Steven Dutch, Professor Emeritus, Natural and Applied Sciences, University of Wisconsin - Green Bay
Wherever possible, diagrams on this page are modified from original diagrams in scientific papers. These are based on the actual diagrams used by the discoverers of plate tectonics.
The diagram above, based on one by G. Brent Dalrymple, superbly illustrates how science proceeds by successive approximation. In the first five papers, there is general agreement that the magnetic field had its present orientation for the past million years, but disagreement over whether the previous normal period ended 2.0 or 2.5 million years ago.
The second and third papers were published only 13 days apart, far too short a time to re-analyze data, and write and publish a rebuttal. All these workers knew one another and were constantly communicating informally. There was no doubt they were all reporting their results reliably. The question is: why were they getting different results?
The sixth paper shows what was happening. In addition to the long periods ("epochs") of normal or reversed polarity, there are also short periods, or "events" where the magnetic field flips from one orientation to the other and back in a few thousand years. The rest of the papers show the later progress of adding progressively finer details.
It is important to note that only the magnetic field of the Earth flips; 1.5 million years ago, a present-day compass needle would have pointed south rather than north. The rotation axis of the Earth does not flip!
The answer to that question was already on hand. In 1954, Hugo Benioff had published a study of earthquakes beneath ocean trenches. The following three figures are modified from Benioff's original paper.
One of the trenches Benioff studied was the Japan-Kurile Trench in the North Pacific.
A cross-section shows how earthquakes lie on a planar zone dipping from the trench beneath the continent.
Benioff interpreted the dipping zones of earthquakes as giant thrust faults. He did not envision continental drift at the time.
Once sea-floor spreading was discovered, it was simple to modify Benioff's concept. The dipping seismic zones are great thrust faults, but oceanic crust is moving continuously along them to be recycled in the Earth's interior.
The Worldwide Standardized Seismic Network created a global system of seismographs all using common timekeeping standards and sending data to common repositories. This system allowed radical improvements in accuracy of earthquake locations. Instead of showing a diffuse smear of seismicity along the mid-ocean ridges, the system showed that earthquakes were confined to extremely narrow zones along the crest of the ridges. Basically, the new maps showed that the earth consisted of large blocks or plates of crust with little earthquake activity, bounded by narrow zones of high activity.
There is some activity within the plates. If it looks like the U.S. and Western Europe are particularly hard hit, that reflects the concentration of sensitive instruments capable of detecting tiny earthquakes in those regions.
This, of course, was all motivated by a desire to achieve a deeper understanding of the unique planet on which we live. You wish. I have a bridge in Brooklyn you might be interested in. Or, more in keeping with the Internet age, I have $20 million dollars to share if you give me your bank account number so I can deposit it. No, the motivation was to see if anyone was cheating on the Nuclear Test Ban Treaty.
The first maps published showed a cluster of dots in the Sahara Desert. Anyone reading the papers could figure out that those were French nuclear tests (at that time France controlled Algeria). When the scientists analyzing the data included a short note to that effect, all sorts of red lights and sirens went off in the Pentagon and the State Department, and the note was removed. Meanwhile, every geologist in the world heard the story. You have to wonder what the Pentagon and the State Department accomplished (their children and grandchildren are now in charge of Homeland Security.)
The diagram above, from a 1969 paper by Isaacs, Oliver and Sykes, shows the different types of plate interaction
This pioneering computer reconstruction, done in 1965, is so good that new data have had almost no effect on it. The actual edge of the continental crust is buried under the edge of the continental shelf and the submarine contour used in the reconstruction is only an approximation. Minor gaps and overlaps are expected and not a concern.
Sea-floor spreading creates symmetrical patterns on both plates. Thus, if one plate is entirely subducted, it is still possible to reconstruct some of its history from the surviving plate. This is the case off the west coast of North America, where a now-vanished plate can be reconstructed.
Created 21 May 1997, Last Update 24 May 2020