Climate science showcase at Dynamic Earth (COP26)

A few weeks late but I thought I should add some photos of the event we attended as part of the COP26 activities. We had a great day chatting with the public about our work and demonstrating some of the robots and instruments used for physical oceanography. Dynamic Earth reported 1400 visitors throughout the day, one of their busiest public engagement events ever! Great to see so many folk after so much isolation over the past year.

Great that the seaglider model I built years ago has been such a valuable outreach tool. It’s still going strong, albeit with a few knocks and scratches. It looks more like the real thing as a result! More information on its construction here:

DY120 cruise animation

Following on from our recent successful completion of the OSNAP cruise on board RRS Discovery, I made a little animation showing how the objectives of the cruise were completed, and how compliant the weather was throughout. The North Atlantic, not known for respecting cruise schedules during the autumn and winter, allowed us a window just long enough to complete all the mooring turnarounds, CTDs and Argo deployments. Of note is the 3-day run back from IB3 in the Iceland Basin to complete the final mooring and get behind the Hebrides before the next low-pressure system arrived. The following Wednesday, waves up to 29 m were recorded west of Ireland, more or less where we’d been the previous week! Along with getting out during COVID, this seems like another aspect of DY120 worth celebrating.

Also on the animation is SAMS glider “Bowmore”, which was deployed in September and will run transects of the shelf edge current throughout the winter. A classic ‘tortoise and hare’ scenario: she doesn’t get anywhere fast, but who’s still out there gathering data long after we’ve run for shelter!

The history of AMOC study, blog post written during the DY120 cruise

Sam Jones 16/10/20

The world’s oceans are in constant circulation. Like a vast, stately conveyer belt, water flows from the tropics to polar regions where it cools and becomes denser, pouring back along the basins and valleys of the seabed towards the lower latitudes. This flow encompasses every ocean basin and takes many hundreds of years to complete a full circuit. A striking example of this process in action is the ocean’s delayed response to climate change: since the industrial revolution, the Polar seas have been taking in additional heat and pollutants from human activities and diligently sequestering them in the deep oceans of the world, where they will reside for many years before eventually resurfacing. Not gone forever, just filed under ‘to deal with later’.
The Atlantic segment of this flow is known as the AMOC: the Atlantic Meridional Overturning Circulation. ‘Meridional’ refers to the south-north nature of the flow, and ‘overturning’ describes its upwelling and downwelling components in the tropics and the poles, respectively. Despite its stately progress, the energy moved by the AMOC is staggering: roughly 1.25 Peta (1015) Watts of heat is continuously transported from the tropics towards the Atlantic Arctic via this mechanism, more than 60 times the present world energy consumption.

Schematic map of the AMOC, with RAPID and OSNAP array locations overlaid.

The idea that human activity could disrupt this immense current system was first considered in response to an unexpected result: in 1957 an ambitious hydrographic survey aimed to measure the total amount of water and heat transported by the AMOC using a line of observations between the eastern and western sides of the Atlantic ocean. This effort successfully delivered a figure for the strength of the AMOC, and a follow-up experiment in 1981 was conducted to see whether anything had changed during the 24-year interval. Contrary to expectations, the transport was found to have decreased significantly. The discrepancy was tentative evidence that this seemingly inexorable current system was slowing down.
A possible culprit for the slowdown was forthcoming: The Polar ice caps store vast amounts of fresh water, and human-induced climate change has added to existing ice melt, causing more fresh water to flow into the Polar oceans. Fresh water reduces the density of seawater so has the potential to disrupt the engine that drives the AMOC; water in the Arctic may no longer have sufficient density to sink and flow towards the tropics. This phenomenon would not be unprecedented: there is much evidence that such slowdowns occurred due to melting at the end of the last ice age, as recently as 8,200 years ago. This is the scenario portrayed in the movie ‘Day after Tomorrow’: it’s true that the warm, salty North Atlantic Current keeps Western Europe relatively mild and that a reduction in this current could lead to cooling in the northern hemisphere. That the current could ‘switch off’ in a matter of days, leading to polar super-hurricanes and an influx of wolves into New York is less scientifically rigorous, but in the interests of dramatic cinema we’ll let that slide.

After the discovery of an apparent slowdown in the AMOC, it was imperative that we learned more about the Atlantic portion of the ocean conveyer belt. How quickly was it slowing down? What would be the implications for climate in the northern hemisphere? How sensitive would it be to future changes in climate? To answer these questions, the RAPID mooring array was instigated in 2004. It comprised a line of deep ocean moorings maintained between Morocco and Florida to continuously measure the AMOC in a more comprehensive manner than had previously been possible. The data provided by this array demonstrated that the slowdown was real, but more gradual than had initially been feared. However, the scale of year-to-year and even seasonal variability in current strength was far greater than expected.

Scientists collect water samples during DY120 (Sam Jones/SAMS).

A large unknown remained: how much of the water which passes through the RAPID array reaches the Arctic to cool and sink, and how much just recirculates in the North Atlantic? To answer this question and better understand the engine which drives the AMOC, it was decided that a second array of instruments – a ‘gateway’ across the entrance to the Polar seas – was needed. This mooring array stretched from Western Scotland, via Greenland, to the east coast of Canada and was named OSNAP (the Overturning in the Sub-Polar North Atlantic Program). This international effort has been maintained since 2014 and has offered further insight into the inner workings of the ocean conveyer belt. We now know, for example, that most of the cooling and sinking occurs east of Greenland, and not in the Labrador Sea as was previously thought. In addition, most of the strong year-to-year changes in AMOC strength originate in the stormy seas of the Eastern North Atlantic where we currently find ourselves.
The moorings we are recovering during DY120 have survived for two years in the Atlantic Ocean, weathering some of the roughest seas on the planet. The data contained in these moorings adds to four years of existing OSNAP data, so increases the total duration of our time series by 50 %, a significant boost to our understanding. Each mooring consists of tens of instruments distributed along a wire, which is anchored to the seabed and kept vertical by a series of buoyant floats. Due to the depth of the ocean basins, most of the moorings are several kilometres long, so the task of recovering each mooring intact onto a relatively small vessel, downloading data and redeploying in precisely the same location is non-trivial. We have now moved from the Rockall Trough into the Iceland Basin, and each successful mooring turnaround feels like a significant triumph given the challenges in getting to this point.

Mooring recovery during DY120 (Sam Jones/SAMS).

New ICES ocean climate and atmosphere report published

The ICES Working Group on Ocean ​​Hydrography (WGOH), of which I am a chair-invited member, has just published its annual report on North Atlantic ocean conditions in 2018.  The report comprises several dozen multi-year time series of ocean observations around the North Atlantic Basin, which together paint a coherent picture of the current status of the ocean climate:


North Atlantic ocean conditions 2018

In 2016, freshening of the upper ocean (0–1000m depth) was observed in the eastern subpolar North Atlantic. This decrease in salinity has since expanded northwards into the Nordic seas, influencing the Greenland Sea and northern Norwegian Sea to Fram Strait, as well as the southern reaches of the Barents Sea. Freshening is also observed spreading westward into the Irminger Sea and eastward into the North Sea.

Throughout the subpolar region, freshening is accompanied by moderate cooling at just a few sites, indicating that the large changes in salinity are dissociated from changes in temperature.

Freshening of central waters in the northeast Subtropical Gyre and intergyre region (Bay of Biscay, West Iberia, Gulf of Cadiz, and Canaries) was enhanced and extended deeper into the water column. In contrast to northern regions, temperatures here decreased in concert with freshening, thereby conserving water mass properties.

Coupled with atmospheric conditions, sea surface temperatures (SST) exhibited a tri-pole pattern, with warm conditions in both the subtropical and Nordic seas regions and cooler conditions in the subpolar region. A cold anomaly observed in the surface and upper ocean of the central subpolar North Atlantic intensified and expanded after weakening in 2017.

The Scotian and Northeast US shelves were warmer than normal, accompanied by notable freshening at several sites.

Extremely warm temperatures were observed near the surface in spring–summer across the Baltic Sea and the North Sea (> 1.5˚C than normal), with less pronounced warming observed from Biscay to Ireland (0.5–+1.0 ˚C).

New paper: “Wintertime Fjord-Shelf Interaction and Ice Sheet Melting in Southeast Greenland”

A computer simulation of the ocean around Greenland was used to study the movement of water in and out of a large fjord. This is important because warm water that gets into the fjord may come into contact with the Greenland Ice Sheet and cause it to melt. The simulation indicates that a significant amount of warm water comes into contact with the ice during the winter. This was previously difficult to measure because of the difficulties in taking direct measurements of the water during the Greenland winter.

Time-averaged wave energy flux through cross-sections of Kangerdlugssuaq Fjord.  In short, there’s more going in than coming out!

Oscillations in the thermocline in the fjord over time, showing the impact of large amplitude coastally-trapped waves on water temperatures.

This work was based on Dr Neil Fraser’s PhD, but the model was forced using realistic winter wind conditions rather than idealised wind.  This was the first time (for me) that using Paraview to analyse model data in 4D has led or contributed to a publication.

Kangerdlugssuaq Fjord, taken during Prof. Mark Inall’s 2010 field work. (c) SAMS

Fraser, N. J., Inall, M. E., Magaldi, M. G., Haine T. W. N. and Jones S. C. (2018). Wintertime fjord-shelf interaction and ice sheet melting in southeast Greenland. JGR: Oceans,

New paper: “Cross-slope flow in the Atlantic Inflow Current driven by the on-shelf deflection of a slope current”


  • Slope water has been tracked on the European Shelf using drifters and gliders.
  • The deflection onto the shelf is not captured in models.
  • The slope water has a higher nitrate concentration that the shelf water, and supplies nutrients to the shelf.

(a) pathway of the shallow drifters over the first 45 days after their release, in grey. The thick black trajectory shows the time mean line, from which the local across and along flow directions are derived. The local bathymetry is shown by thin black lines and Coriolis parameter/depth (f/h) contours by dashed lines. Location A is the point at which the shallow drifters stagnated and turned to cross f/h contours.  (b) The trajectories of all of the drifters throughout their active periods, shaded by the date. Location C shows the drifter release point, Location D shows where the deeper drifters crossed onto the shelf and Location B where the deeper drifters re-joined the slope.


Porter, M., Dale, A., Jones, S.C., Siemering, B., Inall, M.E. (2018). Cross-slope flow in the Atlantic Inflow Current driven by the on shelf deflection of a slope current. Deep-Sea Research Part I, in press.  doi:10.1016/j.dsr.2018.09.002

New paper published: “Decadal variability on the Northwest European continental shelf”

Jones, S., Cottier, F., Inall, M., & Griffiths, C. (2018). Decadal variability on the Northwest European continental shelf. Progress in Oceanography, doi: 10.1016/j.pocean.2018.01.012

This paper details one of the key outcomes from my PhD so it was good to get it finished!  It describes how wind acting over the shallow seas west of Scotland can change the origin of waters on the inner continental shelf (and the coast).  This region typically recieves a mix of salty, nutrient rich water from the Atlantic and fresher, relatively nutrient poor water from the Irish Sea.  1-2 months of sustained easterly winds can block the inflow of Atlantic water and drive a pulse of Irish Sea water into the region, potentially importing much greater abundances of Irish Sea organisms and pollutants than during a typical year.  This body of water is detectable on the continental shelf for several months before it it fully displaced northwards.  Conversely, sustained winter storms can drive Atlantic water far onto the shelf and block the outflow from the Irish Sea, bringing oceanic conditions to what would normally be considered coastal locations.  The strong variability which results is roughly an order of magnitude greater than the changes seen in the adjacent Northeast Atlantic so is thought to mask the well documented decadal changes in these waters.

Map of waters off western Scotland.  Black rectangles show the position of Ellett Line stations on the continental shelf.

plot5 effort 1
Time series of surface salinity from Ellett Line stations on the continental shelf.  x-axis depicts longitude, between the shelf edge (left) and the Scottish coastline (right). (a) shows surface salinity, where brown colours indicate high salinity oceanic water and blue shows lower salinity coastal water. (b) shows the mean salinity across the shelf for each year.  (c) shows the salinity anomaly calculated by subtracting the mean of each station from the data, and (d) depicts the mean surface salinity for each station.  The grey region gives the standard deviation of the data which is a measure of the variability present.