Arctic Ice Dynamics Joint Experiment (AIDJEX) Pilot Study, Mar-Apr 1972

Location and Duration

Manned Station supplied by aircraft in the southern arm of the Beaufort Gyre
Mar 8-Apr 27, 1972 (Representative photographs here)

Narrative

In the spring of 1971 I was finishing my first year of graduate studies in the Geophysics Program at the University of Washington, taking and liking lectures on fluid turbulence from Professor J. Dungan (Jim) Smith. As I was leaving Smith's office one afternoon, he offhandedly asked if I might have any interest in visiting the Arctic. I said I would jump at the chance! Thus began my career as a polar scientist. I soon joined his research group that was doing state-of-the-art turbulence measurements, using a novel current meter Smith had developed that could be configured in an arrangement that measured the three-dimensional current structure, necessary for actually observing how momentum, heat, and other flow properties are transferred in turbulent flow. Measurements like that had not been done in the ocean before.

The following summer I got my first experience with real-world observational science by joining a research vessel anchored in the Columbia River off of Hood River, Oregon. We were measuring turbulence above sand waves that had formed on the river bottom. The work was sponsored by the Atomic Energy Commission, and it was only many years later that I could piece together AEC's interest in the migration of Columbia River sand waves: they were part of a heavy annual sediment load that buried vast amounts of radioactivity that had accumulated during at least two decades while the Columbia River served as the primary coolant for the Hanford plutonium production reactors (this was not divulged until the late '70s, by which time the practice had ended).

That fall we began preparing for the AIDJEX Pilot Study scheduled for the next spring. I was assigned to developing software for a novel "minicomputer" small enough to be deployed in the field with the electronics necessary for sampling and recording data several times per second from dozens of Smith's "partially ducted" current meters. In October I attended a week-long crash course at the Data General computer factory near Boston on programming a Nova 1200 minicomputer in its machine language. This was challenging; the Nova predated by several years microcomputers like those built by Apple and IBM, and computers I had known until then occupied large spaces and were progammed in languages like Fortran, loaded onto punch cards. Working at the basic level of "byte-size" commands was a whole new approach for me, and when we started, the computer and I were essentially ignorant. We both learned in a step-by-step process, including for example, implementing interrupt service routines that permit a computer to "multitask," e.g., pause one routine to accept data from a peripheral device.

I spent months of intensive work and late nights working with our team's electronics engineer, George Bean (whom I regarded as sort of a wizard) on integrating the electronics with the software. In early March, 1972, we boarded a plane in Seattle for Fairbanks, then flew on to Barrow the next morning. The flight north was spectacular, with views of the Yukon River and Brooks Range.

That evening, after meeting Andy Heiberg (who would become a good friend and play a central role in many of my later projects), and checking out a suite of cold-weather gear, I boarded a C-130 Hercules making the first night time flight to the AIDJEX camp, which by then had been established with basic infrastructure in place. I was standing on the flight deck behind the crew as we landed on a smudge-pot lighted runway constructed on a smooth, refrozen lead. In all the flights I've made, it was the only one where I could not sense when the wheels touched down. The meter-thick, young ice must have been elastic enough to cushion the airplane like a pillow.

My memory of that first night on the ice is of short naps between the roar of yet another C-130 landing, with all hands turning out to offload for a fast turnaround. The next morning I emerged from the hut to a dazzlingly bright scene, with color gradations ranging from white to blue, air so cold it stopped your nostrils, and an icescape with miniature plains and ridges as far as the eye could see. Welcome to the Arctic! I was hooked.

The experiment Jim Smith designed included three stainless steel masts suspended from the ice, with current meter "triplets" (i.e., measuring current in three different directions) spaced at set intervals to provide a profile of mean current and turbulence characteristics. The primary mast was deployed through a hydrohole in the floor of the "turbulence hut." It spanned the upper 40 m of the upper ocean with current meter clusters at several levels through a layer of well mixed, relatively fresh water overlying denser, saltier water below. We could access and adjust the orientation of this mast easily.

In addition, a large hydrohole was excavated outside the hut. Gaining access to the ocean through two to four meters of sea ice entails a lot of work. Jim was hardly one to use his position as leader to shirk chipping ice; sometimes that otherwise tedious exercise would turn into a good natured, but intense contest. Two additional masts were assembled over this hole, then moved into position across a nearby pressure ridge keel by members of our team who were qualified divers. They often returned from their dives with vivid descriptions of the unique environment they encountered under the ice.

The entire six weeks I spent at the AIDJEX camp was an adventure that in many ways is still fresh in my memory a half century later. The time was divided roughly between a first half, filled with intense work, both physical and mental, and a sometimes boring second half after we had the basic equipment running well, requiring only routine maintenance (I read a lot of novels).

There were two intense storms during AIDJEX separated by weeks that moved the ice at about half a knot relative to the underlying ocean. When pack ice moves in response to wind, it generates turbulence in the upper ocean. That what we hoped to measure.

After we had deployed the masts with the turbulence measuring gear, it became clear during the first storm that there was a problem with the system. In engineering parlance, it was a hardware/software glitch that was devilishly hard to isolate. Fortunately George Bean and I had developed a good working rapport. We spent hour after hour (day and night) tweaking the circuits, then the computer commands (or vice versa) trying to pinpoint the source. Unfortunately, as the first storm was intensifying, tension was building -- we all knew the entire project depended on isolating and correcting the problem. At some point, Jim was coming into the instrument hut at short intervals demanding to know what progress was being made. This would break our concentration and became more and more annoying. In the wee hours one morning, Jim came in as usual, and as I rose from the teletype terminal, George clearly read my body language, came over, put his arm over my shoulder and hustled me out the door, saying "Coffee time." We later joked that George had saved my graduate career!

Although I had participated in a field program on the Columbia River the previous summer, this was my first time on an extended project. It was a great learning experience filled with both serious and comical moments, sometimes simultaneously. By the time of the second storm, we had isolated and corrected the data system glitch, and the system was recording successfully. One evening during the storm, Jim came into the science hut, looking somewhat sheepish, wondering if I would mind going over to Untersteiner's hut (Norbert Untersteiner was the overall AIDJEX chief scientist) and "maybe smooth things over?" I had no idea what was going on, but did as he asked. I learned when I got there that Jim had a short time before burst into their hut, threatening to throw Alan Thorndike into the nearby lead. Alan was one of Norbert's star graduate students, and would go on to a stellar career as a polar scientist and professor. We later became friends and colleagues during the AIDJEX modeling effort, but at the time I hardly knew him or what this was all about. They were still laughing about Jim's outburst; I explained that he and all of us had been under a lot of strain getting the new computerized system operating, but that things seemed to be working well at this point. That they were laughing was a good sign, but as it turned out Jim (and I, when I was later analyzing our results) certainly did have a legitimate concern. At the time we did not have good satellite navigation and it since it was important to know how fast we were moving in a reference frame attached to earth, acoustic transponders had been dropped to the seafloor, which then could be tracked from the surface. For some reason, Alan had chosen to turn off the bottom tracking system just at a time when the ice was moving most rapidly.

The AIDJEX Pilot Study was a vast undertaking. According to a summary by A. Heiberg in AIDJEX Bulletin #14, peak occupancy approached 80 science and support personnel, with thirty structures, and supplied by numerous aircraft flights, including 19 C-130 flights. A kitchen staff of four kept us exceptionally well fed. Without a great deal of exercise after the intense work of setup, I gained more weight than I wanted to admit. Only late in the project was I able to borrow some skis from more experienced Canadians, which then became a critical addition to my gear in later projects. With frequent flights, we had many short term visitors, including groups from the Soviet Union and Japan. The Japanese scientists, in particular, spent a lot of time in our hut, carefully observing use of a small minicomputer in a viable data collection system.

As implied by the designation Pilot, the main motivation for the experiment may have been to gain experience and select participants for the upcoming year long AIDJEX main experiment (1975-76), yet a great deal of good science was accomplished. Unfortunately, and for reasons I never understood, Smith's novel and imaginative approach for measuring upper ocean turbulence was passed over for inclusion in the main effort.

Results

We used current data gathered with the Smith Rotor apparatus during the second storm to document a spiral structure to the currents in the upper well mixed layer, closely resembling the theoretical Ekman spiral resulting from the impact of earth's rotation (the Coriolis force) on a sheared flow. Although Ekman spirals had been inferred from previous measurement under ice, this was the first example from continuous currents during one storm. More importantly, we were able to measure the statistical Reynolds stress in the upper ocean for the first time. The combination of innovative instrumentation combined with the steady platform provided by the drifting ice, made it possible to measure turbulence through the mixed layer of the upper ocean, something thought impossible at the time.

I worked more or less frantically in the next few months to make sense of the data we had collected in the spring, and submitted an abstract for a presentation at the annual fall American Geophysical Union scientific meeting in San Francisco in December. In those days, the AGU was hardly the huge convention it would later metastasize into; nevertheless, most of the main players in meteorology/oceanography were there. Saundie was more than happy to leave two-year old Erika with Grandma for a break in San Francisco, and was in the audience in the Jack Tar Hotel ballroom for my talk. What I remember most is being so nervous that my knees were literally knocking, and that I flubbed a softball question from a professor who would be on my academic committee. Not exactly an auspicious start!

Analysis and interpretation of the data from the Pilot Experiment, and fitting it into what was known at the time about near surface ocean turbulence became the meat of my dissertation, and was distilled into an an article published in the Journal of Physical Oceanography.

McPhee, M.G., and J.D. Smith. 1976. Measurements of the turbulent boundary layer under pack ice, J. Phys. Oceanog.