Marie Tharp’s Adventures in Mapping the Seafloor, In Her Own Words
The pioneering mapmaker explains how she and colleagues discovered underwater mountain ranges 40,000 miles long, and helped to prove that the continents move.
July 30 marks 100 years since the birth of Marie Tharp, a pioneering geologist and cartographer who created some of the world’s first maps of the ocean floor. This week we’re celebrating her achievements and legacy with blog posts, webinars, giveaways, and more.
Tharp was one of the first women to work at the Lamont Geological Observatory (now the Lamont-Doherty Earth Observatory at Columbia University). That was in 1949, a time when women weren’t always welcome in the world of science. Despite these extra obstacles, she went on to change the course of history for geology and ocean exploration.
In the account below, she explains in her own words how she came to be a scientist, and what it was like to chart the bottom of the sea when so little was known about it. This story, “Connect the Dots: Mapping the Seafloor and Discovering the Mid-ocean Ridge,” was originally published in the book, Lamont-Doherty Earth Observatory of Columbia: Twelve Perspectives on the First Fifty Years 1949-1999.
Not too many people can say this about their lives: The whole world was spread out before me (or at least, the seventy percent of it covered by oceans). I had a blank canvas to fill with extraordinary possibilities, a fascinating jigsaw puzzle to piece together: mapping the world’s vast hidden seafloor. It was a once-in-a-lifetime—a once-in-the-history-of-the-world—opportunity for anyone, but especially for a woman in the 1940s. The nature of the times, the state of the science, and events large and small, logical and illogical, combined to make it all happen.
Right up until World War II, all that water—a few miles deep and hundreds of thousands of miles across—proved an ample barrier, preventing humans from getting any picture of what lay at the bottom. On any map of the world, three-quarters of the Earth was a uniform, featureless blue border for the continents. Scientists thought the ocean floor was almost as featureless—a flat, unchanging plain, a dumping ground slowly filled by sediments eroding from land.
Early depth measurements, collected using ropes and lead weights such as cannonballs, suggested that the ocean floor was slightly more complex, however. With 200 soundings obtained in this way, the Navy’s Matthew Fontaine Maury marked a plateau in the middle of the North Atlantic on his 1854 map. In the 1870s, spot soundings taken during the legendary HMS Challenger expeditions hinted at a broad rise in the central Atlantic, and temperature measurements by the Challenger’s expedition leader, Charles Wyville Thomson, indicated that there was a barrier between the east and west basins of the Atlantic.
In early 1947, [Lamont’s Maurice “Doc”] Ewing undertook a Sigma Xi lecture tour with the official purpose of finding bright students to work in oceanography. Actually, he was scouting for a group of technicians from wealthy families to whom he could offer adventure instead of pay. After Ewing’s talk, Bruce Heezen, who was then a junior at the University of Iowa, introduced himself to Doc, who said, “Young man, would you like to go on an expedition to the Mid-Atlantic Ridge? There are some mountains there, and we don’t know which way they run.”
The following summer, Bruce went to the Woods Hole Oceanographic Institution to join Doc on an expedition on Atlantis, using a continuous echo sounder to take profiles of the seafloor in the North Atlantic. But Bruce didn’t get to go with him. Instead he got his own ship, Balanus, serving as chief scientist, even though he was not yet a senior in college. He got some great on-the-job training, went back to Iowa in the fall to finish his degree and then joined Doc at Columbia.
My course to Lamont was a little more indirect. My father, William Edgar Tharp, was a soil surveyor for the U.S. Department of Agriculture, Bureau of Chemistry and Soils. Papa was a field man and his assignments were to make a soil map of a county, produce a written report describing the soil types and recommended uses and to collect soil samples for analysis in the chemistry division. These were printed by the government and distributed to farmers, insurance companies and the university extension divisions. We were constantly on the move, with Papa working in the Southern states during the winter and the Northern states in the summer. By the time I finished high school I had attended nearly two dozen schools and I had seen a lot of different landscapes. I guess I had map-making in my blood, though I hadn’t planned to follow in my father’s footsteps.
Throughout our travels, Papa always told me, “When you find your life’s work, make sure it is something you can do, and most important, something you like to do.” In college at Ohio University, I changed my major every semester. I was looking for something I was good at, something I could get paid for, and something I really liked, but there weren’t many opportunities for women then, except as a teacher, secretary or nurse. I couldn’t type and couldn’t stand the sight of blood, so I decided to try teaching and began taking education courses, which convinced me that I wouldn’t like teaching all that much. I graduated with majors in English and music and four minors.
I never would have gotten the chance to study geology if it hadn’t been for Pearl Harbor. Girls were needed to fill the jobs left open because the guys were off fighting. A year after the war started, the geology department at the University of Michigan opened its doors to women. In 1943, about ten of us girls responded to one of their fliers, which promised a job in the petroleum industry if we got a degree in geology. It seemed like something I could do. I earned a master’s degree and got a job with Stanolind Oil and Gas Co. in Tulsa, Oklahoma. Some of the girls I went to school with went into micropaleontological work and spent their time looking through microscopes. That seemed tedious, so I went to the University of Tulsa and got a degree in math. Still searching for something more challenging, I went to New York in 1948.
I looked for work at the American Museum of Natural History, but I decided I didn’t want to work there after a paleontologist told me how it took two years to separate a fossil from the surrounding matrix. I couldn’t imagine devoting so much time to something like that, so I tried Columbia to see if I could get a more interesting research job.
Just because I had a math degree, they sent me down to see Doc Ewing, but he was at sea. I went home and waited three weeks for him to come back. When he heard about my background, he was surprised and didn’t know quite what to do with me. Finally he blurted out, “Can you draft?” I had had a part-time drafting job at Michigan, so he hired me.
“I never would have gotten the chance to study geology if it hadn’t been for Pearl Harbor. Girls were needed to fill the jobs left open because the guys were off fighting.”
About two weeks later, Bruce arrived at Columbia. At first I worked for anyone who needed me. But after a few years Bruce kept me so busy that I ended up working exclusively for him, drafting and plotting ocean floor profiles.
During World War II, Ewing and Joe Worzel, working at Woods Hole, had developed the continuous echo sounder for the Navy. With this new instrument, depth measurements could be made nonstop and round-the-clock. A sound signal, usually an electronic ping, would be sent out at a regular interval, and a microphone inside the hull of the ship would pick up the echo. As a ping was sent out, a stylus would be set in motion downward across a continuously spooled strip of four-inch-wide paper. When the echo returned, the stylus would mark the recording paper by burning it with an electric spark. The result was an uninterrupted profile of seafloor depths along the ship’s course. Relatively uninterrupted, that is: The echo sounder depended on the ship’s electric power, which went off whenever someone opened the ship’s refrigerator. When that happened, no echo returned and the sounder recorded depths as bottomless as the crew’s appetite.
With technological advances and Ewing’s drive and direction, tens of thousands of depth measurements in the North Atlantic had been obtained from 1946 to 1952 on cruises of Atlantis. We also had some data from USN Stewart, which in 1921 was the first Navy ship to make a continuous track across the Atlantic. We had interminable rows of sounding numbers that I was supposed to turn into highly detailed and complete seafloor profiles.
Bruce and Ivan Tolstoy at Lamont had devised sheets on Mercator projection to plot surrounding data at a scale of 1:1,000,000, creating the standardized system that is still used today by the Navy and Lamont. Plotting on these sheets, Hester Haring and I went to work in 1952 at drafting tables in a lab on the second floor of Lamont Hall, near Bruce’s office with its coveted private study (a former Lamont bathroom). First, Hester would plot the depths from the sounding data. Then we plotted profiles with significant selected depths along the ship’s course. The profiles had to be drawn in a consistent manner. Any mistakes and someone like Bruce or I would scrawl a message like, ‘Plotted Backwards!’ on the profile and have it redrawn. Bruce and I would then compare the depths on the profiles with the original soundings.
Eventually, after the plotting, drawing, checking, correcting, redrawing and rechecking were done, I had a hodgepodge of disjointed and disconnected profiles of sections of the North Atlantic floor. Plotted on a map, the ship’s tracks looked like a spider’s web, with the rays radiating out from Bermuda, where most of the research vessels took on supplies and water. Sometimes, the tracks zigzagged, as the ships fled from the paths of storms.
After another six weeks to arrange and piece together the profiles in proper order from west to east, I completed six more-or-less parallel, trans-oceanic profiles of the North Atlantic. I noticed immediately the general similarity in the shape of the ridge in each profile. But when I compared the profiles, I was struck by the fact that the only consistent match-up was a V-shaped indentation in the center of the profiles. The individual mountains didn’t match up, but the cleft did, especially in the three northernmost profiles. I thought it might be a rift valley that cut into the ridge at its crest and continued all along its axis.
When I showed what I found to Bruce, he groaned and said, “It cannot be. It looks too much like continental drift.” At the time, believing in the theory of continental drift was almost a form of scientific heresy. Almost everyone in the United States thought continental drift was impossible. Bruce initially dismissed my interpretation of the profiles as “girl talk.”
But I thought the rift valley was real and kept looking for it in all the data I could get. If there were such a thing as continental drift, it seemed logical that something like a mid-ocean rift valley might be involved. The valley would form where new material came up from deep inside the Earth, splitting the mid-ocean ridge in two and pushing the sides apart.
Soon afterward, almost on impulse, we decided to make a physiographic diagram of the ocean floor in the style of A.K. Lobeck, professor of geomorphology at Columbia in the 1920s. Unlike flat contour maps, physiographic maps show the terrain as it would look from a low-flying plane. By 1952, Bruce had been on enough cruises to know most of the features of the Western Atlantic. So, after about an hour of doodling, he produced our first diagram. He was somewhat unhappy with his effort and asked me to do it over. But both of us were pleased with the technique. It allowed us to capture the seafloor’s many textured variations, contrasting the smoothness of the abyssal plains, for example, with the ruggedness of the mountains along the ridges. But we also had an ulterior motive: Detailed contour maps of the ocean floor were classified by the U.S. Navy, so the physiographic diagrams gave us a way to publish our data. In retrospect, our choice of map style turned out to be significant because it allowed a much wider audience to visualize the seafloor.
I started using the physiographic technique to make a more detailed map of the North Atlantic. Our goal was to present it as it actually existed and as it could be seen if all the water were drained away. But, of course, there would never be enough ship tracks to do this. In the face of a minimum amount of data and the immensity of the world ocean, Bruce took a logical and multidisciplinary approach. We used data from wherever we could get it, from different disciplines and different sources, but took great care to ensure that these data from various sources were all plotted on the same scale. We used hypotheses of ocean floor structure to fill in areas where we had meager data. Our final guideline was that the sketching began from the shoreline seaward and from the mid-ocean ridge landward—that is, from the areas that we were most familiar with to those that we weren’t.
More and better data accumulated. By 1952, Lamont had acquired the Vema and had installed on it the precision depth recorder (PDR), invented by Bernard Luskin at Columbia in a hole in the floor of Schermerhorn Hall. The PDR provided much more accuracy than earlier echo sounders, allowing us to differentiate between smoother- and rougher-textured areas and to pick up more subtle seafloor features, such as seamountlets, scarps and sediment drifts. By Vema’s twentieth cruise, the precise sounding data were combined with highly accurate ship tracking, thanks to Joe Worzel, who installed a satellite navigation system on Vema, the first ever on an academic research vessel.
Every other day, the captain of the Vema would read off soundings from the PDR records as the first mate plotted them along the ship’s navigation track. Bruce had always insisted that soundings be read at every peak and valley and at every significant change of slope, rather than at equal time intervals of, say, 15 minutes. The latter would have been easier to do, but it would have tended to miss small seamounts, scarps or canyons. When each chief scientist completed his cruise and was replaced by a new one, he debarked with a roll of sounding data.
Hester Haring, with her meticulous handwriting, using a crow quill pen and India ink on blue linen, maintained the Vema sounding records on standard 1:1,000,000 sheets for many years. These sheets became the bible to which we compared all other institutions’ ship data. Vema data were classified as 9 on a 1-9 scale. Less precise data received lower grades, which were labeled with large red numbers on sheets that began bulging in our ever-accreting files. When laying several sheets from different places on a light table, we used these numbers to evaluate soundings quickly and to use them wisely.
While this work was going on, Bruce got involved in another project that provided another crucial source of data. He and Doc had proved the existence of turbidity currents—slurries of sediment and water that behave as discrete streams within the ocean. They documented that a 1929 earthquake off the Grand Banks had precipitated turbidity currents of such high speeds that they snapped trans-Atlantic cables. Bell Laboratories was interested in laying new cables and asked Bruce to help determine the best locations for them. Bruce hired Howard Foster, a deaf graduate of the Boston School of Fine Arts, to plot the location of recorded earthquake epicenters in the oceans. In this pre-computer era, Howard had to plot tens of thousands of earthquakes by hand. While I was at my map table, plotting the position of the Mid-Atlantic Ridge and the alleged valley, Howard sat at an adjoining table making the map of oceanic earthquake locations. Both maps were created on the same scale, as Bruce insisted.
The earthquake epicenters weren’t as precisely located as our sounding data. Their positions could sometimes only be located anywhere within an abominably wide range of several hundred miles. But when Bruce accounted for this, he noticed that a nearly continuous line of earthquake epicenters ran down the center of the Mid-Atlantic Ridge. Of course, Beno Gutenberg and Charles Richter earlier had noticed that a belt of shallow earthquakes followed the ridge, but Bruce saw that the earthquakes fell within the rift valley. Because all our data were on maps of the same scale, we could superimpose the maps on a light table, and when we did, the earthquake epicenters lined up within the valley. By then, I was certain that the rift valley existed. Bruce had remained skeptical. It was not until the middle of 1953, about eight months after I had worked up the first six profiles, that he accepted the idea.
Recognizing the validity of the correlation between earthquakes and the rift valley, we plotted the position of the valley by using earthquake epicenters for locations where there were no soundings. The extension of the valley into the narrow Gulf of Aden and landward into the Rift Valley of East Africa convinced Bruce in mid-1953 that the Mid-Atlantic Ridge was part of a gigantic 40,000-mile-long mid-oceanic ridge system that extended throughout all the world’s oceans. In fact, the mid-ocean rift valley takes its name from the terrestrial rift valleys of East Africa. We made profiles of some of the valleys in East Africa and noted the topographical similarities between the valleys in the ocean and on land. Bruce also noticed that the shallow earthquakes associated with the East African Rift fell within the valley walls. He began to endorse the existence of a continuous central valley within the mid-oceanic ridge.
Doc began to get interested at this point. He’d heard of this “gully,” as we called it, and he would pop into our lab from time to time and ask, “How’s the gully coming?”
Meanwhile, I had extended the Mid-Atlantic Ridge and rift valley into the South Atlantic, using data from another legendary oceanographic expedition, the 30 trans-South Atlantic cruises of Germany’s Meteor in 1925-27. Sounding data from those cruises would have confirmed right then that the Mid-Atlantic rise extended into the South Atlantic and that it was not broad and gentle, as Maury and Thomson had thought, but narrow and extremely rugged. But the discovery had remained hidden in the unanalyzed data as scientists at the time focused on physical oceanographic measurements of currents and seawater properties, rather than on the seafloor. Then World War II interrupted further analysis.
Around this time, new data from other expeditions also revealed similar ridge features in the Indian Ocean, Arabian Sea, Red Sea and Gulf of Aden. A U.S. Navy expedition had found a large north-south ridge system in the eastern Pacific. While I busied myself with sounding data, Howard was plotting tens of thousands of earthquakes around the world. The pattern we had noticed held. Wherever there was a mid-oceanic ridge, there were earthquakes. When the Indian Ocean earthquake belt was shown to be continuous with the East African Rift Valley, there was but one conclusion: The mountain range with its central valley was more or less a continuous feature across the face of the Earth. Doc and Bruce announced our findings in 1956 at a meeting of the American Geophysical Union in Toronto.
The reaction in the scientific community ranged from amazement to skepticism to scorn. In 1957 Bruce gave a talk on the mid-ocean rift system at Princeton, bringing along a globe we made that showed how the rift system extended all around the world. After the talk, the eminent Princeton geologist Harry Hess, who later developed the theory of seafloor spreading, stood up and said, “Young man, you have shaken the foundations of geology!” The discovery of the mid-ocean ridge system was a revelation, but nobody could explain how it got there.
Bruce believed the rift was a tensional crack caused by the splitting of the Earth’s crust. He still did not believe in continental drift. It was very hard to go in the direction of that theory when the boss, Doc, like nearly everyone else in the scientific world, was violently opposed to drift. I was so busy making maps I let them argue. I figured I’d show them a picture of where the rift valley was and where it pulled apart.
There’s truth to the old clichés that a picture is worth a thousand words and that seeing is believing. Like most scientists, Jacques Cousteau at first didn’t believe in the rift valley. He crossed the Atlantic Ocean in the Calypso, towing a movie camera on a sled near the seafloor. They came to where our rift valley was and found it. He took beautiful movies of big black cliffs in blue water, which he showed at the first International Ocean Congress in New York in 1959. It helped a lot of people believe in our rift valley.
“The reaction in the scientific community ranged from amazement to skepticism to scorn.”
In 1956, we first published the North Atlantic physiographic map as an accompaniment to the Bell Telephone System’s Technical Journal. It was done in pen and ink. The Geological Society of America reprinted the map in 1959. To make the map, we first plotted lines of soundings taken by ships tracking across the ocean. Then we converted the sounding lines into two-dimensional profiles of the seafloor. Then we made three-dimensional sketches based on the profiles and plotted them along the ship tracks. Finally we sketched in areas with no soundings by extrapolating trends observed in profiles made by actual soundings. In other words, we made educated guesses to fill in the dataless gaps. Like the cartographers of old, we put a large legend in the space where we had no data. I also wanted to include mermaids and shipwrecks, but Bruce would have none of it.
We continued on, from one sounding to the next, and one ocean to the next. We weren’t daunted by the tens of thousands of soundings we had to plot. We were daunted more by all the data we didn’t have. For the map of the South Atlantic, in some places we only had spot soundings from the General Bathymetric Chart of the World series. We used data from the Meteor expedition to sketch in the mid-ocean ridge crest and the rift valley. Data from the Vema 9 cruise helped us in equatorial areas. We’d use any data available and change our minds as we got more. For example, we at first thought the rift in the Atlantic was a long valley. Then, in the South Atlantic, it was a long valley with some wiggles. Finally, we recognized the fracture zones, which offset the ridge by hundreds of miles.
One of the more challenging areas of the South Atlantic was the remote Scotia Sea, for which there were little or no data available. Fortunately, the pattern of the Caribbean and Scotia seafloors is strikingly similar, allowing us to make a valid extrapolation. The South Atlantic diagram was published in 1961.
We had planned to study the Mediterranean Sea next, but we were diverted instead to the Indian Ocean, because a diagram of it was urgently needed to help plan the International Indian Ocean Expedition. Now our efforts were thwarted by a long-lasting falling-out between Bruce and Doc. There are two sides to that story, but the result was that Doc banned Bruce from Lamont ships and denied Bruce access to Lamont data. He tried unsuccessfully to fire Bruce, who had a tenured faculty position at Columbia, but he did fire me. From then on, I was paid through research grants that Bruce received from the Navy, and I continued the mapping working at home.
Doc could have scuttled our mapping efforts, but Bruce had forged relationships with researchers all over the world, going to sea on ships from other institutions. By the early 1960s, we had recognized fracture zones in the Atlantic, but we couldn’t confirm their general direction and trends until 1968, when Bruce and I were able to secure a cruise aboard the Navy vessel Kane. We zigzagged over what became known as the Kane Fracture.
Bruce found alternative sources of data. His book with Charley Hollister, The Face of the Deep, had been translated into Russian, and perhaps inspired cooperation from Russian scientists, even during the height of the Cold War. We received extensive soundings from the Soviet ships Ob and Vityaz, which surveyed the Indian Ocean. Japanese soundings between Capetown and Antarctica, and data from the United Kingdom, Australia and South Africa, and several American oceanographic institutions, were all incorporated into the Indian Ocean map, published in 1964—a truly international effort. And, I should note, it contained a big error. I got so overwhelmed with the fracture zones in the Indian Ocean, I didn’t initially recognize a triple junction, where three mid-ocean ridges intersected. We published the map with that error, but corrected it later when new data revealed it.
Inspired by the International Indian Ocean Expedition, the National Geographic Society wanted to commission a map of the Indian Ocean to illustrate an article on it. Some time earlier, National Geographic had received a letter from a little girl in Austria who wrote, “I’ve been looking at your maps and my father can paint better than you can.” Intrigued, National Geographic editors sent their chief topographer to Innsbruck, Austria, to meet the girl’s artist father, Heinrich Berann.
Berann did serious paintings in the style of Leonardo da Vinci often with religious themes which, in my opinion, ranks him as one of the foremost painters of our century. But he couldn’t earn a living doing this. So he began to paint realistic alpine panoramas for advertisements promoting skiing for tourists. National Geographic commissioned him to paint the Indian Ocean floor and hired Bruce and me as consultants. We loved working with Heinrich, and his familiarity with painting the Alps translated beautifully to the seafloor. The three of us published a panorama of the Indian Ocean in 1967 and then continued with the rest of the world’s ocean floors. The final map we produced for National Geographic was of the Antarctic ocean floor in 1975.
The next step was obvious: to paint a panorama of the entire world’s ocean floor. In 1973 the three of us submitted a proposal for the project to the Office of Naval Research. To accomplish it, we had to simplify some of our previous work to accommodate the smaller scale called for by a world map. At the same time we had to update our work to include the vast volume of data that had accumulated over the years.
We’d use all the data we had, but the data didn’t provide complete coverage, so there still were blank areas. That was the biggest challenge: providing data for the blank areas. Over the next three years, we traveled back and forth to Austria. I’d go home, work up a blank area with any data we could get, come back to Austria, and Heinrich would paint that area. Constantly adding new data, we changed our minds quite a bit as the panorama took shape.
Our efforts were aided by the advance of technology over the 25 years since we first started mapping. In 1962, the World-Wide Standardized Seismic Network (which Lamont helped to establish, with instruments Ewing, Frank Press and other Lamont scientists invented) allowed seismologists to map earthquakes much more precisely. The positions of seafloor spreading centers were more accurately located by magnetic data, the bulk of which was collected by Lamont ships. Ironically, by this time, Ewing had moved to the University of Texas, so we could now use Lamont data, long denied to us, to finalize our maps.
The first proofs for the world ocean floor map arrived from the printers in time for Bruce to take them with him aboard the Navy’s nuclear submarine NR-1 on an expedition to explore the mid-ocean ridge off Iceland. In 25 years, the study of earth science had advanced so much, the traditional mountain-climbing geologist with a rock hammer could now sample the seafloor in a submersible. But Bruce died of a heart attack on that cruise, just a few months before the World Ocean Floor panorama was published in 1977.
I think our maps contributed to a revolution in geological thinking, which in some ways compares to the Copernican revolution. Scientists and the general public got their first relatively realistic image of a vast part of the planet that they could never see. The maps received wide coverage and were widely circulated. They brought the theory of continental drift within the realm of rational speculation. You could see the worldwide mid-ocean ridge and you could see that it coincided with earthquakes. The borders of the plates took shape, leading rapidly to the more comprehensive theory of plate tectonics.
I worked in the background for most of my career as a scientist, but I have absolutely no resentments. I thought I was lucky to have a job that was so interesting. Establishing the rift valley and the mid-ocean ridge that went all the way around the world for 40,000 miles—that was something important. You could only do that once. You can’t find anything bigger than that, at least on this planet.