Tag Archives: iceberg

The Turbulence of Van Gogh and the Labrador Shelf Current

Vincent Van Gogh painted his most turbulent images when insane. The Labrador Current resembles Van Gogh’s paintings when it becomes unstable. There is no reason that mental and geophysical instability relate to each other. And yet they do. Russian physicist Andrey Kolmogorov developed theories of turbulence 70 years ago that Mexican physicist applied to some of Van Gogh’s paintings such as “Starry Sky:”

Vincent Van Gogh's "Starry Sky" painted in June 1889.

Vincent Van Gogh’s “Starry Sky” painted in June 1889.

The whirls and curls evoke motion. The colors vibrate and oscillate like waves that come and go. There are rounded curves and borders in the tiny trees, the big mountains, and the blinking stars. Oceanographers call these rounded curves eddies when they close on themselves much as is done by a smooth wave that is breaking when it hits the beach in violent turmoil.

Waves come in many sizes at many periods. The wave on the beach has a period of 5 seconds maybe and measures 50 meters from crest to crest. Tides are waves, too, but their period is half a day with a distance of more than 1000 km from crest to crest. These are scales of time and space. There exist powerful mathematical statements to tell us that we can describe all motions as the sum of many waves at different scales. Our cell phone and computer communications depend on it, as do whales, dolphins, and submarines navigating under water, but I digress.

The Labrador Shelf Current off Canada moves ice, icebergs, and ice islands from the Arctic down the coast into the Atlantic Ocean. To the naked eye the ice is white while the ocean is blue. Our eyes in the sky on NASA satellites sense the amount of light and color that ice and ocean when hit by sun or moon light reflects back to space. An image from last friday gives a sense of the violence and motion when this icy south-eastward flowing current off Labrador is opposed by a short wind-burst in the opposite direction:

Ice in the Labrador Current as seen by MODIS-Terra on May 3, 2013.

Ice in the Labrador Current as seen by MODIS-Terra on May 3, 2013. Blue colors represent open water while white and yellow colors represent ice of varying concentrations.

Flying from London to Chicago on April 6, 2008, Daniel Schwen photographed the icy surface of the Labrador Current a little farther south:

Ice fields seen in Labrador Current April 6, 2008 from a plane. [Photo Credit: Daniel Schwen]

Ice fields seen in Labrador Current April 6, 2008 from a plane. [Photo Credit: Daniel Schwen]

Ice in the Labrador Current as seen by MODIS-Terra on April 6, 2008. Blue colors represent open water while white and yellow colors represent ice of varying concentrations.

Ice in the Labrador Current as seen by MODIS-Terra on April 6, 2008. Blue colors represent open water while white and yellow colors represent ice of varying concentrations.

The swirls and eddies trap small pieces of ice and arrange them into wavy bands, filaments, and trap them. The ice visualizes turbulent motions at the ocean surface. Also notice the wide range in scales as some circular vortices are quiet small and some rather large. If the fluid is turbulent in the mathematical sense, then the color contrast or the intensity of the colors and their change in space varies according to an equation valid for almost all motions at almost all scales. It is this scaling law of turbulent motions that three Mexican physicists tested with regard to Van Gogh’s paintings. They “pretended” that the painting represents the image of a flow that follows the physics of turbulent motions. And their work finds that Van Gogh indeed painted intuitively in ways that mimics nature’s turbulent motions when the physical laws were not yet known.

There are two take-home messages for me: First, fine art and physics often converge in unexpected ways. Second, I now want to know, if nature’s painting of the Labrador Shelf Current follows the same rules. There is a crucial wrinkle in motions impacted by the earth rotations: While the turbulence of Van Gogh or Kolmogorov cascades energy from large to smaller scales, that is, the larger eddies break up into several smaller eddies, for planetary-scale motions influenced by the Coriolis force due to earth’s rotation, the energy moves in the opposite direction, that is, the large eddies get larger as the feed on the smaller eddies. There is always more to discover, alas, but that’s the fun of physics, art, and oceanography.

Aragón, J., Naumis, G., Bai, M., Torres, M., & Maini, P. (2008). Turbulent Luminance in Impassioned van Gogh Paintings Journal of Mathematical Imaging and Vision, 30 (3), 275-283 DOI: 10.1007/s10851-007-0055-0

Ball, P. (2006). Van Gogh painted perfect turbulence news@nature DOI: 10.1038/news060703-17

Wu, Y., Tang, C., & Hannah, C. (2012). The circulation of eastern Canadian seas Progress in Oceanography, 106, 28-48 DOI: 10.1016/j.pocean.2012.06.005

Oceanography and Icebergs in Baffin Bay: LCDR Edward “Iceberg” Smith

In 1928 Edward H. “Iceberg” Smith took the 125 feet long Coast Guard Cutter “Marion” on an 8,100 mile journey from Boston, MA to New York City, NY via Disko Bay, Greenland. Along the way he defined operational Arctic Oceanography to explain and predict iceberg entering the busy sea lanes off North-America. The Titanic was sunk in 1912, the International Ice Patrol was formed in 1914, and LCDR Smith sailed to Greenland in 1928. The data are priceless 85 years later still. I used them to place modern observations from 2003 into a context of climate variations. First, however, let me give credit to one of the pioneers on whose scientific shoulders I stand:

Edward H. "Iceberg" Smith of the US Coast Guard with reversing thermometer.

Edward H. “Iceberg” Smith of the US Coast Guard with reversing thermometer.

“Iceberg” Smith entered the Coast Guard Academy at age 21 in 1910 and served during World War I as a navigator on Atlantic convoy escort duty. After this war his ship was detailed to the International Ice Patrol and he became one of its first scientific observers at age 32 in 1921. As such he was sent for a year to Bergen, Norway in 1925 to learn the latest theories in physical oceanography. Scandinavian explorers like Nansen, Ekman, Sverdrup, Bjerknes, and Helland-Hansen defined physical oceanography at this time by applying physics on a rotating earth to phenomena that they observed from ships sailing at sea or ships frozen in Arctic ice. Much of this revolutionary work is now elementary oceanography taught in introductory courses, but then, nobody knew much about why ice and ocean move they way they do. It was time to put ideas to a thorough test which is what “Iceberg” Smith did, when he got his ship and orders to explore in 1928.

USCGC Marion built in 1927 [from http://laesser.org/125-wsc/]

USCGC Marion built in 1927. Note the scale indicated by a person standing on the lower deck. [From http://laesser.org/125-wsc]

Armed with new ideas, knowledge, and the tiny USCGC Marion “Iceberg” Smith set to out to map seas between Labrador, Baffin Island, and Greenland to explain and predict the number of icebergs to enter the North-Atlantic Ocean. During his 10 weeks at sea he mapped ocean currents from over 2000 discrete measurements of temperature and salinity at many depths. This was before computers, GPS, and electronics. In 1928 this was slow to work with cold water collected in bottles with “reversing thermometers” that cut off the mercury to preserve temperatures measured in the ocean at depth to be read later aboard. Salinity was measured at sea by tedious chemical titrations. Imagine doing all of this from a rocking and rolling shallow draft cutter that bounces in icy seas for 10 weeks within fog much of the time. No radar to warn of icebergs either, and you want to study icebergs, so you move exactly where they are or where you think they are coming from. And they though that the Titanic was unsinkable.

Iceberg in the fog off Upernarvik, Greenland in July of 2003. [Photo Credit: Andreas Muenchow]

Iceberg in the fog off Upernavik, Greenland in July of 2003. [Photo Credit: Andreas Muenchow]

USCGC Healy in northern Baffin Bay in July 2003 with iceberg. Ellesmere Island is in the background.

USCGC Healy in northern Baffin Bay in July 2003 with iceberg. Ellesmere Island is in the background.

The 1928 Marion Expedition was the first US Coast Guard survey in Baffin Bay while the last such expedition took place 2003. Unlike “Iceberg” Smith we then had military grade GPS, radar, and sonar systems. These sensor systems allowed me to directly measure ocean currents from the moving ship every minute continuously from the surface to about 600 meters down. Oh, we also took water samples in bottles, but temperature, depth, and salinity are all measured electronically about 24 times every second. As a result we can actually test, if the physics that had to be assumed to be true in 1928 actually are true. As it turns out, the old theory to estimate currents from temperature and salinity sections works well off Canada, but not so well off Greenland. Furthermore, we found several eddies or vortices in the ocean from the current profiling sonars.

And finally, it took Edward H. “Iceberg” Smith only 3 years to publish most of his data and insightful interpretations while I am still working on both his and my own data 85 years and 10 years later, respectively. Sure, I got more data from a wider range of moored, ship-borne, and air-borne sensors, but I do wonder, if I really consider my data and interpretations as careful and think as thorough as LCDR Smith did. Furthermore, he had no computers and performed all calculations, crafted all graphs, and typed all reports tediously by hand. I would not want to trade, but all this makes me admire his skills, dedication, and accomplishments even more.

Dr Helen Johnson on acoustic Doppler current profiler (sonar to measure ocean velocity) watch aboard the USCGC Healy in Baffin Bay in 2003. [Photo credit: Andreas Muenchow]

Dr Helen Johnson on acoustic Doppler current profiler (sonar to measure ocean velocity) watch aboard the USCGC Healy in Baffin Bay in 2003. [Photo credit: Andreas Muenchow]

P.S.: The New Yorker has three stories on the subject published in 1938, 1949, and 1959. I eagerly await to read those.

ResearchBlogging.orgSmith, E. (1928). EXPEDITION OF U. S. COAST GUARD CUTTER MARION TO THE REGION OF DAVIS STRAIT IN 1928 Science, 68 (1768), 469-470 DOI: 10.1126/science.68.1768.469

Greenland’s Glaciers, Science, Sea-Level, and Teachers

Science Magazine hit climate change hard today. They cover how Greenland’s glaciers and ice sheets change as they interact with the ocean and contribute to sea-level rise feature in 3 related stories. The reality check of these three stories puts a damper on the usual doomsday scenarios of those whose skill is limited to grabbing public attention to move a political agenda. Real science works differently:

May-4, 2012 Science Magazine Cover: A jumble of icebergs forms in front of the heavily crevassed calving front of Jakobshavn Isbræ, one of the fastest outlet glaciers draining the Greenland Ice Sheet. The ~5-kilometer-wide ice front rises ~80 meters out of the water and extends more than 600 meters underwater. Recent research shows that the speeds of Greenland glaciers are increasing. See page 576. [Photo Credit: Ian Joughin, APL/UW]

The solid new research is that of Twila Moon, a graduate student at the University of Washington whose dissertation work relates to the evolution of Greenland’s outlet glaciers over the last 10 years. She uses data from Canadian, German, and Japanese radars flown on satellites. She applies fancy mathematics to the data and feds data and mathematics into modern computer codes. And with all that, she cracks the puzzle on how fast more than 200 of Greenland’s largest glaciers go to town, eh, I mean, to sea. Furthermore, she shows how this flow has changed over the last 10 years.

Twila Moon, graduate student and scientist at the University of Washington and first author of “21st-Century Evolution of Greenland Outlet Glacier velocities” that appeared in Science Magazine on May-4, 2012. [Photo Credit: APL/UW website]

Back in the days of 2008, crude, but simple back-on-the-envelope calculation suggested that Greenland contributes 0.8-2.0 meters to global sea-level rise by 2100. In stark contrast, the 2000-2010 data now reveals, that even the low-end estimate is too high by a factor of 10. A glacier here or there may accelerate at a large rate to give the 0.8-2.0 m, but these rates do not occur at the same time at all glaciers. Ms. Moon’s more comprehensive and careful analyses of accelerating glaciers bring down Greenland’s contributions to sea-level rise to below 0.1 m by 2100, that comes to about 1 mm/year or an inch in 30 years.

A commentary written by Professor Richard Alley relates to the ice-sheets that feed these glaciers. Dr. Alley is famous for his work on Greenland’s ice sheet as he participated in 2-Mile Time Machine, a project that revolutionized the way that we view climate and its variability the last 100,000 years. The title refers to the 2-mile long ice-core from Greenland’s ice-sheet that trapped and stored air and stuff from the last 100,000 years. Dr. Alley is also featured in Andrew Revkin’s dot-earth blog of the New York Times as the Singing Climatologist. His comment on “Modeling Ice-Sheet Flow” references Ms. Moon’s observations as evidence that ice sheets change quickly. It also contains the sentence that “The lack of a firm understanding of ice-sheet-ocean interaction, constrained by reliable ocean data, remains a critical obstacle to understanding future changes.” I could not agree more with this sentiment, these data are darn hard to come by … not as hard as getting to the bottom of the 2-mile time machine, though.

While Ms. Moon addressed changes in Greenland’s glaciers, Dr. Alley addressed the ice-sheets feeding those glaciers, another comment by physical oceanographer Dr. Josh Willis of NASA’s Jet Propulsion Laboratory relates to the sea-level changes caused by accelerating glaciers to make “Regional Sea-Level Projections.” He works mostly on massive computer models which devour massive amounts of data to get climate right. Sometimes this works, sometimes is does not, but he does comment that these earth system models give sea-level projections that are a factor 2 smaller than those derived from statistical relations and semi-empirical models using surface temperature and radiative forcing to extrapolate past trends into the future. The difference probably relates to smaller and more regional processes that involve the physics of ocean circulation and its interaction with ice-shelves off Antarctic and Greenland.

Dr. Josh Willis conducting an oceanographic experiment studying sea temperatures between New Zealand and Hawaii. [Credit: JPL/NASA]

My great oceanography hero, Henry Stommel of Woods Hole oceanographic Institution once wrote in his “View of the Sea,” that “Science is both an individual and a social activity.” I am sure that graduate student Ms. Moon, NASA researcher Dr. Willis, and veteran professor and science communicator Prof. Alley all work hard and lonely at night some nights … and party hard while discussing science and adventures over a beer, dinner, coffee in some city, remote field, or on a ship. The one group of people missing in this picture are … the science teachers, that is, those dedicated, over-worked, and under-paid professionals who encourage, motivate, and helped us to become scientists before we went to college.

The editorial of this week’s Science Magazine is entitled “Empowering Science Teachers.” It compares the social and professional status of pre-college science teachers in Finland and the USA. I will only say in the words of Anne Baffert, chemistry teacher at Salpointe Catholic High School in Tucson, Arizona, that too many science “… teachers work in a command-and-control environment, managed by those who lack any real understanding of how to improve the system.” The editorial suggests on how scientists can improve science teaching, such as “… active involvement in science through structured collaborations with scientists …” Apparently, Finland succeeds while we in the USA are challenged to get our graduate students into a pre-college class room teaching. More stuff for me to munch on here …

Ice Island Flotilla From Petermann Glacier Continue Southward Flow

More icebergs and ice island from Greenland are heading south along northern North-America this year. Petermann Glacier’s first piece arrived last year off Newfoundland causing a local tourist sensation for a stunning display of ice along its shores. There are many more pieces from Petermann to come for a few more years.

Track of Petermann Ice Island from Aug.-2010 through Aug.-2011 traveling in shallow water from northern Greenland along Baffin Island and Labrador to Newfoundland.

April 29/30, 2012 locations of Petermann Ice Island 2010 on their way south along northern North America. [Credit: Luc Desjardins, Canadian Ice Service]

Yet, how come that these arrivals are both so predictable in their pattern, but are almost impossible to pin down for an exact location and time? The answer involves mystical and fake forces, stunningly beautiful experiments, elegant mathematical equations, and, most important of all: spin.

The earth spins rapidly around its axis and neither ocean nor glaciers leave the planet for outer space. The obvious answer that gravity holds all the pieces in place is neither the correct nor the full answer. A subtle balance of several other forces makes Planet Earth the perfect place to keep us supplied with water to drink and air to breath. Additional forces besides gravity relate to the difference in pressure between the top and the bottom of the ocean as well as the rotational force that forces our car off the road if we speed too fast around a curve. The net effect of these is that earth fatter at the equator than at the North Pole. There appears to be more of gravity pulling us in at the North Pole than there is at the equator. Put another way, a scale measuring our own weight dips almost a pound more in Arctic Greenland than it does in the tropical forests of Borneo even if we do it naked in both places. Lose a pound of your weight instantly, travel to the far north. (GRACE)

This makes no sense intuitively, but common sense and intuition help little when it comes to how the ocean’s water and the atmosphere’s air move on a rotating planet. For example, we all know intuitively that a down-pour of rain flows down a slope into the ditch. It requires work to bring water up to the top of a hill or into the water towers to make sure that water flows when we open the faucet. Not true for the ocean at scales that relate to climate, weather, and changes of both. Here all water flows along, not down the hill. Better yet, it requires no work at all to keep it moving that way for all times. This is why Greenland’s ice keeps coming our way as soon as pieces break off. The earth’s spin makes it go around the hill, to speak loosely of pressure differences. Winds and friction have little effect. The ocean’s natural and usually stable state is in geostrophic balance. Geostrophy is a fancy word for saying that the ocean’s water flows along, not down a hill, because it is balanced by a fake and mystical Coriolis force that I will not explain. I teach a graduate class on Geophysical Fluid Dynamics for that.

In technical language, most of the oceans tend to flow along not down a pressure gradient. A kettle of boiling water discharges water from high pressure inside the kettle to the lower pressure in the kitchen. Yet the steam dissolved in the atmosphere moves around high or low-pressure systems. That’s how we read weather maps: Clockwise winds around high-pressure over Europe, North-America, and Asia to the north of the equator, counter-clockwise winds around low-pressure systems. If I apply this spin-law to Baffin Bay containing all the icebergs and ice islands, the spin rule states that these large and deep pieces flow along lines where the earth’s local rate of rotation, lets call it planetary spin f, divided by the local water depth, lets call it H, is a constant. So, to a first approximation, the icebergs and ice islands flow along a path where f/H is constant. If the planetary spin is constant, then the ice island follow lines of constant water depth H. There is more to the story, much more, such as the effects of waters of different densities residing next to each other, but I better continue this later, as I got a dinner date with a sweetheart and “Thermal Wind” can wait 😉

Melting Greenland’s Icebergs and Ice Islands by the Ocean

The BBC keeps asking good and penetrating questions about the fate of Greenland’s many icebergs in general and Petermann Glacier’s Ice Island in particular. A poor telephone connection across the Atlantic this morning prevented an interview, but made me answer a number of questions in writing. When answering these questions, I was thinking of those icebergs and ice islands one finds in abundance in the frigid waters of Baffin Bay, the Labrador Sea, and next to Greenland. I am not talking about what happens once icebergs enter the subtropical Atlantic Ocean and meet the Gulf Stream to the south of the Grand Banks, that is a different story.

USCG Healy besides massive iceberg in northern Baffin Bay, July 2003

1. What can icebergs tell us about oceans?

Icebergs are particles that track averaged ocean currents over the top 200-m or so. These currents are often refered to as “geostrophic” currents which is really a different word for an ocean force balance between pressure gradients (estimated from measurements of temperature and salinity changes with depth at many locations) and the Coriolis force due to the earth’s rotation. This is why RADM Edward H. “Iceberg” Smith of the US Coast Guard spent so much time sailing and taking measurements off northern Canada in the 1920ies and 1930ies as part of the International Ice Patrol. His outstanding publications 80 years old are a first example on how to apply theory (the Scandinavian or Bergen School of the 1920ies regarding dynamical physical oceanography) to a very specific application.

2. How does the trajectory of the ice-islands/ and other icebergs show information about the ocean currents?

Ice islands and icebergs are thick (50-200 m) and mostly submerged below the surface. Hence they are largely moved by the ocean currents about 30-200 m below the surface. I think of them as ocean drifters with a very large and deep drogue element that changes with time as the iceberg melts or tips over. This is different from sea ice that may reach only 5-m into the water column and thus only “sees” the very surface layer of the ocean that is largely influenced by winds. This is not entirely true for icebergs, because they are driven by ocean currents below the thin (10-20 m) ocean “mixed layer” that sea ice is embedded in.

3. What are the current unknowns or poorly understood parameters with regards to iceberg science and iceberg interactions with oceans in the High Arctic?

I think the problem with any prediction scheme of individual particles is that the ocean always has a strong turbulent and unpredictable part to it. This problem is fundamentally no different from trying to predict where oil from a spill will come ashore. We can do it “on average” rather well, but we are very poor trying to do in one specific case for a specific iceberg (or spill). Oh, this certainly also applies to climate predictions, easy to do “on average,” very hard to do in a specific case for a specific time and place that may be affected.

4. What sort of temperatures would the water be?

Meltwater plumes at zero salinity with pieces of ice floating in it have a temperature of 0 C, meltwater plumes with pieces of ice in it at ocean salinities have temperatures of about -1.8C, depending on the amount of sunlight and how much mixing takes place, these fresh and very thin surface layers (1-10 meters) can heat up substantially fast, and cool just as fast. In June/July you have 24 hours of sunlight, so temperatures of 4-10 C are not out of the ordinary, but these waters will do very little melting, because most of the mass is well below a 10 m depth of such fresh surface plumes.

Temperature (left) and salinity (right) distribution off Labrador in the summer of 2009 with depth and distance from the coast (from Colbourne et al., 2010). Note the very cold waters near the freezing point (blue and purple) on the continental shelf below 50-m depth.

5. How would changes in ocean stratification and temperature alter the melting of icebergs?

First, ocean temperature (or heat) and stratification are two very different things. In the Arctic almost all stratification is done by salinity, temperature is a tracer that has very little effect on density stratification. This is also the reason, that most of the ocean’s heat in the Arctic is at depths 200-400 m below the surface. This warm water does NOT rise towards the surface, because it is also salty water, it is often refered to the Atlantic Layer. This deep reservoir of heat is what many oceanographers (myself included) have in mind when they talk about the melting of Greenland’s glaciers by the ocean from below.

Petermann Glacier, for example, has a grounding line at 600-m below the surface. This grounding line is in contact with the heat from the Atlantic Layer that is melting it, but the melted water is fresh and cold, immediately stratifying the water column under the ice, so a source of energy is needed also to move or mix this cold fresh melt water away. An inclined slope may do so, tides may do so, internal waves traveling and breaking on the interface between cold-fresh and warm-salty waters may do so. At Petermann Glacier this Atlantic layer does not reach most of the floating element of this glacier, because it is only 100-200 m thick and thus does not extend into the heat of the Atlantic Layer. So, vertical stratification and location of heat are two different things.

Breaking waves on an interface due to a shear instability, i.e., flow in the (fresh and cold) upper layer is less than the flow in the denser (warm and salty) lower layer.

Second, think of ocean stratification as a blanket that insulates one region from the other. Removing the blanket requires kinetic energy (something or someone has to do the work, doing work requires energy). As you melt freshwater ice via conduction of heat to the ice, the melt has zero salinity and a temperature at the freezing point. This zero salinity water acts as the insolation blanket that reduces the heat reaching the ice. So, again, you need (a) a source of kinetic energy to do work to break down the stratification (of salinity) to (b) enable the transfer of heat from the ocean to the ice.

6. Can you recommend any key journal papers you think we should read?

The classical paper on this subject is [Gade, H.G., 1979: Melting of ice in sea water: A primitive model with application to the Antarctic Ice Shelf and Icebergs. J. Phys. Oceanogr., 9, 189-198], but this paper is a thorough theoretical development. I have professors of glaciology asking me what it means, so the material is not easy to penetrate. In a 2011 publication on the oceanography of Petermann Fjord impacting this Glacier, we made extensive use of the arguments and concepts presented in Gade (1979). That publication is more readable and accessible.

Its first author is Dr. Helen Johnson at Oxford University with whom I have collaborated since 2003 aboard US and Canadian icebreakers. She also published an illustrated diary of our 2007 expedition to Nares Strait.

ADDED Jan-12: My mind yesterday was unclear on the in situ temperature at which glacier ice of zero salinity is melting in seawater such as found on the Labrador shelf. The comment below points this out concisely, while this link to TheNakedScientist perhaps provides the longer and more visual explanation. Another fun explanation of the melting and freezing of ice in a salt solution relates to the making of ice cream. A subsurface temperature at -1.5 C at a salinity of 33.5 psu such as found on the Labrador shelf does melt the zero salinity ice of the iceberg somewhat. A boundary layer consisting of fresh meltwater will lower the salinity adjacent to the iceberg which will increase the freezing point which will reduce the melting until a new stable equilibria is reached.