the hull; chines bring clear water up to the hull. The camera will give us essential information on the thickness of the bubble layer and thus guide us on which approach may work best for a given application. This is an ongoing research initiative of enormous importance because armed with this knowledge we will be able to deploy and operate instrumentation on vessels far more confidently than we have in the past. Exciting times!

http://po.msrc.sunysb.edu/Norrona/

The Norröna was the 3rd ship Charlie Flagg and I were to equip with an acoustic Doppler current profiler (ADCP). It operates out of the Faroes to Denmark and Iceland. The beauty of this vessel’s route is that it spans all warm water entering the Nordic Seas and hence the climatologically crucial deep branch of the Atlantic overturning circulation (AMOC). Thus, the case for instrumenting this ship with an ADCP was a no-brainer. The suggestion to do this was likely made easier from our earlier successes on the Oleander and the Nuka Arctica. But as I wrote earlier, we were in for a nasty surprise. The two huge bow thruster openings blanketed the hull with bubbles that wiped out the acoustics. Fortunately, we were able to turn things around thanks to Charlie’s initiative to put the ADCP in a fairing. Although less than perfect we started to get valuable information on currents in the Faroe-Shetland Channel (FSC). Curiously, they didn’t fit with what we expected.

Historically, the FSC was seen as the main route by which warm, salty North Atlantic water entered the Nordic Seas, most of it as a well-defined boundary current flowing north along the eastern slope of Rockall Trough. It is clearly identified in the famous 1909 Helland-Hansen and Nansen book ‘The Norwegian Sea’. The problem was that the ADCP data were indicating a substantial flow south on the western side of the FSC, this was not expected. This bothered me to the point that I wondered whether we had a measurement problem.

We hadn’t fully solved the bubble problem. Thanks to following winds and seas we got much better data returns steaming from the Faroes to Shetland than vice versa. Might there be a bias creeping in due to the uni-directional sampling? But if so why localized to the western side of the FSC? I discussed this with Sandy Fontana, who managed all the data processing, telling her that we had an issue with the data. But she reminded that she calibrated the ADCP whenever the vessel is in shallow water. This is done by integrating vessel speed and heading acoustically over the bottom and comparing that distanced traveled with travel according to the GPS. It’s an extremely accurate procedure. It started to look like our findings were real. With time we also got data steaming westbound, and these gave the same results: there is a substantial flow south in the western half of the FSC. This of course reduces the net amount of water entering the Nordic Seas through the FSC. The reason it came as a surprise is that this flow is less baroclinic and thus had remained ‘invisible’ to standard hydrographic methods. Look at Figures 2 and 5 in our first report (ref. attached below).

As the project continued we started getting velocity data along the Iceland-Faroe Ridge (IFR). These confirmed what was already known: that the lion-share of North Atlantic water entering the Nordic Seas does so here and not in the FSC. And as we noted earlier, all warm water entering the Nordic Seas comes via the Iceland Basin, not along the eastern slope or through Rockall Trough.

But here’s my point, if you think there’s a possibility of the AMOC slowing down, there is no better place to detect any such trend than here at the FSC and IFR. Bogi Hansen and colleagues got that that right with their ongoing project to monitor overflow through the Faroe Bank Channel. Similarly, that was the power of the Norröna. Sadly, for technical and financial reasons this program was shut down. If I were younger and knowing now how to address the bubble problem, I would make every effort to restart it. Standing by!

Physical oceanography has come a long way over the last 50 years. Starting in the 1970s we’ve deployed moorings with current meters across the world ocean. Thanks to these efforts we have far better ideas about mean flows, time scales of variability, and eddy activity in the ocean. Similarly, Lagrangian floats and drifters have mapped out pathways of fluid motion. They have given us much information about meso-scale activity, dispersive processes, and they’ve shown us the enormous role bathymetry can play in steering flow at depth. Numerous floats have been trapped in coherent vortices for varying periods of time, indicating that these can persist for long periods of time – even measured in years. In more recent years we’ve come to rely upon the global Argo float array to give us valuable information on the hydrographic state of the ocean. But just as with moored instruments and floats, Argo does not have the capability resolve currents and fluxes and how they vary.

Flying west across the North Atlantic is always in daytime so there is plenty to be on the lookout for if like me you have a hard time sleeping. Take a look at this:

Based on other pictures in this old roll of film and my passport it looks I was returning from Europe in late September 2003. Usually after flying Greenland there is little to see until we reach Labrador. But this time we got a wonderful view of currents and swirls revealed by a vast scatter shot of ice as far as the eye can see in the Labrador Sea. What a thrill it is to see these things - that sense of excitement when something out of the ordinary passes in front of your eyes!

Little did we know about the whopping headaches that lay ahead when we started preparing for this project. The Norröna was the third commercially operated vessel to be instrumented with an acoustic Doppler current profiler. Her route was exceptionally strategic, operating out of the Faroes to Denmark and to Iceland, she crosses all warm water entering the Nordic Seas. We, Charlie Flagg and I, had previously instrumented the Oleander and Nuka Arctica, both container vessels, to measure poleward flow by the Gulf Stream system between New Jersey and Bermuda, and farther north by the North Atlantic Current between Greenland and Scotland. The Iceland-Faroe-Scotland choke point was an obvious next step. We had barely left the dock in Hamburg when the problems arose.

It was easy to worry that something was wrong about the installation. The Norröna is a 35,000 ton high-seas ferry with over a mile of lanes for cars and trucks! Installing the cable that connects the ADCP in the hull with the deck box, computer, and navigation equipment up topside was a huge challenge. On research vessels wires and cables are visible everywhere - no big deal, but on the Norröna it had to be a tidy installation: the 100 m long cable had to be strung through decks, corridors and firewalls hidden from sight, a major undertaking the Blohm und Voss shipyard in Hamburg managed to complete on a very tight schedule. It didn’t take long before we realized it wasn’t the equipment but air bubbles drawn underneath the ship blocking the acoustics that was the problem. This came as a quite a shock, unlike anything we had experienced on the other two vessels. And not one readily rectified.

To Charlie’s everlasting credit, he persuaded the vessel operators to allow us to put a fairing around the face of the ADCP; the idea being that it would deflect the bubbles away from the acoustic beams. It was a lightweight aluminum rather frangible structure that didn’t last long, but long enough to prove that a fairing would help. Based on this we sought permission from the Smyril Line, the owners of the Norröna, to install a permanent fairing. This was approved and the installation led to significant improvement in data returns. Captain Jógvan i Dávastovu deserves our unending thanks for his sustained interest and support in this effort.

The bubbles come from the two huge bow thruster openings. These sweep down volumes of air each time they breach the surface which happens as soon as there is any swell. Even on calm days there was often swell between the Faroes and Iceland resulting in poor data returns along that route.

We contracted with the ship designers a hydrodynamic study of flow around the hull to identify, if possible, a better location for the ADCP. They suggested two sites, one near the bow, ahead of the bow thruster openings, and the other in the skeg between the propellors. Charlie and I explored the bow but couldn’t find a level flat spot for the ADCP. The alternative was to build a blister, a flat spot that would protrude from the hull. That seemed risky, and costly so we opted for the skeg instead. We were a bit nervous that being close to the propellors would cause other problems, but in fact the move to the skeg improved performance noticeably – albeit not as much as hoped for. A year later the ADCP quit operating - after more than 12 years of 7/24/365 service. (Can’t complain about that!) Sadly, technical issues and a lengthy covid interrupt coupled with continued bubble issues forced a termination of the program. Sad, yes, but we learned a lot - more on this later.

Charlie’s website provides a lot of information on the challenges we faced. I recommend it highly, it’s a good read about all aspects of the Norröna operation: http://po.msrc.sunysb.edu/Norrona/

The ADCP operation on the Nuka Arctica led to many contributions. The May 2 blog was about the discovery that the Irminger Current in fact consisted of two parallel flows along the western slope of the Reykjanes Ridge. The May 7 and 9 blogs showed some beautiful examples of acoustic backscatter patterns shaped by the diurnal migrations of zooplankton. As noted before, the Nuka Arctica operated on a 3-week schedule between Greenland and Denmark. Eastbound she sailed along a constant latitude, westbound along a great circle, and at times also made a stop at Reykjavik. Thanks to these three routes west from Denmark as well as the Norröna operation between Shetland, the Faroes and Iceland Katelin Childers could construct an accurate map of the Iceland Basin of upper ocean circulation and its linkages with the Nordic Seas.

Given our capability to ballast RAFOS floats to drift on a specific isopycnal surface in the main thermocline, an obvious application that came up was to extend this skill to measure stretching vorticity or local stratification at the float. The approach was very simple: change the volume of the float ever so slightly so it would float up and down to measure the depth of neighboring isopycnal surfaces to get their vertical separation.

To do this Jim Fontaine and Skip Carter developed a small motor-driven volume changer (vocha) that under microprocessor control would repeatedly first increase, and then symmetrically decrease the volume of the float by the same amount, causing the float to first equilibrate on a shallower surface and then on a deeper surface before returning to its resting surface. We learned a lot from this first study.

The vocha and the microprocessor control worked as planned. The lower panel of the figure shows how the float first ascends then then descends toward neighboring isopycnal surfaces. The plot shows that we didn’t allow enough time for the float to equilibrate before starting the next phase of the measurement cycle. In retrospect, we should have given each step more than enough time and then based on these first results decide how much to shorten the steps. Above all the time to get to the lower level was much too short. The reason for cycling through the up-down steps as quickly as possible was to minimize the effect of vertical shear: to get the float back to where it would have been had it not made these vertical excursions. During the 5-day study the float crossed the Gulf Stream to warmer waters such that the isopycnals deepened by about 250 m.

The upper panel shows temperature. The parking (or middle) temperature increased from 8 to 8.5°C, but the main change is between hours 60 and 80 so I interpret this as a drift into water with slightly different T/S characteristics on that isopycnal. This didn’t happen for the top and bottom temperatures. During these ~200 m excursions the temperature ranged from 10.5 to 7°C indicating a 3.5/200 = 0.0175°C/m stratification.

When this study was done, in 1988, the bandwidth for data transmission was limited. Today, with Iridium, the float can collect and transmit vastly larger volumes of data. One can imagine equipping the float with other sensors to zoom in on local property variations such as light level, oxygen, local backscatter, even photographs!

This paper gives a detailed report on that study.

Rossby, T., J. Fontaine and E.L. Carter, Jr. The f/h float-measuring stretching vorticity directly. Deep Sea Res., 41, 975-992, 1994.

Thanks to the high density of backscatterers at depth the ADCP reached to greater than 400 m. It’s fun to speculate about what determines these distribution patterns.

The reason acoustic Doppler current profilers (ADCP) work, whether on moorings or on ships, is thanks to the kind cooperation of zooplankton and fish populations that inhabit the ocean. These little fellows echo back the acoustic signals emitted from the ADCP, and it is from the Doppler shift in frequency of the echo one can determine the relative velocity between the instrument and the zooplankton. But here’s the thing, besides measuring the Doppler shift, the strength of the received echo is also recorded, this is a huge cache of information we pay little attention to (or don’t know how to use). But, for now at least, they allow for pretty pictures, here’s a wonderful example.

The container vessel Nuka Arctica operates on a 3-week schedule between Nuuk, the capital of Greenland, and Aalborg in Denmark. On her east-to-west transits she sails along a great circle path across the subpolar North Atlantic between Scotland and Cape Farewell. In the early years she was equipped with a 150 kHz ADCP that sometimes could reach to 400 m depth thanks to the high concentrations of backscatterers.

The attached figure shows a summer and a late fall crossing from east to west. The lower panels show the vessel tracks –virtually the same great circle route for both crossings. The upper panels show backscatter strength. The plotted intensity has been corrected for signal loss due to spreading and attenuation. While we have no information on absolute signal intensity, the instrument is stable and gain settings are left unchanged so differences between summer and fall (color scale in decibels) reflect real changes in the ocean. The black lines at the top of the panels indicate when the sun is below the horizon: nights are short in summer and long in winter.

While this is well-known stuff, I still amaze at how the zooplankton (and other creatures) rise to the surface to seek food during the night. The rapid change in depth at dawn and dusk of these tiny creatures is incredible. Distances are in km from the axis of the Reykjanes Ridge. These are snapshots. To my knowledge there has been no attempt to conduct a systematic study of these distributions – as a function of time of day, season, and of location (Irminger Sea vs. Iceland Basin). Note the higher backscatter concentrations at depth in the Irminger Sea; this remains true in winter as well. The vastly lower backscatter intensities in winter presumably reflect the loss of zooplankton and only the fish (mainly myctophids?) remain. Is this correct, I’d be happy to host a guest blog on this. The depth variations during summer daylight may reflect varying cloud cover (we should include solar irradiance in the measurement program!). It’s tempting to suspect that some of the spatial variability (local maxima or minima) might be related to coherent eddies acting like bottles preventing mixing and dispersal?

Given that the ADCP is a very stable instrument, a detailed study of these distributions might be instructive (we have over 3 years with the 150 kHz and similarly with a 75 kHz ADCP). I constantly get the feeling there is a lot of both dynamical and biological information here. In the next blog I’ll zoom in on the summer transect.

Here’s an excellent read: Sidebar 2: ACOUSTIC BACKSCATTER PATTERNS by Jaime Palter, Lauren Cook, Afonso Monteiro Gonçalves Neto, Sarah Nickford, and Daniele Bianchi. Page 91 in the September 2019 issue of the TOS magazine Oceanography.

The Nuka Arctica was a container vessel of the Royal Arctic Lines that operated out of Nuuk in Greenland to Aalborg in Denmark on a 3-week schedule, occasionally making a stop at Reykjavik on its westbound legs. Starting in 1999 we operated a 150 kHz acoustic Doppler current profiler (ADCP) for four years. Thanks to the high density of zooplankton backscatterers it reached to O(300) m depths. Starting in late 2012 we operated a 75 kHz ADCP that reached much deeper. Despite traversing some of the worst seas in the North Atlantic, the ADCPs worked well, in no small part thanks to the deep draft the vessel. Thanks to this operation we discovered that the Irminger Current flowing north on the western side of the Reykjanes Ridge consisted in fact of two separate yet parallel flows!

When you fly west from Scandinavia try to get a window seat on the right side away from the wing. The reason is that they (especially Icelandair) often fly along the southern coast of Iceland continuing southwest across the southern tip of Greenland. You might see Greenland from a left side window too, but you’ll be blinded by the sun. If it is a clear day I guarantee you’ll be in for a big treat.

I’ve often wondered why the oceanographic community shows so little interest in working with the merchant marine (MM), considering that the MM could give us incredibly broad and sustained coverage the high seas water column, what we could never expect to achieve on our own. Is it because their vessels are kind of intimidating (surely they wouldn’t let me work with them?), is it because they operate on rigid schedules (UNOLS is far easier to work with), or because it is easier to explore the vast archives of satellite and Argo data (plenty of excellent science there)? A few of us do in fact work with the MM, but we tend to view them as extensions to what we do on UNOLS vessels; we have not explored the possibilities the MM might open up for ocean observation. Earlier, I wrote about the need for scanning currents to greater depths on M-vessels. Here’s another example.

Surely the most conspicuous tipping point in the recent climate record must be the incredibly rapid warming that took place at the Bølling-Allerød event and the end of the Younger Dryas, 14,700 and 11,500 years ago, respectively. In just a few decades air temperature over Greenland increased some 15°C and precipitation nearly doubled. It is generally understood that the north-flowing Atlantic meridional overturning circulation (AMOC) must be playing a major role because it is a nearby and major source of heat and moisture at high latitudes. We also know that prior to these two events the deep North Atlantic was filled with what is called Southern Ocean Water meaning the entire Atlantic Ocean was filled from the bottom up to nearly 2 km depth with dense water from the Antarctic. The water at shallower depths below the main thermocline was formed in the subpolar North Atlantic, which during glacial times was much larger because the Gulf Stream flowed essentially straight east toward southern Europe. So, what happened?

When we talk about ocean exploration we think of expeditions to study the ocean below the sea surface with deep diving submersibles and remotely operated vehicles (ROV). These continue the famous expeditions that started in the 1800s, most notably the 1873-1876 Challenger Expedition. When I walk into the lobby at GSO I often stop by the monitors to marvel at the variety of marine life these modern expeditions reveal to us. But there are other aspects of the deep ocean waiting to be explored, invisible to us but no less real. I’m thinking of how little we know about deep ocean currents and their role in shaping the ocean and world above.

My best friend Bo Ribbing and I were camping on the island of Gotland in the Baltic in anticipation of the total solar eclipse that was to take place on June 30, 1954. We had bicycled from Stockholm to Nynäshamn to take the overnight boat ride to Visby, the capital of Gotland. Not sure, but I think we arrived on Saturday June 26. We had planned to get supplies upon arrival, but all the stores were closed. The reason was Friday was Midsummer eve. Saturday was a holiday, and in those days holidays were holidays. We stopped by my uncle Folke who was vacationing in the city and he helped us out so we could start our bicycle trip around the island.

Given the extraordinary coverage of the ocean surface from satellites, I find it rather curious that there is so little interest in the oceanographic community in the merchant marine (MM) with its potential to provide both sustained ground truth and to extend their vision below the surface. Vessels traverse all oceans. We work with a few of them to drop XBTs and release Argo floats. The data from these give us valuable information about upper ocean heat and salt, and gyre-scale circulation at 1000 m depth. With respect to the deep ocean the decadal Go-Ship program plays the key role with its global surveys a wide variety of physical and biochemical parameters. All together these surveys inform us about the state of the ocean and how it has changed over time. But they tell us little if anything about what the ocean is doing, it is after all a highly dynamic system. We modest effort we can do far better.

If you search the web for information about the Ekman layer you’ll find plenty of entries. You’ll learn that when the wind acts on the ocean, the surface water doesn’t drift downwind, but at an angle to the right due to planetary rotation. In turn, the surface water rubs on the layer underneath causing it to move to the right of the action on it. And so on to greater and greater depths. The result is a right-turning spiral of decreasing velocity with increasing depth. The vertical sum or integral of these velocities is a transport at right angle to the wind. This wind-driven transport is a well-known phenomenon; it is easily observed and plays a crucial role or link in how the atmosphere drives the oceans. In light of this it is curious that the Ekman layer velocities are so small, measured in cm/s compared to ocean currents which can range up to 2-3 m/s in strong currents such as the Gulf Stream. The winds also kick up surface waves and turbulence so while Ekman transport is evident, we wondered whether the Ekman spiral is detectable, and can it be seen in the shipboard data we had been collecting? Amazingly, the answer is yes.

We all have those moments that remain frozen in our minds, moments that make you stop in your tracks and take notice. Like in earlier notes where I wrote about where angels come from, or the discovery of eddies in the 1539 Carta Marina. Another such ‘wow’ moment occurred shortly after I had returned from a sabbatical leave in spring 1974. I was at WHOI visiting Peter Rhines and his postdoc Howard Freeland when Howard showed me a plot he had just made using the SOFAR float data from MODE, it was a stunner.

We’ve always known that the seasonal cycle lags the length of the day. But by how much? To find out I fit a sinusoidal curve to the monthly average daytime highs for Providence, RI assigning the average temperature to the 15th of each month. The fit was rather tight with a 0.34°C standard deviation. The annual average = 15.66°C ranging from 3.26°C (Jan. 22) to 28.06°C (July 24). While solar equinox is 11:06 PM March 19 this year, we must wait another full 5 weeks until ‘thermal equinox’ on April 24. Is there a word for this day?

Since I grew up in Stockholm I had to check the numbers there and found that the thermal lag was 2 days less, i.e., 33 days. Given Stockholm’s maritime climate I had expected a greater lag, not a lesser one (probably not significantly different). The fit wasn’t quite as tight, 0.55°C. The annual average daytime high = 9.83°C ranging from -0.67°C to 20.32°C. The climate tables didn’t show the years used. Given warming in recent years the average temperatures are probably on the low side. This is no doubt the case for Providence as well.

Addendum: At lunch today my colleagues suggested Stockholm may not be as maritime as I thought and suggested I consider a station on Ireland. They were right, ‘thermal equinox’ arrives even later there at May 6, almost 7 weeks after vernal equinox!

I also checked climate data for Bismarck, ND thinking that is about as continental as one can get, but amazingly. ‘thermal equinox’ is only 4 days earlier than here on April 20. Of course, the thermal swing is huge, daytime high is -4.2°C January 19 and 30.0°C July 19.

https://en.wikipedia.org/wiki/Seasonal_lag

Until the 1970s hydrography – measuring temperature (T) and salinity (S) throughout the world ocean – was the one and only tool oceanographers had to determine ocean circulation. Qualitatively, one could infer pathways of spreading from the distribution of heat and salt, and quantitatively one could use T/S to calculate the pressure field and construct a reasonable picture of the upper ocean circulation relative to an assumed quiet deep ocean. We could not measure directly what the ocean was doing. This all changed rather suddenly with the advent of the Mid-Ocean Dynamics Experiment (MODE) and other similar studies in the 1970s.

If you are interested in how neutrally buoyant floats behave in different situations then you’ll find the Prater diagram very instructive. By different situations is meant whether the float is ‘isobaric’ or isopycnal, and whether there is upwelling, vertical and/or lateral diffusion. While not a serious issue the reason for the quotes on isobaric is that it is impossible to design a float to passively remain at constant pressure because temperature or salinity changes will affect water and float density differently. Fortunately, we can turn this fact to our advantage to render a float isopycnal.

Last Sunday the New York Times had an interesting article about a comet that struck Earth somewhere in the upper mid-west about 12,900 years ago. The impact led to megafaunal extinctions, and the soot from extensive forest fires blotted out the sun. The reference below gives a detailed account of the evidence for the impact. Curiously, while the title mentions Younger Dryas (YD), the article says nothing about the connection to the YD other than that it appears to start at the time of the impact. The NYT article expands on this and mentions the role of Lake Agassiz, a huge expanse of glacial meltwater stretching across the northern tier and southern Canada. The comet impact shattered the ice dam holding back the lake causing the run-off to flood the North Atlantic Ocean with the resulting fresh water interrupting the flow of warm salty water into the northern North Atlantic where it normally cools and sinks, what is known as the Atlantic Meridional Overturning circulation (AMOC).

Imagine you are looking at a tank of water on a rotating table. The water is at rest in the tank, meaning it is rotating at the same rate as the table. Now increase the rate of rotation of the table slightly. Relative to the table the water is now moving backwards by the same about since it doesn’t know that the table is spinning faster, it continues to spin as before. We have a similar situation on our rotating planet. While the planet has a 24-hour rotation period, ocean currents respond to the component of rotation in the same direction as gravity. The effect of planetary rotation is therefore strongest at the poles, and absent at the equator; it varies as the sine of latitude. In oceanographic terms this means that if a fluid parcel is displaced poleward where the effect of rotation is stronger, the fluid parcel not knowing this, will begin to rotate clockwise. This is exactly what happened to a cluster of floats in polyMODE.

The 1978 polyMODE float program treated us to a lot of new goodies. Besides the lively little spinner described in a recent blog, another first-class treat was the evolution of a cluster of floats at 1300 m depth. At this depth, near the zero crossing of the first baroclinic mode, fluid motion is largely barotropic; acting as if representing the water column as a whole. The evolution of this cluster gave us a superb illustration of the conservation of not just relative vorticity but potential vorticity. At first the cluster moved NE and in so doing it started to rotate in a clockwise direction. But then the cluster changed direction and went SW over ~200 km, and in so doing arrested the clockwise motion and began to rotate in the opposite direction. But then the cluster again reversed direction and drifted back north again, and again the sense of rotation changed back to clockwise! This is just like being on that rotating table and watching the water while gradually increasing its rotation rate slightly, then decreasing, and then faster again.

But the match between cluster rotation and change in latitude, i.e. relative and planetary vorticity, wasn’t perfect. Jim Price, who was conducting this study, noticed that the bathymetry shoaled from west to east such that the cluster drifted into slightly shallower water in NE and vice versa. When he took the depth variations, which in effect amplified the planetary effect slightly, into account he found much better agreement. This is what is known as the conservation of potential vorticity. The period of this NE-SW motion was about 2 months, consistent with what to expect for a barotropic planetary wave. The paper listed below gives a full account of this study.

In oceanography we don’t have the luxury of the laboratory physicist who can design and execute experiments in a controlled fashion; instead, we must work with what observations can tell us. By pure serendipity this study may have been as close to a design experiment as one can imagine, and what a treat it was!

The figure below illustrates the full 60-day cycle of cluster movement in ~12 day steps in geographical coordinates in the left box, and relative to each other in the right box (where the cluster translation has been removed). The left set of panels in each box shows the floats while translating, the right panels when reversing direction. The right box shows clearly their clockwise motion in the NE and anticlockwise motion in the SW. Beautiful!

Price, J.F. and H.T. Rossby. Observations of a barotropic planetary wave in the western North Atlantic. J. Mar. Res., 40 (suppl.), 543-558, 1982.

The concern about the state of the AMOC continues today. It is appropriate to be concerned, but fears that it will shut down are greatly exaggerated, and there is no evidence that it has slowed in recent times. First, the presumed mechanism for shutting it down, fresh-water run-off from Greenland, besides being a gradual process, takes a circuitous route around Greenland and the Labrador Sea and gains salinity such that when it meets with the warm, salty North Atlantic Current, the northern branch of the Gulf Stream, the dilutionary impact is muted. Second, the AMOC involves two overturning pathways, one is via the Irminger Sea, and the other through the Nordic Seas. We all know this, but it is a crucial distinction that needs to be better recognized because they play quite different roles in major climate change.

Oceanography as a modern science came of age in the 1970s during a period known as the International Decade of Ocean Exploration or IDOE for short. Major surveys of oceanic variability with moored instrumentation were initiated, MODE and polyMODE were two such endeavors. The SOFAR float group contributed substantially to both. In MODE the floats were deployed at sound channel depth, which also corresponds closely to the zero crossing of the first baroclinic mode. This is not a trivial point as we now know that at the meso- and larger scales the vertical structure of ocean currents can to a large extent be accounted for by the barotropic and first baroclinic mode - meaning the floats were reflecting the action of the barotropic field. In polyMODE a second layer of floats at 700 m to explore action in the main thermocline. Sometimes you’re in for a very big surprise.

Almost all oceanographic research vessels are equipped with acoustic Doppler current profilers (ADCP), some reaching deeper than others. A few merchant marine vessels are similarly equipped. One of these, the Oleander, has been scanning upper ocean currents between Bermuda and New Jersey on a weekly basis since late 1992. The main purpose of this ongoing effort has been to monitor the Gulf Stream, and, needless to say, we have learned a lot about its structure and variability over time. But along the way the data we have collected have had much else to tell us. One of these topics is that of coherent vortices, i.e. spinning bodies of water. Besides the Gulf Stream the Oleander was also at times slicing through warm and cold core rings. What could we learn about them?

August Pettermann published a beautiful chart of July 1870 sea-surface temperatures. I’ve attached it below. Copy it and blow it up. You’ll see he is fascinated by the Gulf Stream; he writes Golfstrom in several places, even far north in the Barents Sea. This is vivid testimony to the European fascination with the Gulf Stream and its role in their climate. This, of course, lives on today with our concerns with the state and health of the Atlantic meridional overturning circulation (AMOC). In Pettermann’s days they weren’t aware the distinction between the AMOC and the Gulf Stream, but today we know the Gulf Stream plays a dual role as a link in the AMOC and as the western boundary of the North Atlantic wind-driven circulation. The wind-driven part cannot shut down – the physics responsible for the westerlies and trade winds were active in glacial times as well. The AMOC may shut down, but not the Gulf Stream.

Immediately following the very successful Mid-Ocean Dynamics Experiment (MODE) discussions developed between the MODE leadership and oceanographers at the Shirshov Institute of Oceanology in Moscow, USSR about a possible joint study. They had recently completed a somewhat similar study in the eastern Atlantic called POLYGON. To explore this, Prof. Allan Robinson at Harvard, Dr. Bill Schmitz at WHOI, and I at Yale were invited to Moscow that fall (1973) for initial discussions and planning for what became known at polyMODE. I was the junior member of the group and included, I think, because the Shirshov people were very interested in our Lagrangian approach to studying oceanic motion using SOFAR floats. But, in fact, almost all planning pivoted around where to site polyMODE, the details of current meter array, and the hydrographic surveys.

Look at almost any map from the Renaissance period and you will realize the importance of the oceans for travel and commerce. A voyage at sea was far easier than a journey over land. The maps bear witness to this: the most accurate and detailed aspect of many of these maps is surely the coastline and the location and names of ports important for navigation and trade. But if the maps inform about the coastlines, they remain conspicuously quiet about the ocean, which almost always appears as a void, nothing of note. But there is one truly remarkable exception, namely the 1539 Carta Marina, a map of the Nordic countries published in 1539 in Venice by an exiled Swedish priest, Olaus Magnus. In this physically huge map, made up of nine wood lithograph prints, Olaus draws little descriptions of human activity, of hunting and fishing, swimming and ice skating, sleigh-riding and skiing in the mountains.

In the middle of the Iceland Basin there’s a region with very high eddy kinetic energy (EKE). Curiously, it is quite localized, it doesn’t connect or relate to boundary currents, it doesn’t seem to be fed by a series of eddies as with the Lofoten Basin Eddy, i.e. eddies being shorn by the unstable Norwegian Atlantic Current (LBE, see January 11, 2024 blog). While it sits in the middle of the North Atlantic Current (NAC), the NAC is not a well-defined baroclinic current that meanders and breaks off eddies like a Gulf Stream, say. Sometimes there is a well-defined anticyclonic eddy, as such known as the Prime Eddy. It can last for months, maybe even longer, but so far as I know there is no evidence that it is permanent in the sense of the LBE. See Martin et al., 1998 for an excellent description of what has become known as the Prime Eddy (Plankton Reactivity In the Marine Environment). So what’s going on?

We’ve come a long way. Fifty years ago we were just entering the era of modern physical oceanography – meaning that was when we for the first time were able to accurately and systematically see what the ocean was doing, i.e. measure ocean currents. Arrays of current meters could map out currents in defined areas and clusters of floats could reveal patterns of fluid motion and how water spreads. Two decades later NASA and partners started orbiting satellites that could map out surface currents around the entire globe. Argo floats, envisioned in their original form as a tool for mapping ocean currents at depth, have evolved to become an incredible global hydrographic resource. But sustained study of ocean currents at depth remains a major challenge. Mapping currents everywhere is out of reach, but with acoustic Doppler current profilers (ADCP) we can reach 600-1200 m depth. By putting ADCPs on merchant marine vessels we can study oceanic variability along routes of interest. Thanks to these operations we have learned much about Gulf Stream variability, and currents across the subpolar North Atlantic, and in the Antarctic. But profiling is limited to the upper ocean, the deeper ocean is still out of reach. I am convinced we can do better. Why and how might we go about this?

An exciting day! In summer 2010 I was on a cruise with my Bergen colleague Henrik Søiland in the Norwegian Sea. After fulfilling the primary objectives of the cruise, we had a day to spare to take a closer look at an area where we knew from past surveys that the water is notably warmer and saltier, and to much greater depths than the surrounding ocean. This is interesting because a pool of warm water reaching to great depth must exhibit clockwise motion, most likely in the form of an anticyclonic eddy. We frequently see eddies near strong currents like the Gulf Stream or the Kuroshio Current in the Pacific Ocean. These eddies result from pinch-offs of the meandering current. But why a pool of warm water all by itself over the deepest part of the Lofoten Basin in the Norwegian Sea? What did it look like? In just one day we learned so much, it felt like a discovery, all thanks to an instrument called the acoustic Doppler current profiler (or ADCP for short).

As water moves through the ocean its properties will gradually change due to mixing processes or more suddenly when exposed to the atmosphere. In a major study called the Atlantic Climate Change Experiment (ACCE) in the late 1990s we deployed 50 isopycnal floats in the North Atlantic west of the mid-Atlantic Ridge to explore water movements on the 27.5 kgm-3 density surface. All floats were equipped with Clark polarographic oxygen sensors. To minimize drift the sensors were first aged for a period (weeks), then calibrated, after which they were kept wet in a dedicated container and mounted on the floats only just before deployment. While most of them exhibited some measure of gradual drift, some up, others down, the records were extremely informative of ventilation processes at the surface when the isopycnal the float was drifting on outcropped at the surface. During the warmer months they also detected sharp but brief drawdowns in oxygen presumably resulting from detritus ‘raining’ from the surface.

People often think of the ocean as ‘turbulent’, but we could just as well talk about a structured ocean for ocean currents in fact exhibit an amazing array of patterns, some of which can persist for years if not decades, see my blog ‘spinning disks’. Here I’d like to highlight another class of structured motion, namely how water wends its way through the ocean at depth.

In 1962 Stommel wrote a paper “On the smallness of sinking regions in the ocean”. Even though that was a very simple model that couldn’t approach the complexities of the real ocean he was well aware of how incredibly small these regions in fact are. The deep water of the world oceans is supplied by narrow overflows from the Nordic Seas and from several sites around Antarctica, especially the Weddell and Ross Seas. The outflow of warm, very salty water from the Mediterranean is another well-known example. But due to earth’s rotation these overflows don’t run down a slope like a cataract, instead they slide sideways along the bottom increasing in depth, at first rapidly but soon adjusting and thereafter only very gradually. We used RAFOS floats to study overflow from the Nordic Seas through the Faroe Bank Channel.

In a previous blog I wrote that it was good to know that it wasn’t crucial if isopycnal floats didn’t settle on exactly the same density surface since once they were more than O(10) km apart, lateral shear would dominate over vertical shear and control their subsequent rates of separation. This because the lateral shear is controlled by the mesoscale velocity field, which is strongly coherent in the vertical, meaning it doesn’t matter exactly which isopycnal a float is on. But suppose you really wanted to know more about lateral dispersion in an isopycnal layer, i.e. in a layer in which vertical shear is ‘guaranteed’ absent. In that earlier study we could minimize but not guarantee absence of vertical shear by restricting ourselves to floats that were within 0.1°C of each other. But is there such a place? I think so, it’s an absolutely fascinating area.

Given the assumed role of fresh water in a potential slow-down of the AMOC we will need to know where and how fresh water out of the Arctic via the east Greenland Current mixes with saline water from the south to reduce its density. We already know that contact between these water masses is limited since the North Atlantic Current water reaches the Nordic Seas relatively unaffected. While the great salinity anomalies do freshen the North Atlantic Current some, they appear to have had no measurable impact on the rate of overflow through the Denmark St and Faroe Bank Channel, the source waters for the deep AMOC. The reason for this must lie in the multidecadal buffering of the Norwegian and Greenland Seas, the principal overturning basins in the Nordic Seas. This is good news. On the other hand, the fresh water may have a more immediate impact on the subpolar gyre overturning, the shallow AMOC. Here’s a way we could monitor this flux, both efficiently and cost-effectively.

The SOFAR float program took its first steps in 1966 when Henry Stommel introduced me to Doug Webb and learned about his efforts to develop a very low frequency acoustic float that could be heard at great distances. Stommel had already made arrangements for us to use the Broad Ocean Area SOFAR hydrophone network to do the listening. Doug was to develop the float (see ‘A close call in MODE’) and I was was to implement the float tracking system. Compared to Doug’s challenge, my task was rather straightforward. It was to install receiver electronics and digital data recorders, using at first 7-track, later 9-track computer-grade magnetic tape at each of the SOFAR hydrophones. We initially planned to install receivers in Bermuda, Eleuthera, Puerto Rico, and Antigua. But there were issues with the Antigua site (and it was distant) and the Bermuda phone was shadowed toward the west. There was a hydrophone at Grand Turk Island (GTI) but for some reason we couldn’t use it. So, we proposed and received funding from NSF to install our own hydrophone at GTI.

Little did I know that GSO (Graduate School of Oceanography) would become my professional home when I first arrived in summer 1975. Shaped by my family background, my frame of mind was somewhat that of a nomad, born in the US, grown up and educated in Sweden and then back to the US, first graduate school at MIT, and then a junior teaching position at Yale. Without giving it much thought I assumed I would be moving on after perhaps some half-dozen years. And now it is 48 years since I arrived, what happened?

We have long known that the overturning of the oceans is highly asymmetric. In 1962 Stommel wrote a paper on the ‘Smallness of the sinking regions in the oceans’. A year later (summer 1963) he tasked me to show this experimentally. The reason for the asymmetry was obvious, sinking waters are buoyancy-driven, leading to an advection of cold water into the abyss. Getting those waters back toward the surface requires downward diffusion of heat to make them buoyant, a far less efficient process requiring vast areas where this can take place – hence the enormous asymmetry. But how do we measure or quantify this overturning? Especially now that we know that the sinking at high latitudes and the slow rising elsewhere are far more complex processes than originally envisioned. This is where the beauty of western boundary currents comes in.

The Gulf Stream is the name we give to the stream of warm water emanating from Gulf of Mexico through the Florida St and continuing north and east past Cape Hatteras. Starting with the Ben Franklin map and continuing into the present this perception of the Gulf Stream as a open ocean river is firmly etched in our minds. We also know this warm water continues east and north across the Atlantic contributing in a major role in the mild climate of Europe. Accordingly, you might imagine that anything floating in the center of the current will be swept east and north toward Europe. So why is it that surface drifters positioned in the center of the Gulf Stream at Cape Hatteras, say, never stay in the current but escape to the south into the Sargasso Sea? There are two parts to this story, the first one about surface drifters and the other one about drifters at depth, for they too, will leave the current to the south.

My first oceanographic cruise in September 1963 did not go well – at all. I had learned that Pierre Welander, a Swedish oceanographer at the University of Gothenburg was planning to test a new technique he and his engineer (I forget his name so I’ll call him Lars) had developed to measure currents. If my memory serves me right, it consisted of a kilometer long flexible tube they were going to lay on the bottom. As the current varied, so would the pressure difference between the open ends vary. This would induce a weak flow in the tube, a flow which they hoped to measure by looking at how heat dispersed from a point source in the tube – did it go in one direction or the other, something like this is what I recall. Sounded pretty crazy, perhaps I misunderstood something.

Historically speaking, oceanographers were hydrographers – they map out distributions of heat and salt from which they can estimate transports relative to an assumed quiet deep ocean. This worked reasonably well at low latitudes, but not in western boundary currents nor at high latitudes. Believe it or not we couldn’t do much better than this until the 1960s when much effort was spent developing current meters and neutrally buoyant floats which could be deployed for extended periods of time. These gave us the tools to see what the ocean was doing. what we could only dimly perceive in the past we now could see the amazing richness of eddy activity in the ocean. This was an exciting and transformational period in physical oceanography. What also became apparent was that point measurements of currents, no matter how long, are ineffective for estimating transport, another approach was needed. The first steps in this direction were taken with the advent of the acoustic Doppler current profiler (ADCP) coupled with GPS navigation. An ADCP equipped ship in regular traffic can measure currents accurately from the surface to 300-1000 m depths, depending upon the instrument – repeatedly, year after year.

This has nothing to do with Oceanography, but at least it is about one of the little wonders of Nature. I was finishing up a three month mini-sabbatical at the University of Stockholm. I’m pretty sure it was January 5, 1996, a very cold Friday. The sky was clear and the air still when I got to my office early that morning. When the sun came up I saw the 22° circle halo, but what really caught my eye were the bright sun dogs to either side. I had seen them many times before, but this time they were so bright! Needless to say I took lots of pictures, in fact I wasted much of what I had left in the camera. Not good for I was going visit a friend in the Stockholm archipelago the next morning, and because the winter had been so cold the city harbor was full of pack ice. The boat ride was bound to be exciting. I had to go into the city to buy some more film.

Much of what we know about the ocean in those early years of modern science is thanks to the presence of the merchant marine on the high seas. This came about through the organizing efforts of Matthew Fontaine Maury, who arranged for ships to file reports on weather, winds, currents and sea surface temperatures. Charts of prevailing winds based on these reports led to much improved shipping times. Similarly, what we know about sea surface temperature 150 years ago is thanks to the reports from those vessels. Why do I make this point?

One thing that really baffles me is the lack of interest in the oceanographic community for teaming up with the merchant marine (MM) for ocean observation. You would think it would be exactly the other way around, for who has a presence on the high seas if not the shipping industry. Most everywhere and all the time, now and forever. There are, of course, some important exceptions, the volunteer observing programs (VOS) provide weather and sea surface reports, and the Ship of Opportunity programs (SOOP) include XBTs, surface water properties, and surface plankton distributions. In addition, a very few ships are equipped with acoustic Doppler current profilers (ADCP) to scan upper ocean currents. But we could be doing so much more!

One day at lunch Prof. Martin Mork and I were discussing the remarkable steps in temperature and salinity revealed by the Bissett-Berman STD (salinity, temperature, depth) profiler. Each layer appeared to be quite uniform with a thin transition to the next layer. We couldn’t but help wonder what the velocity field might look like. It occurred to me it might be possible to resolve velocity at high vertical resolution using a fast-response hot-film sensor on a free-falling profiler. He liked the idea and invited me to try it out on an upcoming cruise he had scheduled south of Spain that summer (1971). I assembled a simple profiler in a few months and took it as excess baggage with me to his 8-day cruise out of Malaga. The instrument worked and we wrote a technical report on the effort, but we really had no idea how to interpret the measurements. This would be the end of story - except that later that summer while visiting family in Sweden I got an urgent message from Martin to please come and visit him in Bergen. That become the start to what became Yvette.

It was in a cover photo from Scientific American that I first became aware of the Green Flash. If I remember right the picture was of a sunset somewhere in the Mediterranean. At the time I was living in Stockholm so the opportunities for seeing a green flash in person were poor at best. But the hope that I might someday see it never left me. But it would be another 20 years or so before I saw it for the first time, sitting at a hotel bar in Bermuda. I was so excited I put out a loud ‘wow, did you see that’! The others at the bar looked at me with a jaded ‘so what look’, I couldn’t interest them in what had just happened. Already then I was well aware that the green flash was considered ‘an old wife’s tale’. That was certainly the case some years later when a colleague joined me on an oceanographic cruise out of San Juan on the research vessel Gillis.

This is a troubled story, but I tell it anyway for the promise it holds. For some years my colleague, Prof. Godi Fischer, had together with his graduate students been developing a fishtag that included a complete RAFOS signal receiving system on a silicon chip. This development led to several PhD theses in recent years. Along the way various tests showed this development to hold promise. In 2017 another colleague, Prof. Jeremy Collie asked if it might be possible to use the fishtag to track lobster – and we saw no reason why not. What was even better, the tag was just about ready for its first serious application in the field, so why not on lobster? We had three sound sources available to provide the acoustic navigation.

Soon after we started deploying isopycnal floats in the Gulf Stream one of my students, Arthur Mariano, insisted that we should deploy floats in pairs or clusters to study how they initially disperse – or equivalently for how long will they stay close. I was reluctant to do this as this meant using twice as many instruments for the basic objective of tracing out fluid pathways in the Gulf Stream. I’d rather use them to get independent realizations. But Arthur was insistent, so as our confidence in the new float technology grew and could send them on longer missions, I relented and began deploying floats in pairs knowing that they would eventually part ways and become independent tracers of fluid motion. The results were striking to say the least.

For a float to drift at a certain depth two requirements must be met. First, its weight must be adjusted to match the weight of the water it displaces – what is known as Archimedes’ principle. Second, it must be less compressible than water. This, so that if it is displaced from its equilibrium depth, downward say, it will not compress as much as the surrounding water resulting in a restoring buoyancy force. If the float is to follow an isopycnal surface, we must add a further requirement, that its compressibility match that of the surrounding water. We do that with a compressee (see blog on the compressee). For a float to remain on the same isopycnal we depend upon the fact that the ocean is thermally stratified such that water gets denser with depth due to decreasing temperature. But the resulting restoring forces are smaller, so it becomes far more challenging to target it to the desired density surface.

In an earlier blog “A brief history of the RAFOS float” I mentioned the development of the SOFAR float and its use in the 1973 Mid-Ocean Dynamics Experiment (MODE), and later in polyMODE. It was a collaboration between me and Doug Webb, an amazing engineer I was most fortunate to work with. Doug’s challenge was enormous: to develop the SOFAR float which as it drifts freely at depth would tell us about the ocean currents we cannot see. In principle the concept was simple: as it drifts it would transmit precisely timed acoustic signals which could be used to determine its position. My task was to listen for those signals and determine their arrival times at hydrophones (underwater microphones) at Bermuda, Bahamas, Grand Turk, and Puerto Rico. Doug joked that his task was to put the sound in the water, mine was to get it out. But make no mistake, his challenge was far greater than mine. Behind those brief words hides an amazing period of engineering to develop the SOFAR float: its conceptualization, construction, and testing, addressing shortcomings, and testing again. a continuing process both in the shop and later at sea. After just one opportunity to test a completely re-designed float in 1972, Doug built 20 floats to be deployed in MODE. It was an immensely successful study, but what the scientific publications don’t tell you is what a close call it was.

In the previous blog I discussed the frustrations we had trying to realize the ‘one-shot’ float concept, a subsurface instrument that would give us only its end point after a long period of drifting at depth. A single vector may not be very illuminating, but a large cluster of vectors would tell us about mean drift and dispersion, both simple statistics of great interest. Clusters deployed from merchant marine vessels in regular traffic would keep operating costs to a minimum. Repeat deployment of clusters would give us a window into very low-frequency (sub-annual) oceanic variability, something we know very little about. We often imagine the deep ocean circulation as slow and not very energetic, but we also know that the kinetic energy spectrum is red meaning that there is variability at ever lower frequencies.

In the ten years since those earlier efforts a host of new technologies gave new life to the deep-drifter concept. In this note I describe our approach: a glass pipe float drifts wherever the currents take it, and that after a prescribed time surfaces and reports its position. It measures pressure and temperature along the way (easy to do), but it is not tracked along the way. We used it to tag a Mediterranean salt lens (meddy) so we know where to find it when the drifter surfaces at a later time.

Early in the SOFAR float program a parallel float concept started to take shape, namely the ‘one-shot’ float. While acoustically tracked floats give us tremendous detail about particle motion in the ocean, where they end up over long time will result from a combination of mean flow and eddy activity. Resolved trajectories provide a wealth of information on mesoscale activity, but their scatter makes it difficult to estimate mean flow with any accuracy. Thus, we began to explore the concept of one-shot drifters, i.e. floats that drift silently until end of mission when they make a single signal that can be used to determine their endpoint. Key to this thinking was that the float be simple and inexpensive so it could be deployed in large numbers. It is a frustrating story.

In an earlier blog I give a brief history of the glass pipe RAFOS float. One of the really cool attributes of this glass-pipe instrument is that it can be made to follow isopycnal surfaces in the ocean. Up until then all floats, whether the original Swallow float, the SOFAR float, or even the basic RAFOS float, will drift with the water at essentially at the same depth or pressure. But for those of us who worked with floats, designing a float to follow an isopycnal surface as it moves up and down was always high on our list of objectives, we just couldn’t figure out to do this! Yes, we could instrument the float to measure temperature and salinity and adjust the float’s volume to stay at a desired density, but that would be a very costly approach at best. I don’t recall when we realized this, but the properties of the glass pipe allowed us to convert the float from a constant pressure to a constant density (or isopycnal) device.

Oceanographers use two very different approaches to describe fluid motion. One, the Eulerian method, considers the speed and direction of fluid motion at a point. We use current meters to capture this information. The longer the record, the more we can say about the temporal properties of fluid motion at that point. This point may be representative of flow in nearby regions, but it will tell us little about how fluid motion varies across space. The other method, the Lagrangian method, follows a fluid parcel that has been tagged with a marker of some kind. In our case we use a neutrally buoyant float to follow its movement or trajectory, giving us information about flow patterns in the ocean. This approach gives us a wealth of information about the spatial structure of currents. Some trajectories are easy to interpret, others are not. Such floats were first developed by John Swallow in the early 1950s. They were tracked from a ship.

Throughout the 20th century oceanographers have been mapping out the water masses of the oceans. Long ago we learned that the Atlantic is saltier than the Pacific, and the Mediterranean even more so. We know that the cold waters of the global abyss originate in the Antarctic and that during interglacial times also come from the Arctic by spilling into the deep North Atlantic. We’ve excelled at measuring temperature, salinity, oxygen, and other water properties, but we have not been able to measure currents! Only much later, barely 50 years ago, did we begin to map out currents at depth by tracking instruments that drift with the water. It is these floats that really opened our eyes to how the ocean below the surface behaves.

I’m not sure where this fascination with a possible AMOC collapse started. Perhaps with Broecker’s widely cited paper in Science 1997? I bring this up because a number of publications have been appearing lately about the slow-down and eventual collapse of the AMOC. There seems to be almost an axiomatic belief that because there is increasing ice melt, this fresh water must inhibit the production of deep water and thus weaken the AMOC, perhaps to the point of collapse. I’ve always found the notion that as the globe warms the Nordic countries will head toward cooler times rather implausible.

Starting in Fall 1992 we have been operating an acoustic Doppler current profiler (ADCP) on the freighter Oleander, a container vessel that operates on a weekly schedule between Hamilton, Bermuda and Port Elisabeth, NJ. The key objective of this project has been to monitor the strength of the Gulf Stream, how it varies over time including whether there is any evidence of a long-term trend. This matters because the Gulf Stream is the only place where water is moving poleward, everywhere else across the Atlantic and at depth water is moving equatorward. What makes the Oleander route so effective it that besides spanning the transport of warm water by the Gulf Stream, it also traverses the westward flow of cold water the Slope Sea and shelf waters to its north, and the westward flow of Sargasso Sea water between the Gulf Stream and Bermuda.

In the instruments section there is a brief description of the Pegasus profiler that we developed in the late 1970s. It profiles currents in the ocean much like an ascending balloon profiles winds aloft. Whereas a balloon can be tracked visually or by radar, the Pegasus profiler is tracked acoustically relative to acoustic beacons or transponders that have been placed on the ocean bottom. From the lateral displacements of the instrument as it sinks and rises, we can profile absolute velocity to cms-1 accuracy. At Ants Leetmaa’s request, a physical oceanographer at AOML in Miami, we designed and built the profiler (on a very fast time scale) to study the Somali Current (see ‘Near miss in the Seychelles’). That project was so successful we decided to initiate a similar program to study the Gulf Stream. We got funding to build two more for a campaign of bi-monthly sections across the Gulf Stream over a 3-year period.

AMOC stands for Atlantic Meridional Overturning Circulation. That sounds simple enough, in 25 words or less you might say it is ‘the flow of warm water north where it gets cooled, sinks and flows back south at depth.’ But as soon as you start to look under the hood, as it were, things get a lot more involved – there are so many aspects to it.

To be a sea-going oceanographer demands a lot of preparation, that goes without saying. But when you are staging a new expedition from a distant port with brand-new instrumentation that hasn’t been tested the challenges and hurdles quickly multiply. What follows is an example of the high levels of anxiety one might have to cope with when preparing for a cruise.

All oceans have ‘Gulf Streams’ or western boundary currents as they are known because they all flow along the western margins of oceans. These include the Gulf Stream off the US east coast, the Brazil Current in the South Atlantic, the Agulhas Current in the Indian Ocean, and the Kuroshio and East Australia Currents in the North and South Pacific oceans, respectively. Regardless of where they are, they all flow poleward, i.e. toward higher latitudes. In so doing they close the loop on the ocean-wide gyres driven by the trades that blow from east to west at tropical latitudes and the westerlies that blow from west to east at mid-latitudes. Given that the wind systems never stop blowing, not even in glacial times, so will these currents always flow. So why is it that people worry about the ‘shut-down’ of the Gulf Stream and the climate changes that might follow? It’s an interesting story.

Far in the NE Atlantic halfway between Iceland and Scotland sits a small group of islands known as the Faroes or Faroe Islands. The islands have been inhabited for over 1000 years since Vikings from the west coast of Norway first settled down there. For centuries sheep farming was a major source of livelihood, supplemented by grain and other supplies through trade with Norway. In the last hundred years or so, fishing has steadily grown in importance and comprises today by far the greatest source of activity and income to the islands. Indeed, the waters around the Faroes are immensely productive because they are both warm and nutrient-rich thanks to the warm waters flowing past the islands from the North Atlantic into the Norwegian and Greenland Seas, more generally known as the Nordic Seas.