logo seis

SEIS on Mars: First Results

SEIS on Mars: a review following publication of the first results

SEIS, the first broadband seismometer on Mars, has been patiently listening to the Red Planet’s activities day in, day out for almost one year now. Following the InSight lander’s touchdown on Elysium Planitia on 26 November 2018 and the seismometer’s deployment on the Martian surface between December 2018 and February 2019, the instrument has detected over 300 events and provided planetary scientists with numerous results. Forty-four years after the Viking probes’ first heroic attempt, SEIS has just allowed a new planetary discipline—Martian seismology—to finally see the light of day.

The Martian environment’s noises, or why SEIS works mainly at night

In its slow and patient pursuit of marsquakes, SEIS faces a formidable hurdle. At any point in time and often completely randomly, a wide variety of noises can interfere with seismic signals and obscure the tiny surface tremors.

Some of these disturbances are caused by the lander, when operators move the robotic arm to take pictures for example, or when electrical currents flow through the solar arrays. Others are generated by the instrument itself, within its delicate electronic components, or when mechanical structures are deformed under heat stress. The main source of disturbance, however, remains the Martian environment.

Working in conjunction with InSight’s sophisticated weather station (APSS), which operates continuously, SEIS has shown that each Martian day can be broken down into three periods of varying degrees of noise, making each period more or less suitable for detecting quakes.

Artist's view summarizing the different sources of external noise affecting InSight’s SEIS seismometer (© IPGP/Nicolas Sarter).Artist’s view of InSight landing site during day and night. Between 5 pm and midnight, most interfering signals disappear and SEIS then benefits from exceptional conditions for listening to the seismic activity of Mars (© IPGP/Nicolas Sarter).
From sunrise at 7 am (local solar time) until about 4 pm, the Martian atmosphere’s activity becomes more and more turbulent. For signals with a period of less than one second, these disturbances reach the same level as those of Earth, which is a very noisy planet (for ten-second periods, however, daytime noise remains ten times lower than on Earth). At certain frequencies, our world causes a real hubbub due to constant human activity or the pounding of ocean waves. The frequency saturation affecting SEIS is mainly caused by the Martian atmosphere’s convective currents, which are the result of the huge temperature differences between the planet’s surface and its atmosphere.

Fortunately, the noise level suddenly drops drastically from around 5 to 6 pm and the Red Planet remains quiet until about midnight. This short respite is clearly the best time for detecting tremors. During this period, the noise level is some 500 to 1,000 times lower than that on Earth (SEIS has recorded levels even lower than those recorded on the Moon by the Apollo seismometers!). SEIS is then free to listen to what is happening in frequency bandwidths that are totally inaccessible to terrestrial seismometers because they are deafened by interfering noise.

Summary spectrogram indicating the origin of quakes detected from sol 72 to sol 400 (© MQS/NASA InSight/SEIS Team).Summary spectrogram indicating the origin of quakes detected from sol 72 to sol 400 (excluding solar conjunction), depending on the time of day on Mars. Between around 5 pm and midnight, a hush falls over the planet, offering SEIS an excellent window of opportunity for detecting tremors. During this period, the noise level drops by a factor of 100, making it 500 to 1,000 times lower than that on Earth. This enables SEIS to explore frequency bandwidths unusable on Earth because of the high ambient noise (credit: © MQS/NASA InSight/SEIS Team).

Finally, from midnight to 7 am, disturbance begins gradually increasing again until sunrise, which marks the beginning of another noisy day. Such is the life of a seismometer on Mars, punctuated by periods of calm conducive to observations, and moments of noisy uproar that interfere with the signals.

Although this situation may appear to be ideal, it is unlikely to be permanent as there are seasonal variations in the Martian environment’s energy level. With the arrival of summer and its famous dust storms, it is likely that SEIS will temporarily lose its ability to distinguish the subtlest tremors at night.

Characterizing near-surface subsoil

While the ultimate goal of SEIS is to unlock the secrets of the deep internal structure of Mars, one of its earliest, more modest but nonetheless fundamental occupations was to determine the properties of the near-surface subsoil.

The subsurface of dusty Elysium Planitia—InSight’s landing site—consists of three distinct layers, some of whose properties (such as rigidity and depth) have now been determined using three different techniques.

Artist's view showing the probing of the landing site subsurface by three different techniques (© IPGP/Nicolas Sarter).Artist's view showing the probing of the landing site subsurface by three different techniques : resonance of the seismometer's levelling cradle, hammering of HP3 penetrator and minute uplifting of the ground by dust-devils (© IPGP/Nicolas Sarter).

When it was deployed on the ground by InSight's robotic arm during sol 22 (19 December 2018), the three conical legs of the seismometer's levelling cradle settled into an indurated layer a few centimetres thick known as the duricrust. By analysing the way the legs resonate when there is a tremor, seismologists have been able to work out the elasticity of this superficial armour.

Being a more cohesive layer than those underneath, the duricrust appears to be hindering downward progress of the HP3 penetrator, designed to benefit from the friction offered by looser soils. The thousands of jolts generated in an attempt to bury deeper have, however, made it possible to study the regolith, the superficial layer of material crushed by countless impacts over billions of years and also found on the Moon.

During its hammering sessions, HP3 behaves like an active seismic source by making the subsoil vibrate. Scientists have therefore been able to estimate the propagation speed of P-waves in the first metre below the Martian surface, which in turn has enabled them to deduce certain physical characteristics specific to the regolith.

Finally, the subsoil has been sounded on a scale of some 10 (vertically) to 100 (horizontally) metres using the deformations induced by the movement of dust devils. The latter, which appear to enjoy wandering nonchalantly over Elysium Planitia, cause sharp pressure drops visible in the data from the ultra-sensitive pressure sensor of InSight's weather station, the APSS. The minute uplifting of the ground (which can be compared to a kind of suction) that they create on their erratic passage is also very clearly recorded by SEIS.

However, Elysium's dust devils—which SEIS considers to be micro-seismic sources—have kept a surprise in store for the mission's scientists. For more than a year now, although InSight's weather station and SEIS seismometer have made thousands of indirect observations, none of these formations have ever been photographed by the lander, particularly by the ICC camera located below the lander deck.

To be seen, a dust devil must raise a sufficient amount of the extremely fine ochre-coloured dust that covers the Martian surface. There is plenty of dust at the landing site, as shown by the copper-coloured deposit on the lander deck and solar arrays. So why have none of the Martian atmosphere’s whirling dervishes that have whisked past the lander on Elysium Planitia ever been seen?

To unravel this mystery, scientists are counting in particular on the Fluxgate magnetometer in the weather station suite. This magnetometer can actually detect the triboelectric effect (i.e. the production of static electricity) that occurs due to the friction caused when dust particles rub against each other as they are sucked up into the vortex of a dust devil.

Artist's view of the landing site showing the layered subsurface structure : from top to bottom: duricrust, regolith and fractured bedrock (© IPGP/Nicolas Sarter).Artist's view of the landing site showing his layered subsurface structure : from top to bottom: duricrust, regolith and fractured bedrock (© IPGP/Nicolas Sarter).

The first marsquakes: the Red Planet is seismically active

Since its commissioning in February 2019, SEIS has detected around 300 events, including around ten quakes of magnitude 3 to 4. This particularly high number of tremors is comparable to that observed in areas on Earth not affected by plate tectonics (absent on Mars), or associated with hot spots. Current data suggest that the Red Planet’s seismic activity is two to three times lower than Earth’s intra-plate seismic activity, but ten to 20 times higher than that of the Moon.

Classifying marsquakes

To be studied, marsquakes have been classified into two main populations: “low-frequency” quakes (of an energy mostly below 1 Hz, i.e. one vibration per second), and “high-frequency” quakes (of an energy above 1 Hz).

Martian seismologists believe that this division reflects different seismic phenomena in terms of source, depth and trajectories followed by the seismic waves. Low-frequency (LF) quakes, which are quite rare and can reach magnitude 4, are thought to come from deep down—at least 50 km below the surface—and quite far away from the lander. The depth of the origins of such quakes would explain the absence of surface waves, which have so far never been clearly observed by InSight.

High-frequency (HF) waves are far more numerous but of a lower magnitude. They are thought to be more superficial. Their weak signals lack information, thus preventing any possibility of determining their position, however partially, on the map of Mars. Some HF quakes are even so subtle that they can only be detected through the amplification offered by a strange specific resonance located at 2.4 Hz.

Resonance behaving like a seismic amplifier

Very early in the mission, seismologists identified on the spectrograms a strange continuous excitation at a frequency of 2.4 Hz.

Present throughout the day, this resonance does not appear to be related to meteorological activity, but is stimulated by more than half of the HF quakes. Unlike other phenomena centred around 4 Hz and 6.5 Hz, it does not seem to correspond to the vibration of lander structures such as the solar arrays under the effect of wind, or to certain operational activities such as movements of the robotic arm or the uplinking of radio data to the relay satellites in Martian orbit.

For the time being, the origin of this 2.4 Hz signal (or “mode” in seismologists’ language) is unknown. Scientists believe that it is related to the particular structure of the first few kilometres below the surface of the landing site, without being able to elaborate further at this point in time. The one thing they are sure about is that this resonance acts as a natural seismic amplifier by intensifying or revealing some very weak seismic tremors that would otherwise be inaudible. While Mars has a habit of putting obstacles in the way of explorers trying to decipher its mysteries, it also sometimes makes things a little easier, often in totally unexpected ways.

An ever-increasing number of quakes

A second intriguing aspect of HF events is that their number has been steadily increasing since the beginning of the observation campaign. While the seismometer struggled to detect such tremors at the beginning of its mission, it now records an average of two a day, a rate that is slowly but surely increasing without seismologists really understanding why.

Summary spectrogram showing the increase in the number of events (mainly high frequency) from the beginning of the mission to now (© MQS/NASA InSight/SEIS Team).Summary spectrogram showing the increase in the number of events (mainly high frequency) from the beginning of the mission to now (© MQS/NASA InSight/SEIS Team).

Analyses have shown that this increase is not a measurement artefact linked, for example, to a change in the SEIS instrument’s behaviour or parameter settings. It is therefore quite possible that this increase is natural. It could, for example, be due to a periodic phenomenon involving seasonal warming, or the position of Mars along its orbit.

A question of quality

The quakes are not only classified according to type (HF or LF), but are also rated (A, B, C or D) according to their quality, defined as a set of parameters to which mission seismologists pay close attention. One of the key parameters is the signal to noise ratio (SNR), which must be as high as possible. However, the quality also depends on being able to reliably determine the quake’s distance from the seismic station, its azimuth (i.e. its direction with regard to the north) and its magnitude. There is no procedure like this on Earth: it is specific to Mars, as seismologists are becoming increasingly aware that the tools developed for Earth quickly become limited when dealing with the Red Planet.

To be classified as “A”, a given seismic event must have clear, identifiable phases (P- and S-waves) and high-quality polarization information that can be used to specify its position. When the polarization cannot be determined but the phases are clear, the event is assigned to category “B”. Its distance can still be determined, but its azimuth can no longer be established. When a signal is clearly identified but no phase information can be extracted, it is assigned to category “C”. Finally, very weak or ambiguous events are classified as “D”.

Quakes in quality category “A” are the most useful to seismologists, but they remain very rare. Whenever one is detected, its arrival is greeted with great enthusiasm by SEIS researchers. So far, InSight has identified only two of them, during sols 173 (23 May 2019) and 235 (26 July 2019) respectively.

The first marsquake, on sol 128

The quake on sol 128, detected on 7 April 2019, was the first ever recorded on the Red Planet. In a way, this date is a milestone marking the birth of Martian seismology. Surprisingly, this first Martian tremor was observed almost 130 years to the day after Ernst von Rebeur-Paschwitz recorded the first terrestrial seismogram in Potsdam. Early in the evening of 17 April 1889, the horizontal sensor of a seismometer installed in Germany reacted to the arrival of an earthquake of magnitude 5.8 that had shaken Japan an hour earlier.

The HF quake of sol 128 was very weak, and was attributed a magnitude of only 2.1. The signals did not enable the seismologists to distinguish between the first wavefront of the P-waves (which arrive at the seismometer first) and the later S-waves (which travel slower through material and often come second).

To detect and analyse this first marsquake (and the vast majority of the following ones), the mission seismologists did not work on waveforms, which are representations that express the amplitude of tremors as a function of time, somewhat reminiscent of an electrocardiogram. The signals received being very subtle, they opted instead to study the spectral content via coloured spectrograms (which show the energy of a quake from start to finish as a function of frequencies).

Spectrograms and waveforms of the first marsquake ever recorded. It was detected during sol 128 (7 April 2019) by both the SEIS seismometer’s short-period (SP) sensors (at the top) and its very broadband (VBB) sensors (at the bottom). The signal has been processed (sonification) to make it audible to the human ear. From left to right, you can hear the wind sweeping across the landing site, the quake itself and finally movement of the robotic arm (NASA/JPL-Caltech/CNES/IPGP/Imperial College London).

Although the origin of the sol 128 quake remains unknown for the moment, the focus from which the waves emanated is thought to lie over 5 km deep. The quake is believed to have originated in the crust somewhere along a circle with a radius of 530 km from the InSight lander, but the direction could not be determined.

The sol 173 quake: initial identification of P- and S-waves and first position determination

The second major marsquake recorded by InSight was on sol 173. It is one of the most interesting tremors observed on Mars so far. Unlike the event of sol 128, the sol 173 tremor was a low-frequency quake. With a magnitude of 3.6, it also released much more energy than that of sol 128. Even more noteworthy, it was the first quake to offer seismologists the possibility of unambiguously identifying the wavefronts (i.e. pinpointing the arrival times) of P- and S-waves in the signals.

Containing much more information than the quake of sol 128, the sol 173 recording enabled seismologists to pin down the epicentre. This measurement is particularly difficult when there is only one seismic station on a planet (as is the case with InSight), especially when the speed at which seismic waves travel through materials is not yet precisely known. As if that were not enough, the astonishing absence of surface waves (which, as their name suggests, propagate along the surface) that the mission seismologists were relying on to facilitate position determination, adds another challenge to the process.

The frustrating absence of surface waves

To overcome the limitations inherent to the use of a single seismic station on Mars, mission seismologists placed a good deal of hope in the exploitation of surface waves. To triangulate the origin of a quake, at least three instruments are normally required. However, to determine the origin of marsquakes, InSight's seismologists had developed a method of listening to the repeated passage of the surface wavefront emitted by a quake powerful enough to generate vibrations capable of circling the planet several times.

Up to now, no surface waves have been detected. Why not? Mission seismologists have put forward several hypotheses: one suggests that the quakes observed so far could be too deep, another that the surface waves could be affected by the cratered, crushed surface of Mars. If the second explanation is correct, it may also be more complicated than first thought to detect meteorite impacts, which first and foremost generate surface waves when the meteorite hits the ground, excavating a new crater.

Thanks to sophisticated mathematical processing, planetary scientists have nonetheless managed to get the signals from the sol 173 quake to talk, and to locate the tremor’s point of origin on a map of Mars. All the information gathered so far indicates that on 23 May 2019, the ground began to shake 1,600 km from the InSight lander, and that the resulting waves arrived from a region to the east called Cerberus Fossae, which seismologists have been watching for a long time.

Cerberus Fossae: the first active seismic zone ever found on Mars

Cerberus Fossae is a huge fault system located east of Elysium Planitia, InSight’s landing site, which was probably generated during the formation of Elysium Mons, the second largest volcanic complex on Mars after the Tharsis bulge and its giant volcano Olympus Mons. The analysis of images provided by satellites orbiting Mars suggests that the Cerberus Fossae faults are still active, and could continue to generate quakes today. A former site of volcanic, fluvial and wind activity (the most recent evidence of which dates back only 10 to 2 million years), the Cerberus fracture region has apparently accumulated numerous stresses in its history, some of which have not yet been released.

Artist’s view of the Cerberus Fossae region (© IPGP/Nicolas Sarter).Artist’s view of the Cerberus Fossae region. Studies of data on this region, provided mainly by Mars orbiters, have revealed numerous rock falls along the near-vertical fault walls. These could be caused by seismic tremors (© IPGP/Nicolas Sarter).
The event of sol 235: first detection of an aftershock

It may or may not be a coincidence, but the third major quake detected by SEIS during sol 235 (26 July 2019) is also located in the Cerberus area. The epicentres of the quakes on sols 173 and 235 are actually only about 450 kilometres apart.

Classified as an A-grade LF quake, the event on sol 235 stands out due to the fact that it was accompanied by an aftershock, the first ever detected on Mars. Thirty-five minutes after the main quake, SEIS recorded a secondary wave train coming from the same direction and following exactly the same path as the main quake.

The first aftershock ever detected on Mars occurred during sol 235 (26 July 2019) (© IPGP/NASA InSight/SEIS team).The first aftershock ever detected on Mars occurred during sol 235 (26 July 2019) (© IPGP/NASA InSight/SEIS team).

The first seismic activity map

In addition to the events of sols 173 and 235 originating in the Cerberus Fossae region, mission seismologists have approximately located another LF quake on the map of Mars. The epicentre of this one, detected during sol 183, appears to be located near an enigmatic elliptical structure known as Orcus Patera and thought to be either a meteorite impact crater or an ancient volcano.

That is about as far as quake position determination has been able to go up to now. It has only been possible to determine the distance of the few hundred other events identified by SEIS (mostly HF events) from InSight, and even that has sometimes proved difficult. Since their direction in relation to the north could not be calculated, the seismologists have placed these events along circles centred on the lander. The circumference of each circle depends on the tremor’s distance from the seismic station, it being possible that the furthest ones could be 3,000 or even 4,000 kilometres away. While such mapping can be frustrating, it nevertheless allows us to begin to link certain seismic zones with possible tectonic structures on the surface of the Red Planet.

First global seismic activity map of Mars (© NASA/InSight SEIS team).First global seismic activity map of Mars. The position of the vast majority of marsquakes cannot yet be precisely determined. Only their distance from the seismic station can be estimated; their orientation relative to the cardinal points remains unknown, resulting in an approximate distribution along circles (© IPGP/NASA InSight/SEIS Team).
Characteristics of the upper Martian crust

By studying the largest quakes detected to date and presented above, the mission seismologists were able to determine the attenuation and scattering (or diffusion) values of seismic waves as they passed through the Martian crust, and to identify the presence of a major discontinuity.

Diffusion and attenuation of seismic waves

Quite early on, as soon as the first signals were received, SEIS revealed that the Martian crust reverberated seismic waves quite intensely, a behaviour that is unfortunately not advantageous for seismologists. Crushed by an incredible number of impacts over billions of years, the crust of the Red Planet reflects and disperses waves in all directions. The crust’s fractures and fragments of all sizes behave like mirrors for seismic vibrations, with the unwanted result of generating considerable interference affecting the signals received by the seismometer.

In a homogeneous medium such as the crystalline bedrock that forms the Earth’s mountain ranges, seismic waves propagate in quite a disciplined way. When an earthquake occurs, the signals collected by a seismometer are very clear, and the wavefronts arrive in the right order at the monitoring stations, usually within a very short time. For a seismometer, the whole process is as crystal clear as the sound of a bell announcing midday in a small countryside village.

On the other hand, in an environment conducive to the physical phenomenon of widespread scattering, each quake turns into a veritable cacophony. Some wave trains are considerably delayed by all the reverberations, and the signal relating to the tremors is diluted over time, becoming much more difficult to decipher. It may therefore become impossible to identify the arrival time of P- and S-waves via their polarization. Just like the Moon before it, the planet Mars is thus posing seismologists many challenges.

SEIS data have furthermore shown that attenuation—the slow but inexorable loss of energy that seismic waves experience along their propagation path—is three to four times larger than that of the Moon, which could be explained by a higher crustal hydration level.

Stratification of the upper Martian crust

The largest marsquakes were also used to carry out an initial analysis of the crustal structure. Based on the so-called “receiver function” method, which identifies the conversion of seismic waves when they encounter discontinuities (e.g. conversion of P-waves to S-waves), it is possible to perform a seismic survey.

This technique revealed an 8- to 11-km-thick stratum of highly damaged or fractured volcanic material in the Martian crust. It could slow down some seismic waves by about 50%. It is thought that deeper down, there is a more homogeneous and coherent layer that could extend down to the Mohorovičić discontinuity (or Moho for short), a boundary marking the beginning of the mantle. However, the mantle has not yet been detected.

Heading into the heart of Mars

Although the Martian crust is beginning to reveal its secrets, InSight's ultimate goal is to unveil the planet’s deep internal structure. However, the relatively large number of events detected so far is no guarantee that seismologists can shed light on the inner workings of Mars, and in particular its mantle and core.

The vast majority of events that shake the Martian surface are of very low intensity, and significant quakes of a magnitude greater than 4 are clearly rarer than expected. The seismic activity of Mars is thus quite different from that of Earth. It could therefore take longer and be more difficult than expected to determine the internal structure of the Red Planet.

The initial results provided by the SEIS seismometer during its first year of operations on Mars are, however, very encouraging. The data, which are made public as they become available, are being carefully studied by many international teams. Although a very young discipline, Martian seismology is already stimulating new lines of research, and laboratory projects are abounding in many different areas.

Planetary seismologists have been hard at work analysing the huge data set provided by SEIS, but they are also waiting for the first huge marsquake that will approach or exceed magnitude 4.5, and will finally reveal the closely-guarded secrets of the Martian interior.

Nature Research Papers:

Useful links:

Philippe Labrot, 25 february 2020 (last update on 27 april 2020 to modify the research papers links and to fix an error regarding seismic attenuation).


logo IPGP bottom

Realized by IPGP (P. Labrot)



 NASA JPL - Oxford University - Imperial College London - CNES - ISAE - MPS - IPGP - ETH