Soil on a Cable: How InSight Tackled Seismic Noise on Mars

Dropping soil beside a sensitive scientific instrument sounds like an unlikely way to improve its measurements. In March 2021, however, InSight's operations team deliberately used the lander's scoop to let soil trickle down its seismometer's protective dome toward a connecting cable. The soil was not intended to strengthen earthquakes inside Mars. It offered a possible way to weaken disturbances introduced near the instrument itself. Understanding a distant planet could depend on managing the small movements of a connection lying on its surface. [2]

That intervention illustrates a distinction between sensitivity and identification. An instrument may register extremely slight motion without every registered motion being the event its designers hope to study. SEIS, the Seismic Experiment for Interior Structure, was supplied by the French space agency CNES and its partners and placed on the ground by InSight's robotic arm. Its purpose was to investigate waves traveling through Mars. On the surface, though, it also belonged to a local mechanical environment. This account examines that engineering problem through mission reports published in 2019 and 2021, within a knowledge cutoff of 13 October 2026. It does not present their historical event counts as a final mission inventory. [1][2]

Motion is not yet a source identification

A change in a seismic record cannot identify its origin by itself. A marsquake produces waves traveling through the ground. A moving robotic arm can also introduce movement into the instrument's surroundings, and wind can generate interfering vibrations. Those disturbances need not be imaginary or evidence of a defective sensor. They can be real physical movements that answer a different question from the one being asked about the planet's interior.

The October 2019 mission report made this distinction particularly clear. Of more than a hundred events detected by that date, approximately twenty-one were strongly considered to be quakes. Other origins had not yet been ruled out for the remainder. The same report described a camera survey performed by the robotic arm on 6 March 2019: individual arm movements registered conspicuously in SEIS. Producing a list of events and confidently assigning each event a geological cause were separate accomplishments. The early figures describe the team's assessment at that time, not a limit on everything the instrument would eventually detect. [1]

Operational information therefore matters alongside the seismic trace. If an unusual feature appears while the arm is moving, a mechanical explanation deserves investigation. This is not a rule that every coincident signal should be discarded automatically. It is a requirement to ask what else was happening when the measurement was made. Separating an instrument's record from its operating circumstances can make an uncertain event look more unambiguous than the evidence warrants.

A quieter interval can change what becomes detectable

Wind changed those circumstances over the course of a day. The 2019 report identified twilight as one of the useful periods for finding quakes. Daytime warming produced more wind interference than nighttime conditions. A calmer interval could allow weak seismic activity to stand out more clearly. That is a change in the conditions of observation; it need not mean that Mars was producing more earthquakes during the same interval. [1]

The April 2021 account extended the problem to seasons. During the preceding northern winter, wind interference had prevented quake detection. Clear events became available again as winds calmed. It would be incorrect to turn the absence of detections into a demonstration that the planet had stopped shaking. As an explanatory analogy, someone speaking in a crowded room may become inaudible without becoming silent. The analogy is not a quantitative model of Mars, but it captures why counts from different observing conditions require care. [2]

Two clear events in March 2021 were associated with Cerberus Fossae, a region already linked to earlier strong quakes. Together, those records were useful for studying the interior. Their value was not simply that someone had added two entries to a catalog: the signals were especially clear. An observing program can accumulate many records while only some are suitable for a particular analysis. Treating the size of a catalog as a complete measure of scientific quality conceals that difference. [2]

The shield did not enclose the whole measurement chain

SEIS already had a dome-shaped Wind and Thermal Shield. Its presence nevertheless did not eliminate wind-related vibration sufficiently to reveal every quake. Protection against a disturbance is a matter of degree, not a guarantee that the disturbance has vanished. The instrument was also connected to the lander by a cable. Protecting the sensor's immediate surroundings did not automatically give the exposed connection the same protection. The measurement arrangement extended beyond the object underneath the dome. [2]

The 2021 report described local temperatures ranging approximately from minus one hundred degrees Celsius at night to zero during the day. Expansion and contraction of the cable under those changing conditions might contribute to sudden pops and spikes in the record. That wording matters. The report presented a possible mechanism, not a controlled demonstration that every anomalous spike had the same cable-related origin. It nevertheless supplied a concrete target for an engineering intervention: partly insulating the connection could reduce an unwanted movement before it reached the recorded signal. [2]

On 14 March 2021, the mission's 816th Martian day, the scoop began delivering soil by letting it fall over the shield and trickle toward the cable. The approach allowed material to reach close to the shield without interfering with its seal against the ground. Adding insulation could not be treated as an isolated task: the team also had to preserve the protective arrangement already in place. The source documents the beginning and purpose of the operation. It should not be read as proof that the entire cable had been buried successfully or that all associated noise subsequently disappeared. [2]

What a listener hears is another representation

Public audio examples from SEIS introduce a further distinction. The earthquake recordings released in 2019 were sped up and slightly processed to make them audible to people. They were not ordinary microphone recordings played at their original speed. Sonification carries aspects of a measurement into a form that human hearing can explore. The resulting listening experience and the original physical measurement are related, but they are not interchangeable. An arresting sound cannot independently establish the geological origin of the event that produced its underlying data. [1]

The same mission account discussed smaller mechanical sounds associated with parts of the instrument expanding and contracting. Cooling was proposed as a factor in some evening sounds. An unusual whistling effect was thought possibly to involve electronic interference. These explanations retained uncertainty about the mechanisms. Rewriting them as settled diagnoses would remove an important property of the evidence: different possible sources were being considered, not all instantly resolved. [1]

Recognizing those limits does not make the observations less useful. It shows how their usefulness was established. The 2019 report describes learning to recognize environmental and operational disturbances. The 2021 report describes a physical attempt to reduce one suspected route by which a disturbance entered the measurement chain. Classification after recording and partial insulation before recording address different stages of the same problem. Neither makes the other unnecessary, and neither promises complete recovery of every weak event.

Read together, these documented episodes support a practical conclusion about planetary instrumentation. To investigate motion arriving from inside Mars, the team also had to understand motion generated around its equipment on the surface. The scientifically important boundary was not simply between instrument and planet. It ran through the arm, the shield, the cable, the weather, and the interpretation of their combined record. Soil trickling toward a cable was a modest-looking intervention, but it addressed that boundary directly.

Sources

[1] NASA/JPL. NASA's InSight 'Hears' Peculiar Sounds on Mars. 1 October 2019. https://www.nasa.gov/solar-system/nasas-insight-hears-peculiar-sounds-on-mars/

[2] NASA/JPL. NASA's InSight Detects Two Sizable Quakes on Mars. 1 April 2021. https://www.nasa.gov/centers-and-facilities/jpl/nasas-insight-detects-two-sizable-quakes-on-mars/