The heat that makes the ground tremble

Variations in the thickness of the lithosphere caused by differences in temperature clarify the origin of earthquakes in Brazil

It used to be impossible to explain the reason why the earth shakes a few times a year in regions so different as the west of the state of Goiás, the east of the Pantanal (the swampland), the northeast of the state of São Paulo, and the western part of the state of Minas Gerais. It was also difficult to understand why an almost eternal calm reigned along the Paraná river, in the north of Minas Gerais, in the east of Goiás, or in almost all the interior of Bahia. All these areas are to be found within the vast blocks of rock that form the Earth’s surface, the tectonic plates – and they ought to be equally stable.

It was already known that the big earthquakes spring only from the limits of the tectonic plates: the collision of one plate with another, like a piece of marble pushing another, generates enormous pressures that deform and break off the edges – strictly speaking, it is only when they are intense that they ought to be called earthquakes. An attempt was made to understand the smaller tremors in the inside of the plates by means of apparent signs that they may have left, such as cracks, differences in the level of the blocks of rocks, and other open air scars.

But no explanation leapt into sight. The reason for these phenomena seems to be further down, hundreds of kilometers from the surface, in the lithosphere, the most rigid and the coldest layer of rocks that covers the planet. On the basis of information gathered over the last 12 years, geophysicist Marcelo Assumpção, from the Astronomy, Geophysics and Atmospheric Sciences Institute (IAG) of the University of São Paulo (USP), concluded that the areas most subject to tremors inside the plates have a thinner lithosphere. This is the reason why they release more easily the pressure resulting from the movement of the plates, the origin of earthquakes. Contrariwise, in the regions with a thicker lithosphere, also inside the plates, the tension is diluted and only rarely causes tremors. In other parts of the world, the fragility of the lithosphere is directly associated with the frequency of earthquakes, as in the Mississipi valley, in the United States, at points of the Alps, the extensive mountain range in thesouth of Europe, and in the north of China.

In Brazil, many people think there are not, but there are indeed earthquakes – or rather many seisms, the technical name that designates tremors in general. There are about 80 to 90 a year, most of them of a magnitude lower than 4 on a scale that goes up to 9 – they are therefore relatively weak, above all when compared with the violent tremors that occur, for example, in Japan, when one of the most intense earthquakes ever recorded destroyed, in 1923, about 440,000 houses and killed 100,000 persons. Located in the continental part of the South American plate, Brazil is a region regarded as stable, albeit subject to pressures from the Nazca plate to the west, which constitutes the bottom of the Pacific and generates the shocks in the Andes, and in the Mid- Atlantic Ridge to the east.

It is precisely the Nazca plate diving down onto the South American plate that causes the intense tremors in Acre: the most recent, at 3:24 in the small hours of June 20 of last year, reached magnitude 7, but as it was deep down was hardly felt by the inhabitants of Cruzeiro do Sul – last century, there were five seisms in the region with a magnitude of over 7. The largest seisms from the interior of the tectonic plates were a little blander, reaching magnitude 6, but even so releasing the energy corresponding to 30 atom bombs like the one dropped on Hiroshima at the end of the Second World War. In general, these tremors occurred in the central and southeastern regions, which have been accompanied by researchers for a longer time than the others. This was also the area studied by the team from the IAG, in conjunction with scientists from the University of Brasilia (UnB).

With his team, Assumpção analyzed an area of almost 2 million square kilometers, equivalent to one quarter of the Brazilian territory – a rectangle delimited at the north by the cities of Cuiabá, in Mato Grosso, and Milagres, some 150 kilometers from Salvador, in Bahia, and at the south by Assunción, in Paraguay, up to a point in the Atlantic Ocean located about 300 kilometers from the city of Santos in the state of São Paulo and 150 kilometers from Cabo Frio, on the coast of Rio de Janeiro. In this space, about ten tremors with a magnitude equal to or higher than 3 occur, enough to be perceived without the assistance of seismographs, devices that detect seismic waves, as the vibrations caused by the quakes are called.

The researchers determined the thickness of the lithosphere in an indirect way, by means of seismographs scattered over 59 localities which since 1992 have recorded seismic waves. There are two kinds of waves generated by seisms, and both of them, like sunlight plunging into a swimming pool, undergo reflection and refraction when passing through rocks that are more or less hard: the P (primary) waves that cross through any part of the inside of the planet and arrive first at the surface, while the S (secondary) waves propagate at a lower speed and only in solid rocks.

Analyzing the arrival times of the P and S waves at the surface, after having been generated by an earthquake on the other side of the world, has been the means by which the study of the deeper layers of the Earth is advancing. Through these vibrations that it was deduced, in 1906, what the center of the planet ought to be like – an immense and compact sphere of iron that remains liquid at a temperature close to 3,500°C. They also made it possible to map the seismic regions around the globe more subject to tremors, which coincide with the limits of the tectonic plates, by which the internal energy of the Earth escapes most easily.

Now, duly interpreted by Assumpção’s team, the P waves are revealing where it is most probable that the Earth will shake – these are the regions in yellow and in red on the map. “At these points, the seisms are not the result of the meeting of the plates, but of the internal fragility of the plates”, Assumpção says. In the regions with a thinner lithosphere, more subject to the building up of tensions, the point of origin of the tremors – the hypocenter – is to be found at less than 5 kilometers from the surface. “The seisms are superficial, but the causes are deep.” The researchers did a CAT scan of the lithosphere, in the same way that doctors examine the inside of the body. They analyzed the constitution of the depths of the planet every 50 kilometers down until they arrived, obviously with a poorer definition, at 1,300 kilometers, almost one fifth of the distance to the center of the Earth.

It was also by tomography of the deepest layers of the planet that Assumpção, years ago, presented the hypothesis that the tectonic plates did not move by drifting, like a raft without a sail. According to his model, these immense blocks of rock move apart or collide, making the continents wander very slowly around the globe and sometimes causing earthquakes, as a consequence of the movements of a large part of the mantle, the layer below the crust, at depths that in Brazil can reach 700 kilometers. Before this study, published in Nature in 1995, it used to be imagined that only the upper layer of the mantle, at no more than 200 kilometers, was capable of pushing the plates (see Pesquisa FAPESP issue No.53, of May 2000).

Assumpção, this time, accompanied 10,000 records of P waves, the speed of which can vary from 6 to 13 kilometers a second, with the purpose of analyzing the profile of the lithosphere – the hard layer of rocks that includes the crust, the layer up to 40 kilometers thick that covers the surface, and the outermost band of the mantle, at a depth of from 100 to 200 kilometers. Finally, he concluded that the greatest seismic activity occurs preferably in regions where these waves were up to 2% slower at depths from 150 to 250 kilometers. The slower speed was interpreted as a result of higher temperatures, since the waves are propagated more slowly in hotter rocks.

According to this approach, the hotter regions correspond to the higher limits of the asthenosphere, the malleable part of the mantle, with temperatures close to 1,300°C, which occupies the first 200 kilometers below the lithosphere. After discovering where the asthenosphere was closest to the surface, it was easy to determine the thickness of the lithosphere: with an average temperature of 1,000°C, the extent of this layer corresponded to the distance that was missing to reach the crust. Accordingly, if the asthenosphere were not very deep, the lithosphere would be thinner.

This was how a set of maps was born that indicate that the thickness of the lithosphere in Brazil can vary from about 100 kilometers, precisely where there are more tremors, to about 300 kilometers, where quakes are very rare. According to this study, which will be published in the Geophysical Journal International, the thinnest point of the lithosphere, with a depth of from 100 to 150 kilometers, is to be found in the region of Iporá, a municipality to the west of Goiás, where on average two tremors a year are recorded with a magnitude equal to or higher than 3. What seems to be little is in actual fact a lot, when compared with the environs of Goiânia, the south of Goiás, and the region of Belo Horizonte, in Minas, where the lithosphere is thicker – from 250 to 300 kilometers – and only one or other little tremor every 200 years has come to notice.

This study elucidates the reasons why some of the largest earthquakes in Brazil have occurred. One of them, with a magnitude of 5.4, occurred in 1964, in the region of Miranda, east of the Pantanal, in Mato Grosso do Sul – Assumpção found that this was another area where the thickness of the lithosphere should not exceed 150 kilometers. This must also have been the reason for the 6.2 magnitude tremor recorded in 1955 in Porto dos Gaúchos, a municipality 300 kilometers to the north of Cuiabá, which is to be found at the limit of the area analyzed in this research.

Proofs on the surface
In the region of the Northeast, the tremors are more frequent, but not as strong as in the central region of the country. At the end of the 80’s, in the course of four years, from 1986 to 1989, there was a succession of tremors in João Câmara, in the state of Rio Grande do Norte, the strongest with a magnitude of 5, damaging hundreds of houses. Two months ago, in June, in the region of Belo Jardim, 50 kilometers from Caruaru, in Pernambuco, there was a series of small tremors, of magnitude 3. “We do not yet have precise measurements, but in the more active regions of the Northeast, in Rio Grande do Norte and in Ceará, the lithosphere is also possibly thinner”, Assumpção comments.

Even centered on the behavior of the waves at hundreds of kilometers below the surface, his study is not disconnected from the contours of the landscape, because the regions with a thinner lithosphere, being weaker, are also where the magma produced by the heat of the asthenosphere can escape more easily. This is how the so-called intrusions are originated, which are rocks melted from the base of the lithosphere that rise to the surface – it is the same material that, in larger quantities, form volcanoes. The Agulhas Negras peak in the Itatiaia National Park, on the border between the states of São Paulo, Rio and Minas, at an altitude of 2,787 meters, is one of those more fragile points of the lithosphere, where the lava found a way to spill out, some 60 million years ago. In Goiás, something more modest and more ancient: the Engenho hill, 200 meters high, in Iporá, consists of the remains of intrusions occurred 80 million years ago.

Assumpção intends to expand the area studied and conclude the mapping of the lithosphere of the whole of the country. Done today at the pace that is possible, the work enjoys the collaboration of a group of Brazilian institutions – amongst them, the Federal Universities of Mato Grosso do Sul and Rio Grande do Norte, besides the State of São Paulo Technological Researches Institute – and of foreign partners, such as the Institute of Earth Sciences, of Spain, and Northwestern University, of the United States. It is not easy: the areas of the Center and of the Southeast alone have taken up almost ten years of work.

But Assumpção still hopes to be able to count on the participation of companies, because this kind of mapping – already concluded in countries like the United States, Russia, China, and Australia, amongst others – facilitates the search for minerals: it is more probable to find diamonds, for example, in old regions with a thicker lithosphere. Perhaps it may be even more difficult to discover where the heat comes from that makes the asthenosphere shallower, thins the lithosphere, and opens the cracks so magmas can break through. The specialists are far from a consensus.

Assumpção believes that these phenomena may be linked to the Trindade Plume, a column of very hot rocks for the mantle and relatively thin (about 100 kilometers in width). As proposed 30 years ago by geochemists to explain the occurrence of intrusions, like the Engenho hill and the Agulhas Negras peak, the plume could be formed by almost molten rocks coming from a fixed point at the base of the mantle, close to the core, some 3,000 kilometers in depth, which rise up and heat the lithosphere. Millions of years ago, part of this plume that reached the surface may have formed the submarine mountain range near to Vitória and the Trindade archipelago, off the Brazilian coast.

As the source of heat close to the core cannot have been extinguished, more hot rock continues to rise up and to cause volcanism and intrusions on the crust, while the surface moves with the tectonic plates. If it were so – here is the first problem of this idea -, the dozens of intrusions attributed to the plume ought to be, in some way, aligned in accordance with their age. But they are not: they seem to be mixed up, the older ones with the more recent ones together, without any clear order. The second problem is that the plume is thin and deep enough to escape from the examinations of tomography and its effect on the seismic waves is almost imperceptible. There are those who prefer to believe that the magma does not come from such deep regions, but from shallower portions, no more than 700 kilometers down. In this case, the raised parts of the asthenosphere could be caused by convection currents, like those that move boiling water in a saucepan, confined in the upper part of the mantle. It is not easyto prove either.

The Project
Crustal Structure and Seismicity of the Southeast (nº 01/06066-6); Modality Regular Line of Research Grants; Coordinator Marcelo Sousa de Assumpção – IAG/USP; Investment R$ 89,141.42 (FAPESP)