Dora the Explorer's Indian Ocean (Banda Aceh) Earthquake Science Page

Dora the Explorer's Indian Ocean Earthquake Science Page

Credit goes to the Georges of the geology newsgroups for nicknaming me Dora the Explorer. Tee, hee. I decided the term fits!

Page edited on 3/28/05

Page restructured on 4/16/05; catastrophic volcanos and U.S. quakes are now at Dora the Explorer's Catastrophic Volcanos and Earthquakes Page

Coastal Sumatran district capital city of Banda Aceh before and after the quake

Coastal Sumatran village after the quake

This is not entirely from the tidal wave. The quake dropped parts of northern Sumatra below
sea level, leaving previously low lying land permanently under water.

March 28, 2005 (Easter) Northern Sumatra Quake

On March 28, 2005, the day after Easter, an earthquake of atleast magnitude 8.5 struck northern Sumatra. Experts say it affected the same fault, but ruptured southward, relieving pressure just south of the area that slipped during the Christmas earthquake.

It appears that no tsunami or only a small tsunami was generated by this quake. Energy may have been sent southward toward a region where there is no major land. There was vigorous damage, and it appears to have been worse on outer islands such as Simeule and Nias.

NGIS map of area affected by March 28 quake.
Red star is the epicenter of the March 28 quake. Yellow star is the epicenter of the December 26 quake. In between is Simeule Island, and the yellow star is approximately on a geological quake directing feature (a twist and discontinuity in the fault system that doesn't let slippage along the fault get past it) called the Nias Elbow - and probably affected the area immediately south of the Nias Elbow. Main rupture is said by NSGS to have been 30 km deep.

Kerry Sieh, one of the main experts on the tectonics of the region, was quoted in an article after the Dec 26 quake, as wondering when the region immediately south of the part of the fault that slipped in the Dec. 26 quake, was going to go. He didn't necessarily expect this to happen immediately. It was a logical place to think tension was building up, but the Nias elbow itself has not experienced serious quakes in recent history and Sieh bemoaned that they had no particular system in place to watch this area nor the region affected by the Dec 26 quake.

CNN coverage of April 27, 2005 quake in northern Sumatra

Overview of the Geology of the Dec 26, 2004 Quake

The December 26 Indian Ocean Earthquake, otherwise known as the Banda Aceh Earthquake, was centered off of Sumatra, in Indonesia, near the small island of Simeulue. Sumatra is located on a system of subduction trenches where many plates meet, in the western Pacific Ocean. The majority of the world's most violent earthquakes and volcanoes occur along this system of subduction trenches.

The quake occurred along the subduction trench where the Indian and Australian plates and the Sunda plate converge. Along this trench, the Indian and Australian plates consist of oceanic crust, and because they are heavier than the continental Sunda plate crust, the Sunda plate tends to slide over the two oceanic plates, while the oceanic plates slide under the Sunda plate. Simplistic models of the trench depict the Indian plate sliding under the Sunda plate - but in the case of this earthquake it appears to have actually been the overriding Sunda plate (sliver arc portion thereof) that moved, even though the initial slippage appears to have affected the Indian plate. That explanation is subject to change as scientists better understand what happened.

The quake specifically caused a strip of the sliver plate along the edge of the subduction zone, attached to the Sunda plate, between nine and twenty five or fifty miles wide, and roughly 600 or 700 miles (1200 km) long, beginning just off of the Sumatran island of Simeulue and extending northward along the subduction trench into the Andaman Sea, to jump upward some thirty feet, by one estimate, and to jump many yards westward over the Indian plate (in the direction of the coast of India - if I have my directions confused again!). It appears that the subduction trench itself, and an entire section of the plate, moved westward, taking part of northern Sumatra with it - but this remains to be confirmed. That entire section also dropped in height - such that parts of northern Sumatra that were just above sea level are now below sea level, and are permanently flooded.

These maps and sketches of the area show roughly where the quake occurred and the region where the aftershocks occurred, which bears some relation to the portion of the plate that slipped in the quake. These maps show something of the geology involved. The reason why I have included more than one of them is that the NEIC's map misrepresents the boundaries of the sliver plate, which is less commonly called the "Burma microplate". While characteristics of a point of the plate just south the epicenter did affect the course of the quake, the sliver plate extends without interruption from northern Burma past Sumatra. If you notice the map for the March 28 quake above, the Burma microplate has disappeared. The term was never very popular among geologists. It is semifictional.

The star off of the (unlabelled) small island of Simeulue depicts the location of the quake's epicenter.

The yellow dots indicate the locations of aftershocks. The dots are sized by magnitude.

This version of NEIC's aftershock map shows some of the fault systems, instead of the inaccurate Burma plate boundaries that are on their earlier versions.

Maps from NEIC web site, Kerry Sieh's Sumatra project web site, and Danny Natawidaja's web site.

The Indian Plate is part of the Indian-Australian plate and meets its mate, the Australian plate, at a subduction trench, just south of where the epicenter of the quake occurred. The subduction trench runs from the southern tip of Burma, around the outer Indonesian island arc, and curves around south of Java eastward toward New Guinea. At various locations along the trench, it is called the Andaman trench, the Sunda trench, and the Java trench. The Sunda plate is part of the Eurasian plate, and the fact that it moves in a subtly different direction than the Eurasian plate contributes to the forces along the subduction trench.

The quake directly affected a strip of the outer border of the Sunda plate along the subduction trench. Early maps and descriptions by NEIC, and media reports that quoted them, and a few other discussions of the tectonics of the area, refer to a sliver plate, the Burma microplate, and the Burma plate. To their credit, in their most recent maps, the NEIC have finally stopped referring to it at all. They simply draw the subduction trench and some indication of the transform fault system behind it. Sometimes this "sliver plate" is simply called the subduction fault zone - which is basically what it is. Since the term "Burma microplate" is falling back out of use, knowing what the term refers to is less critical to understanding the geology of earthquakes in teh region than it was in January of 2005.

The Meaning and Significance of the Burma Plate

This sliver plate consists basically of a steadily widening strip of slip-strike faults behind the subduction trench, that is also folded and uplifted in bands as this strip of plate rises and rides over the subducting India plate. Because along the western portion of the trench, the Indian and Australian plates converge with the Sunda plate at an oblique (not 90 degree) angle, slip-strike faults formed all along the edge of the Sunda plate, to move the plates north and south relative to each other as the Indian and Australian plates slide under the Sunda plate.

The term sliver or Burma plate is used, defined, and depicted in confusing ways. On NEIC's maps of the quake's location and tectonics, you will see southern and northern boundaries drawn for the Burma plate, where they are not located. The northern boundary of the sliver plate is not the northern end of teh Andaman Sea, and the sliver plate does not end immediately south of the epicenter of the December 26 quake. The "sliver plate" runs along the edge of the subduction fault from northern Burma past Sumatra and possibly past Java.

NEIC wrongly took their northern and southern borders for the "Burma plate", from the data of Peter Bird (see below). Peter Bird drew artificial lines around groups of tectonic entities in an extremely abstract effort to explain analytically how the movements of these entities affect each other. On his web site he specifically states that some people (including NEIC) have, in trying to explain the recent quake, mistakenly taken his lines literally, and they were never meant to be taken literally. Peter Bird has a special term for his analytical groupings of tectonic features, that need not trouble us here; and he marked them on his maps with special grid lines. Peter Bird explains on his web site, that somehow the version of the maps that USGS and other researchers got was missing some of the grid markings that identified teh lines as not literal geological boundaries. To complicate the matter, the groupings of tectonic features were arrived at by trigonometric calculation, which Peter Bird explains on his web site is just one of a number of equally valid possible solutions. Nevertheless, the notion of a "solution" to how to divide the tectonic features in the region was born - and apparently cannot now be erased from the cerebral databases of technicians at NEIS. Even though a literal interpretation of those lines as plate boundaries is obviously nonsensical, and almost noone agrees with the notion of putting plate boundaries for the sliver plate in those locations.

Many geologists think it is more proper to call this zone a subduction fault zone, then a plate. Kerry Sieh and his students call it a fault zone some times, and a sliver plate or the Burma plate at other times. Danny Natawidjaja, in his dissertation on the tectonics of Sumatra, completely and clearly described the tectonics of the region without ever mentioning a Burma or sliver plate at all. Actually, the student definitions of a subduction fault zone and of oblique subduction would lead one to expect a strip along the overriding plate with uplifted ridges, cracks, and slip-strike faults, that behaves entirely the way this "sliver plate" does. This leads one to question why the specialists are persistently telling us poorly defined details that we don't need to know, as well as using poorly defined terms like "Burma plate" as if the terms by themselves, with illogically drawn lines on maps, explain the violence of the December 26 earthquake.

The principal reason to refer to this strip of crust as a plate, is that geologists like to think of the trigonometry of the motion of crust within plates, in terms of subplates. Tectonic plates are actually very dynamic entities; for instance, volcanic activity causes them to spread and split apart into entirely separate plates. Pressure and friction from surrounding plates cause large areas of plates to move slightly differently from the rest of the plate. If a region of a plate is proven to be moving over time differently from the rest of the plate it is part of, geologists refer to it as a subplate. That notion allows them to make all sorts of geometric and trigononetric calculations about the region and its motion. If they didn't think of it as a plate they could still predict how the rock would behave, and explain how it does behave, but this way they can generate computer models, and draw confusing lines on maps.

The Burma/ sliver plate is called a plate mainly because the region behind the subduction trench is spreading apart, and the fault zone accomodates rather complex north/ south motion between the plates, this strip of plate often moves in ways that are fairly unique to it, relative to the plates around it.

Milsom's version of this, from his chapter on the tectonics of Sumatra (below), is "The suggestion, originally made by Fitch (1972), that the oblique motion is to a first approximation accommodated by orthogonal subduction at the trench and dextral slip along the Sumatran Fault, is now widely accepted. To the extent that this is true, the forearc region must be decoupled from both the Indian Ocean and Eurasia. The commonly used term 'sliver plate' (e.g. Curray 1989) suggests more strength and rigidity than could reasonably be expected of such a long and narrow strip of lithosphere, and any analysis of subduction beneath Sumatra must take into account the probability of independent movements of forearc fragments (e.g. McCaffrey, 1991)."

In other words, the globe must be peppered with "forearc sliver plates"; everywhere where plates converge at a shallow oblique angle. The following web site confirms that suspicion. Plate Techtonic Modeling: Tools and Methods . Sliver arcs are in fact routinely computed for subduction faults.

The term "forearc" means along the outside of the semicircular line along which the plates converge, at the subduction trench.

In addition, the origins of this strip of plate are controversial. It may contain some geologically unique rock. Some researchers think it is the splintered edge of the Sunda plate, or speak of it as having separated from the Sunda plate; others say that it broke off from the great supercontinent of Gondwanaland at a later time than surrounding plates and merged with the Sunda plate at a late point in time. Understand that, unique rocks notwithstanding, the historic mechanism of the creation of the sliver plate, that is really only a fancy term for a fault zone to begin with, by spreading slip-strike faults, otherwise to be thought of as cracks in the ground, appear to me to require that the sliver plate originated as the edge of the Sunda plate.

Earthquakes in the region commonly take the form of subduction into the trench followed by aftershocks along the slip-strike faults, but not always. The December 26 quake apparently involved subduction into the trench, and according to what Joseph Curray of the Scripps Ocean Institute told me, there may have been some slight compensatory motion along faults on the back side of the sliver arc in the Andaman Sea.

The sliver plate actually begins in northern Burma, where its eastern border is defined by the Sagaing fault zone, that runs the length of Burma. At the Andaman Sea the western border of teh sliver plate is defined by the subduction trench, and the eastern border is defined by a series of slip-strike faults that have spread steadily eastward over time over a basin that is spreading outward. At the northern tip of Sumatra the fault system converges in a slightly complex manner into another system of faults that include a fault off the western coast of Sumatra that runs the length of Sumatra and is no longer active, and the Sumatran Fault Zone. The Sumatran Fault Zone runs through the central mountain range.

Immediately east of the central Sumatra fault, is a line of very explosive volcanoes, caused by the melting of the Indian and Australia plates underneath the Sunda plate. This creates a particularly sticky, gassy and explosive kind of magma that has contributed to such historical volcanic disasters as Toba, which nearly brought the human race to exctinction 75,000 years ago, and Krakatowa, which caused massive environmental catastrophe in 1883.

The mountain range and in fact the entire outer island arcs were created by a combination of uplift in the fault zone that is created when the plates slide over and under each other, and the volcanic activity.

The fault system and its geological irregularities

The course of earthquakes along the subduction zone is determined by the interaction of plates and subplates that meet off of the northern coast of Sumatra (such as the Indian and Australian plates), the curvature of the subduction zone, the changing angles at which the plates come together, and discontinuities and changes over time in the transform faults, behind the subduction zone, that take up north-south motion as the plates move together at not a 90 degree angle.

The transform fault system has changed, moved out, expanded, and formed new faults over time. The location and extents of many of the older and smaller sections of the transform fault system are not well understood. Confusingly, some faults and sections of faults are named differently by different scientists, and the extent of faults with continuities and discontinuities in the faults are drawn differently on different maps, and understood and interpreted differently as well.

An ancient line of faults that once joined with the West Andaman fault, and on some maps still does, runs the length of Sumatra, off of its western coast, approximately halfway between the subduction trench and the western coast of Sumatra. The locations of epicenters major quakes and their aftershocks, relative to this fault and other transform faults that connect it to the juncture of the Australian and Indian plates (called an investigator fault zone) and to the main Sumatran Fault that is the currently active fault and runs down the middle of the Sumatran mountain range, makes it appear to me that this ancient and allegedly quiescent line of faults is probably more important than it is commonly believed to be.

This ancient fault was probably the original inner boundary of the Burma microplate/ sliver arc, but as spreading occurred the active back arc fault moved to the current Sumatra fault zone that runs down the center of the island of Sumatra. Picture a sort of triangular wedge, the long side of which is formed by the semicircular subduction trench, where the two plates meet at a not 90 degree angle. They slide under each other perpendicularly, more or less. But much of the north south motion is taken up by transform faults at the back of the triangle.

The name of the long fault off the coast of Sumatra depends on who you ask, partly because geologists disagree on how much of it to call an extension of the West Andaman Fault. Some call it part of the West Andaman Fault. It is given other names, and some don't give it any name and also don't depict it as a continuous line. One popular name for it is "Mentawi Fault", and some show it connected to the current Sumatran Fault Zone by the Batee Fault, which is on the Nias Elbow, just south of the epicenter of teh quake. Some actually have the West Andaman Fault ending just south of the epicenter of the December 26 quake, which is almost on the southern boundary of the section of plate that moved.

The West Andaman Fault continues to physically appear related to the location of aftershocks and the boundaries of the section of plate that moved, all the way through the Andaman Sea region. However, there are a multitude of slip-strike faults throughout the back arc region in the Andaman Sea, and experts in the area have told me that most likely motion occurred along a number of them.

The location and strength of quakes and volcanoes along the subduction fault zone is affected by the fact that the angles at which the Indian and Australian plates and the Sunda plates meet at the subduction trench, changes along the course of the trench. The several reasons why that happens help determine where the angles at which the plates meet changes. First, the fault line itself curves. That means that in some places the plates meet almost dead on, in some places they almost slide past each other as if it were a slip- strike fault like the San Andreas fault in California, and at some places they meet at an angle, and that angle changes.

Second, alot of small tectonic plates meet in the southwest Pacific ocean off of Southeast Asia and Australia, and the huge Eurasian plate breaks in outheastern Asia into several subunits that have their own motion with respect to each other. The pressure and friction of the numerous small plates in the area against each other as each moves according to its own course, effectively prevents any tectonic plate in the region from moving in a very straightforward manner. The plates and subplates actually turn and twist and appear to sort of rotate relative to each other, and the jostling causes the plates to internally distort.

Third, the Australian and Indian plates converge together at a slight angle as they subduct under the Sunda plate. The point at which that happens is not by coincidence just south of the epicenter of teh December 26 quake. As they meet, they actually deform and shrink. Much of this action probably takes place in something called the Investigator Fault Zone in the Indian Plate roughly perpendicular to that point.

Another contributing factor in the way the plates move together is that there are discontinuities in the system of slip-strike faults that form the back of the sliver arc, as that sliver arc formed and then widened over time. In the Andaman Sea, between Burma and Sumatra, the back of the sliver arc was once the prominent West Andaman Fault, which still plays determinant roles in many of the quakes and aftershocks in the region, even though the main fault system in the region is now further east. These faults all converge sort of on the West Sumatran fault system. A couple of million years ago, the sliver arc was bounded by the West Andaman Fault, which ends near where the epicenter of the December 26 quake occurred, and another, now inactive, fault that ran the length of Sumatra off of the coast of Sumatra. As the plate spread out across the sea behind the arc where the plates converge, and the sliver plate widened, the fault system spread eastward, to its current boundaries.

Where the angle of convergence of the plates changes sharply and the faults are discontinuous, transtension zones form. Most of the changes in the angle of convergence and fault discontinuity occurs near Sumatra. There is a transtension zone just north of Sumatra, a transtension zone in the straight between Sumatra and Java where Krakatoa is not coincidentally located, and a transtension zone immediately south of the epicenter of the December 26 earthquake, roughly near the island of Nias. The epicenter of the December 26 quake was just off of the island of Simeulue. John Milsom (see geology chapter on Sumatra below) thinks that the Toba volcano crater is associated with this transtension zone - or with the zone where the Indian and Australian plates compress together as they subduct into the trench, which is associated with this transtension zone.

Kerry Sieh is one of the researchers who call it a transtension zone. Researchers really aren't sure what all goes on at that point, and Kerry Sieh has guessed that several things might be going on at that site. In the dissertation of one of his students, below, it was termed the "Nias elbow", and the theory was that the plate is locked and motion sort of "hinges" at that location. Some explanations actually include the notion that there are several more microplates, they have a boundary at that location, and two more of them are located along the coast of Sumatra. That appears to be distinctively a minority opinion. Joseph Curray told me he doesn't think there are real discontinuties at that location - and told me I really need to read his forthcoming article, the publication of which he is trying to speed up. It appears that his ideas on the subject have changed since he wrote a paper on the subject some years ago.

Some things are definitively known about the transtension zone near the island of Nias. Earthquakes in the region have occurred either in the Andaman Sea, or off of Sumatra affecting exclusively sections of the plate SOUTH of Nias. The section of the plate between northern Sumatra and Nias has been completely quiet for apparently as long as anyone knows. It has been quiet for so long that Kerry Sieh commented that the quake came as a complete surprise. I have to speculate that some manner in which the plates were locked at that point had something to do with the violence with which it came unstuck, and with the amount of motion that occurred.

The epicenter of the December 26 quake was just north of the Nias transtension zone, and all of the motion of the quake occurred immediately around and NORTH of the epicenter. No motion spread south of that point. Such a thing is slightly unusual but hardly unheard of, and when it happens, it is usually determined by bends, twists and discontinuities along fault zones, any of which could be going on near Nias. For instance, someone on one of the geology newsgroups said that there is a bend along the San Andreas faults in California that earthquakes don't pass.

One thing that is certain is that the section of the plate around Nias will not fail to move forever. The entire subduction zone is moving westward. That entire edge of teh plate is steadily moving relative to the Indian and Australian plates. In his press statements and comments immediately after the quake, Kerry Sieh said that he thinks that that section of the plate may be locked in such a manner that it can be expected at some point to break loose. He headed promptly for Sumatra to do more investigating of that section of the plate.

The Significance of Kerry Sieh's Work on the Coral Reefs off Sumatra

Kerry Sieh and his graduate students have done extensive work on changes over time in the geology of the Sumatran fault system along the southern part of the subduction zone, focusing specifically on the western coast of Sumatra. They present much of their work in pretty clear terms on their web sites, below, and most of their published papers and a key dissertation are also available on those web sites, as are good fault maps of the region.

I found the specific information on rises and falls of the coral reefs over time as they pertain to current risks of earthquakes at specific locations along the Sumatra coast, hard to follow, but since January, things that Kerry Sieh has explained on his running journal of his reconaissance work in Sumatra after the quake and about the quake and the danger of further quakes in the region, to the media, have made his findings and the reasons for his concerns alot clearer.

As Sieh explained on a February 15 radio interview, two days after he returned from his reconnaisance work in Sumatra, until ten years ago it was broadly known that there had been very big quakes historically in the Andaman Sea and off of Sumatra but little was specifically known about them. The work of Sieh and his graduate students initially focused on demonstrating that growth patterns in coral reefs over time reveal changes in sea level, or, rather, in the level of the sea floor relative to sea level; and that reverses in patterns of change in sea floor depth corresponded with when the quakes occurred. He was then able to map what land was affected how much by the historical quakes, and also determine where the greatest magnitude quakes would have occurred. This led to such knowledge that we currently have as the fact that quakes always propagate (rupture) either south or north of the Nias elbow.

Sieh explains in his media interviews, that as the oceanic (Indian-Australian) plate sinks into the subduction trench, it tends to pull the edge of the overriding continental plate downward with it. That accumulates some of the tension that is eventually resleased by the big earthquakes. The subduction occurs at a very shallow angle; 10 degrees, according to Sieh's media explanation; so when the overriding crust finally breaks loose from the downsliding plate, it jumps up a little but over alot. So by his various guesses, the Dec 26 quake caused the sea floor to rise a foot or two (or maybe a yard), and move westward ten to twenty meters (tirty to sixty feet). And as I explained that he explained above, the land behind it may sink, leading to a fall below sea level of previous dry land along the back part of the arc.

As the sea floor on the continental plate off of the coast of Sumatra, near the fore arc islands, is pulled downward into the trench, the coral reefs sink - or the sea level rises relative to teh sea floor, and the corals that grow off of the sea floor.

When the overriding continental crust finally breaks and the quake occurs, the sea floor abruptly snaps upward. This raises the coral reefs out of the water. Of course, the parts of the coral that are above the water stop growing at that point.

Much of Sieh's work consists of minutely sampling and examining the coral reefs off of the coastal islands of Sumatra, and apparently one of his graduate students has focused his life's work on this coral - as well as on mapping the faults around Sumatra.

Sieh and students focused their work on the central and southern parts of the Sumatran coastline where historical quakes have occurred the most recently (in the 19th century). They say that they really weren't sufficiently watching what was going on with the sea floor off of Northern Sumatra, and, as far as I could tell from my own reading of their work, they were not watching the Nias elbow itself or were not watching it very closely. They didn't expect major action to occur there since it hadn't done lately, yet Sieh thought they really ought to get back there when there was time and resources and cover that territory.

They did see enough to worry Sieh. He saw a pattern of slowly and steadily sinking land off of the central coast of Sumatra, south of the Nias elbow. He knew that the last major quake there was maybe 170 years ago and they tend to come every 200 years. He wasn't that worried about the northern coast of the island. He began warning people along the entire island chain - and some of them listened. The people of Simeulue, the northernmost island immediately south of the epicenter of the Dec 26 quake, turned out to need the least warning, because their traditions told them that if there was a quake they must run to high ground, and they did. Simeulue right next to the epicenter had the lowest death rate from the quake and the tsunami of all of the entire Indian Ocean coastline. Many people in other areas who got Sieh's warnings weren't particularly concerned about it.

Immediately after the quake, Sieh said that he was expecting a major quake to occur off of central Sumatra and that strain was still there and probably made worse by the quake to the north. He really didn't explain very specifically to the public how he arrived at that conclusion by noticing the warning sign of steadily sinking sea floor.

But when Sieh got to Sumatra, after the quake, he reported on his online journal of his reconnaisance work, on his web site, that he saw all kinds of significant changes along the coast of Sumatra. Land on northern Sumatra sank below sea level. Along Simeulue, and the northern part of the island, coral reefs on newly released sections of sea floor stood out of the water, and other ground that wasn't previously there appeared as well. But off of the island of Nias itself, Sieh reported noticing dead palm trees in standing water. They hadn't just been killed by the tsunami, nor by land that sank because of the quake. They were killed by sinking of the sea floor over time - and the sea floor off of the coast of Nias was still pulled down. The December 26 quake did not release it.

The consequence is that Sieh spent February and early March, warning the public that a major quake is due near Nias and off of central Sumatra sometime in the next thirty years.

In the radio interview someone asked if quakes in the area are linked. Seven of the ten giant earthquakes of the twentieth century occurred between 1950 and 1965, five of them around the northern Pacific rim, so that he intuitively suspects there is some link. (I've fixed my probable misunderstanding of what Sieh said with a smiliar statement in an article by Sieh in a recent issue of Nature.)

McCLoskey, in his recent Nature article, says that subduction-zone earthquake4s are often coupled. In the Nankai trough subduction zone to the Southeast of Japan, five of the seven large earthquakes on teh Nankaido segment in the past 1500 years were accompanied by similar events on the contiguous Tonankai/Tokai segment within five years, and three of those occurred in the same year.

Meanwhile a warning by another group of scientists about accumulated strains that should lead to another quake off of central Sumatra appeared in the journal Nature just a week before the March 28 quake. This not available online, as nearly as I can determine.

Kerry Sieh also reported in mid to late February that there is little data available yet from the GPS stations that exist in teh region, nor from the statistics that ahve been gathered. He referred to five stations put by India in the Andaman Islands, the data from which was not yet available. I believe that could be this data; Preliminary estimates of the geodetic GPS survey in the Andaman & Nicobar Islands conducted by CESS. No explanation of what CESS is. They report on GPS data from the Andaman Nicobar Islands, and they report that that line of islands moved southwestward 10 meters. They did not yet have the vertical movement analyzed. Kerry Sieh also collected data from his GPS tracking stations but does not yet have it analyzed. Or did not in mid February.

In the radio interview, Kerry Sieh mentioned controversy concerning the magnitude of the quake and the length of the rupture, caused by the long period of time it took for the rupture to occur. Recent articles in Nature and Science journals argue that one reason for the long rupture time is that either two or three segments of the fault ruptured, at different rates of speed.

Researchers are now explaining that the Sumatran and Andoman - Nicobar sections of the fault have different kinds of crust on the subducting edge of the Australian-Indian plate. At Sumatra the subducting crust is young and thin, subducts at a shallow angle, and tends to hang up closely on the overriding Burma microplate. In the Andoman Sea, the subducting crust is old and thick and tends to subduct at a sharper angle. This is one reason why there is more back arc spreading in the Andaman Sea than at Sumatra. The rapidly sinking older crust in that area actually subducts downward faster than it subducts inward, and pulls the edge of the overriding continent apart. Ritzwoller's paper in a recent issue of Science makes confusing references to the temperature of the crust in the Andaman Sea, which as nearly as I can tell, may have contributed to the situation by causing the northern crust to fracture more slowly. I could not follow their advanced and rather obscure, controversial, arguments about current flows in the plastic upper layer of the mantle.

There were effectively two or three earthquakes within several minutes of each other, which early analysis mistakenly analyzed as one quake. Each segment of the fault that ruptured had its own quake. The initial segment off of northern Sumatra ruptured very quickly, generated alot of energy, and probably 2/3 of the tidal wave. The northern segment(s) ruptured more slowly, but at a greater magnitude, so that as the quake kept going it gained magnitude. The speed of the waves that were measured to measure the magnitude of the quake may also have changed, because in the north the subducting crust is cooler and denser than in the south and apparently in the north the temperature of the mantle under the spreading back arc is higher, so that the rock is softer.

General Information about the Recent Quakes in Sumatra

2004 Indian Ocean Earthquake Wikipedia free encyclopedia article with news links - this is an extremely comprehensive and well done piece. It has also been updated regularly with new information as it becomes available.

South Sumatra Forest Fire Management Project: 2004 Indian Ocean earthquake News: Thu, 30-Dec-2004 12:28 WIB(205

National Geographic News: Tsunami-Battered Sumatra Ripe for More Disasters

New Scientist: Asia Primed for Next Big Quake

BBC News: New Asian quake threat warning; BBC News resport on a nature article that discusses stress placed on the surrounding plate by the DEc 26 quake.

Nature News report on Nature journal article. A map provided with it makes it apparent that McCloskey et al use the same mistaken interpretation of Peter Bird's data that NEIC was using in January of 2005. They put a southern boundary of the Burma microplate where there isn't one.

Tues Feb 15, radio interview with Kerry Sieh, two days after his return from Indonesia to gather his info.

Exerpts from articles about and interviews with Kerry Sieh online

Newsgroups and mailing lists:

sci.geo.earthquakes and sci.geo.geology
Be prepared to be patient. Look out for the Georges - but if you are persistent people will answer questions. Daryl Krupa,, was particularly helpful at finding maps and technical papers I missed.

Maps of Regional Tectonics

Danny Natawidjaja's sketch of the tectonics of Sumatra

"The Sumatran Fault" a web site by one of Kerry Sieh's students

"The Sumatran Fault System, Indonesia" Plate 1

Prediction of earthquake from spatial/temporal clusters by statistical analyses Includes maps of the Sumatran and Andaman fault systems.

Tectonic configuration of SE Asia - Hall

The 1881 Nicobar Earthquake and Tsunami. Ortiz and Bilham. ESE 11-2.

Tectonics of southeast asia.

Southeast Asia Tectonic Framework

Very large detailed scematic of NEIC version of plates, faults, earthquakes and volcanoes in a rea. Keep in mind that the southern and northern boundaries they draw for the "Burma plate" do not exist. Keep in mind also that over time they might do away with those lines - I'm far from the only one arguing with them.

Here is a cleaned up USGS map without the faulty fault lines, and a more accurate picture of the faults in the area substituted. The map focuses on the locations of the quake and aftershocks.

Small plate map of the world

A lesson in plate tectonics - Plate boundaries of the world (small)

Major Tectonic Plates of the World

"Earthquake Spawns Tsunamis" Computer generated model of the movement caused by the quake.

What actually happened

The Science behind the Aceh Earthquake - press statement by Kerry Sieh at Caltech.

"Little Feared Fault Produced Indonesian Quake" also article/0,2403,BSUN_19093_3428793,00.html daily/12-04/12-28-04/a02wn378.htm

'Megathrust' Sumatra earthquake equal to one milliion A-Bombs CDNN News. Cyber Diver News Network

Quiet fault produced catastrophic results, Thomas H. Maugh II and Rosie Mestel, LA Times, 12/28/04

Caltech Seismological Laboratory Preliminary Analysis of Quake-related Data

Kerry Sieh's reports from the field

There are a number of good recent articles in the journal, Nature, and Michael Ritzwoller et al recently published an article, "Structural Context of the Great Sumatra-Andaman Islands Earthquake" in Science. The online pdf file did not helpfully give the publication date, and the article ahs not been indexed yet. It was submitted to Science on March 18, 2005, and I am writing this on April 4.

NEIC/ USGS's rather strange technical reports on the earthquake

As I explained above, NEIC's intial reports on the quake were vague and confusing, and in some respects outright incorrect. Most importantly, the imaginary northern and southern boundaries they gave the Burma microplate, on their maps, came from an often repeated misinterpretation of Peter Bird's data, as I explained above. Their most recent maps, particularly those of the March 28 quake a hundred miles south of the earlier quake, have dropped the Burma microplate and its incorrect boundaries completely.

Magnitude 9.0 OFF THE WEST COAST OF NORTHERN SUMATRA Sunday, December 26, 2004 Preliminary Earthquake Report U.S. Geological Survey, National Earthquake Information Center with their tectonic charts (unless they come to their senses, and remove their tectonic charts - their director wrote me that they only recently started providing such a thing).

Aftershock map as of 1/7/05

NEIS News Release on the quake

FAQ: Everything else you wanted to know about this Earthquake and tsunami (NEIS)

Completely nonsensical as far as I can tell analyses of the earthquake

Source Rupture Process of the Great Sumatra, Indonesia Earthquake (Mw=8.9) of 26 December 2004 Preliminary Results José Fernando Borges, Bento Caldeira and Mourad Bezzeghoud

The 26th December, 2004 Sumatra-Andaman Earthquake & Tsunami

Tectonics and Seismology of the Area

Active tectonics of the Sumatran plate margin and its seismic threats to Indonesia and southeast Asia by Danny Natawidjaja and Kerry Sieh.

Sumatran Plate Boundary Project - Kerry Sieh and Danny Natawidjaja of Caltech. Page contains links to a number of their articles on the tectonics of the region around Sumatra

Danny Natawidjaja's page on the Sumatran Plate Margin

Kerry Sieh's current home page

Danny Natawidjaja's home page

Kerry Sieh's journal from his reconnaissance work in Sumatra after the quake

Other version of Kerry Sieh's online journal

New USGS web page on Tectonics of Sumatra-Andaman Sea Islands Explains some aspects of the plate boundary rather clearly; particularly the kinds of relative plate motion that take place in the fault zone.

The 13 September 2002 North Andaman (Diglipur) earthquake: An analysis in the context of regional seismicity C. P. Rajendran*,#, Anil Earnest*, Kusala Rajendran*, R. Dev Das† and Sreekumari Kesavan Current Science 84(7), Apr 2003. Also has a good map.

Neotectonics of the Sumatran Fault and Paleogeodesy of the Sumatran Subduction Zone. Dissertation, California Institute of Technology, Danny Hillman Natawidjaja, 2002. (Kerry Sieh advisor; the coral reef project is his)

Peter Bird. An Updated digital model of plate boundaries. Geochemistry Geophysics Geosystems, 4 (3), 1027, doi: 0.1029/2001GC000252. Peter Bird was cited to me by the director of NEIC as the source of their Burma fault boundaries - and his web site says that they missed the information that his boundaries are not to be taken literally.

Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computer-based reconsructions, model and animations. Robert Hall. 2001. Journal of Asian Earth Sciences. 20 (2002) 353-431. Available only through library or library data base. Or online by subscription.

Cenozoic plate tectonic reconstruction of SE Asia Trebor Hall. 1997. Earlier and shorter version of above.

Curray, Joseph. New paper on SE Asian tectonics - in an upcoming issue of Journal of Asian Earth Sciences. He sent me the abstract and a key plate. He hopes the article will appear soon.

The Evolution of the Sumatran Fault System, Indonesia, Andy McCarthy PhD

Geology of Sumatra: Chapter 2, Seismology and Neotectonics chapter in forthcoming book, by John Milsom, University Collge London, UK. John Milsom wrote me that something is going on at the point where NEIS sketched the southern boundary of the Burma microplate - but what is going on is probably not a southern boundary of the Burma microplate. Peter Bird wrote on his web site that a literal interpretation of his model is one possible solution of a number of possible valid solutions, and John Milsom wrote me that he thinks that is the worst possible solution.

The Mogok Metamorphic Belt, Burma (Myanmar) and its continuation around the eastern Himalyan syntaxis to Tibet, M. P. Searle, D. J. Waters, S. Noble, A. H. G. Mitchell, U Tin Hlaing.

The Sumatra subduction zone: A case for a locked fault zone extending into the mantle Martine Simoes1 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 109, B10402, doi:10.1029/2003JB002958, 2004

Transtensional basins in the western Sunda Strait. Heidrum Lelgemann et al. Geophysical research letters, 9(6), pp 0-0 M 1 2000.

Shallow seismicity, stress distribution and crustal deformation pattern in the Andaman-West Sunda arc and Andaman Sea, northeastern Indian Ocean. M. Radha Krishna and T. d. Sanu. Journal of Seismology. 6(1): 25-41. Jan 2002.

Tectonics and History of the Andaman Sea Region abstract of Joseph Curray paper - source not given. Joseph Curray told me on the phone that this paper is out of date, and refers one to his upcoming article in Journal of Asian Earth Sciences, which he hopes will be out soon. The one detail he told me, besides confirming the hard to learn detail that the Burma microplate is another name for the sliver plate (and has the same "boundaries" and behavior), is that he doesn't think that the sliver plate is discontinuous at the place where NEIS draws a fault line, and that is consistent with what his plate shows. He has the sliver plate extending from northern Burma, bounded on the east by the Sagaing fault, a system of slip-strike faults in the Andaman Sea, and then the Sumatran fault zone (the central mountain range), and extends atleast past Sumatra.

Geodyssea: GEODynamics of S and SE Asia.

Crustal motion and block behavior in SE-Asia from GPS measurements. Gero W. Michel et al. Earth and Planetary Science Letters 187 (2001), 239-44

Historical Studies of Earthquakes in India, Roger Bilham , CIRES & Geological Sciences University of Colorado

Modelling of geological hazards and dynamics. (collection of reports on the region)

Crustal motion and block behavior in SE-Asia from GPS measurements. Gero W. Michel et al. EPSL Earth and Platetary Science Letters 187 (2001) 239-44. Probably from TEXSHARE database. Apparently available at multiple sites online.

Crustal motion in E- and SE-Asia from GPS measurements. Gero W. Michel et al. Earth Planets Space, 52, 713-20, 2000.

For information about catastrophic volcanos of the world, including Indonesia, New Zealand, and teh U.S., and catastrophic U.S. earthquakes, Dora the Explorer's Catastrophic Volcanos and Earthquakes Page

Basic geological information and terms

Canadian: Frequently asked questions about megathrust earthquakes

The Pacific Ring of Fire The "Ring of Fire" is an arc stretching from New Zealand, along the eastern edge of Asia, north across the Aleutian Islands of Alaska, and south along the coast of North and South America. It is composed over 75% of the world's active and dormant volcanoes.

Plate Tectonics web site The story of Plate Tectonics is a fascinating story of continents drifting majestically from place to place breaking apart, colliding, and grinding against each other; of terrestrial mountain ranges rising up like rumples in rugs being pushed together; of oceans opening and closing and undersea mountain chains girdling the planet like seams on a baseball; of violent earthquakes and fiery volcanoes. Plate Tectonics describes the intricate design of a complex, living planet in a state of dynamic flux.

NASA Plate Tectonics tutorial

Global Tectonics "Tectonics Links" Page

PBS Online: Savage Earth

Caltech UC Berkeley Seismological Laboratory: What is a Fault?


IRIS: Fault Motion

USGS Visual Glossary Faults

NSGS Earthquake Image Glossary

Teaching Resources in Structural Geology Definition of island arc and related terms

A Lesson in Plate Tectonics

Detailed introductory site on subduction zones with geothermal terminology Up a level is a general site on earth physics.

Margins User friendly online introduction to basic and advanced tectonics.

Plate motions and crustal deformation Charles DeMets Department of Geology and Geophysics, University of Wisconsin, Madison

"Oblique Subduction" When two plates converge at an oblique angle, they need one fault for subduction, and a separate system of faults for transverse/ slip-strike motion.

Plate Techtonic Modeling: Tools and Methods You will understand where they came up with the notion of a Burma microplate. You will understand it even better if you know advanced trigonometry.

Skywise Seismic FAQS page

Types of Volcanoes - from Volcano World

Volcano Types from Volcano Live

Plate boundaries: strike-slip components; terminology Includes such terms as transform fault, oblique convergence, and transpression. Plate boundaries: geologic activity; terminology (fault motion) Structural Geology Syllabus Links to specific units, with glossaries and explanations of terms.

The Structure and Dynamics of the Mantle Wedge A broad introduction to the thermodynamics of mantle wedges at subduction zones. He also points to another basic introductory article.

Effects of Thermal Structure and Mantle Wedge Flow gives a brief idea of what the notion of mantle wedge is talking about.

J. Stern's "Subduction Zones" paper (Reviews of Geophysics 40(2002); widely cited as a good introductory article on the thermodynamics of subduction zones.

Surface Expression of Susurface Structure - Glossary

The following terms were not found defined in any convenient online way (though many of them are discussed in the four above links provided by Daryl Krupa).

Back Arc Basin Basin located on the over-riding plate behind the volcanic island arc at a subduction zone

Fore Arc Region on the trench side of a volcanic arc

Island Arc Curved chain of volcanic Islands, many of which are located among the Circum-Pacific margins e.g. the Aleutian Islands and the Islands of Japan. Diagram as an imagemap !?!

Marginal Basin
Basin located on the over-riding plate behind the volcanic Island arc at a subduction zone. The basin may be split into Active Zone - A zone of extension which splits the active arc and trench complex from the remnant arc and continent. Inactive Zone - Region containing submarine ridges (the remnant arcs).

Remnant Arc
Inactive arc located behind the active volcanic arc and separated from it by an inter-arc basin

Slab Pull Force
The mechanism whereby the gravitational pull of a downgoing slab in a subduction zone exerts a lateral force on the plate attached to the slab

Emplacement of part of the oceanic crust onto the continental crust at a destructive plate margin.

Transform fault. A special variety of strike-slip fault that accommodates relative horizontal slip between other tectonic elements, such as oceanic crustal plates.

transcurrent fault - another word for strike-slip fault, or else one of two kinds of strike-slip fault, with transform fault being the other kind of strike-slip fault. A transcurrent fault may be a type of strike-slip fault that accomodates lateral motion along an oblique subduction fault.

Trancurrent motion - ?

dextral - one plate moves laterally to the right with respect to the other

sinistral - one plate moves laterally to the left with respect to the other

orthogonal - one plate moves straight into the other


Benioff Zone
Inclined zone of earthquake foci that dips beneath an Island Arc or active continental margin at approximately 45'. Benioff zone-A zone of earthquake epicenters beneath oceanic trenches that extends from near the surface to a depth of 700 km (430 mi). The foci of such earthquakes were plotted in the 1940s and 1950s by Hugo Benioff, for whom they are named. They are thought to indicate active subduction.

Interplate coupling
Interplate coupling is the ability of a fault between two plates to lock and accumulate stress. Strong interplate coupling means that the fault is locked and capable of accumulating stress, whereas weak coupling means that the fault is unlocked or only capable of accumulating low stress.

Locked fault
A locked fault is a fault that is not slipping because frictional resistance on the fault is greater than the shear stress across the fault (it is stuck). Such faults may store strain for extended periods that is eventually released in an earthquake when frictional resistance is overcome.

Normal stress. That stress component perpendicular to a given plane.

Tsunamis - safety information and the Pacific warning system

During the Dec 26 quake, not only did people along coastal areas of the Indian Ocean fail to realize that an earthquake could mean a tsunami and to recognize immediate warning signs of a tsunami, such as the water disappearing from the beach front or suddenly high water; but unbelievably, crowds of them, natives and tourists alike, went down to the waterfront to stare at the strange ocean behavior and speculate about it, while children ran out into the suddenly empty ocean to gather seashells and stranded fish!

Meanwhile, Pacific Ocean and U.S. earthquake and tsunami monitoring stations, particularly officials at the U.S. NOAA, alerted governments in India, Indonesia, and Thailand - and the officials who got the warnings didn't know how to make sense out of what they were told. ?!!!! An hour after the first tidal waves hit, the Thai government put out warnings on the INTERNET! I suppose they thought people who were still alive were swimming around looking at their laptops.

Many people there and in other regions of the world still do not understand that while tsunamis reached nearby coasts in minutes - and most people who knew to run for high ground or the upper floors of a reinforced building as soon as the quake stopped would have been spared - people across the Indian Ocean wehre the greatest damage and numbers of deaths occurred had five hours to escape danger if they had been warned.

I understand these governments also had trouble understanding the sense of spending two million dollars on a warning system. Now they will spend many times that on the damage. Someone tell me that all of this is because India, Thailand and Indonesia are governed by impoverished fishermen?

Since I learned all about tsunamis, the obvious warning signs, and what to do if they occur, in a fictional story in a children's magazine story when I was ten, and the story concerned an illiterate boy in a poor fishing village whose uncle had told him about a previous tidal wave, I find such extreme and total ignorance among such a wide range of people people in such a high risk area beyond belief.

Below is basic safety information about tsunamis as well as information about the warning system in place in the coastal areas of the northern Pacific ocean. People'd best read up. Especially those whose idea of vacation includes a tropical island.

West Coast & Alaska Tsunami Warning Center

The Book: Tsunami Even a child can understand this resource.

Tsunamis: Office of Emergency Preparedness brochure

Tsunami safety rules

Tsunami Hazard Mitigation

Pacific Tsunami Museum FAQS

This page by Dora Smith
Email me at or

This page hosted by
Get your own Free Home Page