On October 19, 2016, exactly at the time when its mother ship was entering orbit around Mars, the 577-kilogram Schiaparelli lander attempted a six-minute descent onto the Martian surface, however something went wrong around the time when the main parachute of the lander was jettisoned and soft-landing engines fired for the first three or four seconds...
The Schiaparelli landing sequence. Credit: ESA
During its seven-month cruise between the Earth and Mars, the EDM Schiaparelli lander was in hibernation mode to save the very limited power of its small batteries. (The battery capacity is the main limiting factor for the life span of the lander on the surface of Mars).
Schiaparelli separated from the TGO orbiter on October 16, 2016, at 14:42 GMT (10:42 a.m. EDT, 16:42 CEST, 17:42 Moscow Time, three days before reaching the surface of Mars. The timing of the operation was chosen as a compromise between the acceptable accuracy of the Schiaparelli's landing and the required time for its "mother ship" -- TGO -- to maneuver away from a collision course with the planet.
TGO and Schiaparelli were mechanically linked via the main separation assembly, MSA, which was attached to the orbiter with 27 screws. The MSA held onto Schiaparelli with three separation mechanisms made of compressed and angled springs that were attached with non-explosive actuators, NEA. When the NEA's were released, Schiaparelli was gently pushed away from TGO, at the same time was imparted with a rotation that helped to stabilize its atmospheric entry.
Before the lander reached Martian atmosphere on October 19, the European Mars Express spacecraft, which has been orbiting the Red Planet since 2003, was able to record information from Schiaparelli beginning at 13:22 GMT (9:22 a.m. EDT). Around 1.5 hours, before hitting the Martian atmosphere, the lander was programmed to wake up from hibernation and begin its transmissions five minutes later and its first signals were to reach Earth around 13:36 GMT (9:36 a.m. EDT).
The TGO orbiter also reached a position to listen to Schiaparelli beginning at 14:20 GMT (10:20 a.m. EDT).
Just 22 minutes later, the lander hit the upper reaches of the Martian atmosphere at 14:42 GMT at an altitude of 122.5 kilometers above the planet and a speed of 21,000 kilometers per hour (5.8 kilometers per second).
In just about three minutes, the lander had to decelerate from Mach 35 to Mach 5 thanks to aerodynamic braking with a disk-shaped heat shield, which was expected to heat up to 1,500-1,850 degrees Celsius.
During the descent, an autonomous computer sequence had to control ejecting the front and back aeroshells of the lander, operating the descent sensors, deploying the braking parachute and activating the three groups of hydrazine thrusters to control the touchdown speed.
At 14:45 GMT (10:45 a.m. EDT), at an altitude of around 11 kilometers and a speed of 1,650 kilometers per hour (Mach 2), a 12-meter supersonic parachute made out of nylon fabric and held by Kevlar lines was to be deployed with the help of mortar to reduce speed to below supersonic.
The parachute's so-called disk-gap-band canopy was to unfurl in less than a second to a length of 27 meters, and, 40 seconds later, allowing for oscillations to die down, the front shield of the aeroshell was to be jettisoned at an altitude of seven kilometers and a speed of 320 kilometers per hour.
At an altitude of around seven kilometers, Schiaparelli was to activate its Doppler radar altimeter and velocimeter to locate its position with respect to the Martian surface by measuring its distance and speed relative to the surface.
Around a minute after the front shield was jettisoned and the module dropped to an altitude of around three kilometers, a down-looking engineering Descent Camera, DeCa, was to begin shooting a series of 15 black and white images, looking over the edge of the lander at the approaching and spinning surface below. Transmitted to Earth some 24 hours later, the resulting pictures could help pinpointing the exact landing trajectory and touchdown location.
Photos were to be taken at 1.5-second intervals and could continue capturing the descent until an altitude of around 1.5 kilometers or possibly even lower. At the beginning, the camera could see an area as wide as 17 square kilometers, which would zoom to as little as 4.6 square kilometers at the end of the photo-session. However there was a chance, that wind storms in the Martian atmosphere typical for this time period, could veil surface details with clouds of dust.
As soon as the sensors detected that the spacecraft was 1.3 kilometers (or 1.2 according to some sources) from the surface around 14:47 GMT (10:47 a.m. EDT), the back shield of the lander would be jettisoned along with the parachute, just two minutes after its deployment, as the spacecraft zoomed toward the surface at a speed of between 250 and 270 kilometers per hour.
Immediately, three clusters of hydrazine engines with three engines each had to fire in pulse mode to reduce the descent speed from 240 kilometers per hour to just 15 kilometers per hour around two meters from the surface. At that moment, the engines were to cut off, leaving the dish-shaped lander in free fall.
The final shock of the touchdown at around 11 kilometers per hour would be cushioned by a crushable structure built into the module. The entire descent was to be completed for Schiaparelli in around six minutes over a distance of approximately 700 kilometers.
The nominal touchdown was scheduled to take place at 14:48:11 GMT (10:48 a.m. EDT, 16:48:11 CEST) on October 19, 2016.
Schiaparelli’s historic descent was being recorded on Earth in near-real time by scientists using the Giant Metrewave Radio Telescope, GMRT, located near Pune, India, and operated by the National Center for Radio Astrophysics, part of the Tata Institute of Fundamental Research. GMRT comprises an array of 30 radio telescopes, each with a dish diameter of 45 meters, and it is one of the world’s largest interferometric arrays.
This activity promised to provide extremely important confirmation of the module’s descent and landing, and signified a major area of international cooperation between ESA, NASA and India.
NASA's Jet Propulsion Laboratory in Pasadena, California, installed experimental receiving equipment at the GMRT telescope to help it see not just the astrophysical events but also track spacecraft like Schiaparelli.
From the outset, the signal acquired by GMRT was expected to be very weak because Schiaparelli was never designed to transmit all the way to Earth. ESA characterized GMRT tracking of Schiaparelli is a 'nice-to-have' experiment to allow to watch the descent in real time. During practice testing during Schiaparelli separation from the TGO orbiter, ESA experts already saw the weak trace of the signal from the lander at GMRT. At the time, only 18 and, later, 16 antennas were available. By the time of the landing ESA was to have 28 of the antennas working, but the orientation of the probe and some other factors could still prevent the reception of the signal.
The primary vehicle monitoring the descent was ESA’s Mars Express probe, which has been orbiting the Red Planet since 2003. It recorded signals from the lander with the help of a UHF radio device onboard called Melacom, for Mars Express Lander Communication System. It was previously used to communicate with Spirit, Opportunity and Curiosity rovers, as well as with the Phoenix lander. Mars Express also had a special 'open-loop' recording mode that could track a signal tone from a lander as it arrives at Mars. This has been used successfully to monitor the entry, descent and landing of the NASA's Phoenix lander and the Curiosity rover.
Although Melacom was not designed to decode data in the open-loop mode, the Doppler shift that it did receive was very sensitive to all minor changes in the speed of the lander. However Mars Express had to point the 45-centimeter Melacom antennas at Mars to record the signal and, as a result, it could not relay the received signal in real time to Earth. Instead, as soon as Schiaparelli landed, Mars Express re-oriented itself toward Earth and replayed the recording, which was received by ESA's Cebreros ground station and transferred to ESOC, where it was processed and made available to ground controllers around 1.5 hours after Schiaparelli touchdown. Unfortunately, its data was not conclusive.
The TGO orbiter was also above the horizon for the Schiaparelli until 14:56 GMT (10:56 a.m. EDT) or eight minutes after the touchdown. Another seven minutes later, (at 15:03 GMT), Schiaparelli was programmed to stop its first transmissions from the surface and go into hibernation for the first time on the surface of Mars. Five minutes later, Mars Express discontinued its recording of information from the surface and began transmission of captured data back to Earth half an hour later. The transmission session lasted around 20 minutes, concluding around 16:32 GMT Earth's time (12:32 p.m. EDT).
TGO was fitted with an Electra UHF radio, which unlike GMRT or Mars Express, could record actual data that was being streamed to it on the Schiaparelli signal, giving precise details and measurements of the descent progress all the way down.
This would be invaluable for engineers in order to piece together the exact performance of the Schiaparelli journey to the surface. However, TGO was busy at the time with its own orbit entry, and as a result, this data could not be downlinked to Earth until sometime later and then processed in the early hours of the morning after landing, on October 20.
NASA's Mars Reconnaissance Orbiter, MRO, was expected to fly over the Schiaparelli landing site for the first time just under two hours after landing. At that point, Schiaparelli was programmed to wake up ready to talk to MRO's Electra UHF radio and to establish an eight-minute-long, two-way communication link to get all the latest status information and data from the lander. This data would then be sent to Earth around 1.5 hours later and could give mission control the final word that Schiaparelli is safely on the ground and functioning on Mars.
Members of the ExoMars mission team look at data monitors at the European mission control center in Darmstadt, Germany, moments after Schiaparelli stopped sending signals on Oct. 19, 2016.
Immediately after the Schiaparelli was supposed to make its soft landing on the surface of Mars, ESA confirmed that a ground-based telescope in Pune, India, had heard the signal from the lander, indicating that key milestones, including the opening of the parachute and the activation of the rocket-propelled landing system had taken place, however no signal from the surface had been detected. Signals recorded by both the Pune station and by the Mars Express orbiter stopped shortly before the module was expected to touchdown on the surface.
Fortunately for the ESA engineering team, the TGO orbiter recorded up to 600 megabytes of data from the Schiaparelli's entry into the Martian atmosphere and its descent to the surface. All this information was successfully downlinked on the ground in the early hours of October 20.
The very preliminary analysis of the data revealed a number of serious problems in the final phase of the parachute descent. The telemetry showed that the back heat shield holding the parachute had been ejected earlier than scheduled -- 50 seconds instead of 30 seconds before the touchdown. (Other sources said the parachute had separated 15 seconds prematurely). Also, the lander was apparently descending at a speed higher than planned. There were also indications that the soft-landing engines had fired for only three or four seconds and all communications from the lander were cut 19 seconds later, or shortly before touchdown. By that time, Schiaparelli's landing radar had been activated.
Less than 24 hours after the botched landing, ESA engineers expressed confidence that they had received enough data to find out exactly what happened, but it would take some time to sift through all the information and interpret the findings. As a result, even in failure, the lander still provided much of the engineering data sought from the mission. There was also early hope that measurements from onboard engineering sensors collected by the AMELIA payload during the entry and descent could also be recovered.
ESA promised to continue attempts to communicate with the lander in the coming days using available orbiters and to make an effort to locate the lander or its remnants on the surface of Mars.
Sure enough, by October 21, NASA's sharp-eyed Mars Reconnaissance Orbiter, MRO, imaged the likely wreckage of Schiaparelli on the surface of Mars almost exactly at the center of the planned landing ellipse.
Below the main image of the landing ellipse area are a pair of before-and-after images, taken by the Context Camera, CTX camera on NASA's Mars Reconnaissance Orbiter on 29 May 2016 (left) and 20 October 2016 (right), respectively at a resolution of six meters per pixel. The Oct. 20 image revealed two new features appearing following the arrival of the Schiaparelli test lander module on the martian surface on October 19. One of the features is bright and can be associated with the 12-m diameter parachute used in the second stage of Schiaparelli’s descent, after the initial heat shield entry.
The parachute and the associated back shield were released from Schiaparelli prior to the final phase, during which its nine thrusters should have slowed it to a standstill just above the surface. The other new feature is a fuzzy dark patch roughly 15 by 40 meters in size and about one kilometer north of the parachute. This is interpreted as arising from the impact of the Schiaparelli module itself following a much longer free fall than planned, after the thrusters were switched off prematurely.
The landing ellipse is 100 kilometers and 15 kilometers, and is centered on 2 degrees south in latitude and 353 degrees east longitude, in the Meridiani Planum region of Mars, close to the planet's equator. The image measures about 100 km; north is up. The dark spot on the image, associated with the Schiaparelli module, is located approximately 5.4 kilometers west of the center of the landing ellipse.
More detailed photos of the Schiaparelli's crash site from NASA's Mars Reconnaissance Orbiter, were released on Oct. 27, 2016, two days after they had been obtained by the orbiter's HiRISE camera, the most powerful imaging instrument ever orbiting Mars:
According to ESA, the zoomed insets provide close-up views of what are thought to be several different hardware components associated with the module’s descent to the martian surface. These are interpreted as the front heatshield, the parachute and the rear heatshield to which the parachute is still attached, and the impact site of the module itself.
In the image, north is up; west to the left. Schiaparelli was travelling from west to east. The image scale is 29.5 centimeters per pixel. The brightness of the individual zooms have been adjusted to best reveal the features against the martian surface in each case.
The 100-meter scale bar in the main image is only indicative, as the HiRISE image was taken at an oblique angle. The distances given between the various components in the main text have been corrected for this effect.
According to NASA, the impact of Schiaparelli left a 2.4-meter crater with an estimated depth of 0.5 meters. A dark curving line extending northeast from the crater was not typical for an impact event and could not be immediately explained, NASA said.
On November 1, MRO photographed the crash site through red, green and blue filters, enabling to produce color images. Photos were taken looking slightly to the west, while the earlier image was looking to the east, providing a contrasting viewing geometry, ESA announced.
According to the agency's press release on November 3, "...the latest image set shed new light on some of the details that could only be speculated from the first look. For example, a number of the bright white spots around the dark region interpreted as the impact site are confirmed as real objects – they are not likely to be imaging ‘noise’ – and therefore are most likely fragments of Schiaparelli.
Interestingly, a bright feature can just be made out in the place where the dark crater was identified in last week’s image. This may be associated with the module, but the images so far are not conclusive.
A bright fuzzy patch revealed in the colour image alongside the dark streaks to the west of the crater could be surface material disturbed in the impact or from a subsequent explosion or explosive decompression of the module’s fuel tanks, for example.
About 0.9 kilometers to the south, the parachute and rear heatshield have also now been imaged in colour. In the time that has elapsed since the last image was taken on Oct. 25, 2016, the outline of the parachute has changed. The most logical explanation is that it has been shifted in the wind, in this case slightly to the west. This phenomenon was also observed by MRO in images of the parachute used by NASA’s Curiosity rover.
A stereo reconstruction of this image in the future will also help to confirm the orientation of the rear heatshield. The pattern of bright and dark patches suggest it is sitting such that we see the outside of the heatshield and the signature of the way in which the external layer of insulation has burned away in some parts and not others – as expected.
Finally, the front heatshield has been imaged again in black and white – its location falls outside of the colour region imaged by MRO – and shows no changes. Because of the different viewing geometry between the two image sets, this confirms that the bright spots are not specular reflections, and must therefore be related to the intrinsic brightness of the object. That is, it is most likely the bright multilayer thermal insulation that covers the inside of the front heatshield, as suggested last week.
Further imaging was planned in about two weeks, and it will be interesting to see if any further changes are noticed.
The images may provide more pieces of the puzzle as to what happened to Schiaparelli as it approached the martian surface.
Following its successful atmospheric entry and subsequent slowing due to heatshield and parachute deceleration, the internal investigation into the root cause of the problems encountered by Schiaparelli in the latter stages of its six-minute descent continues. An independent inquiry board has been initiated, ESA announced on November 3.
Within days after the Schiaparelli's botched landing, ESA engineers suspected that the guidance and navigation, GNS, software had been a culprit in the failure, commanding the premature cutoff of the propulsion system, which led to a free fall from an altitiude between four and two kilometers, possibly, with the module's propellant tanks still nearly full. The spacecraft slammed into the surface with a speed of higher than 300 kilometers per hour and, possibly, exploded.
Surprisingly, radar altimeter data, accelerometers and other sensors were delivering consistent and expected data, as far as ESA experts could reconstruct the events from the limited set of data captured by the TGO orbiter.
Investigators immediately focused on the main flight control computer, which had made a decision to cut off the engines. According to Paolo Ferri, Director of Flight Operations at ESA, between two and three weeks would be required to find the culprit.
Also, on October 21, Thales Alenia Space, which built the Schiaparelli module, issued a statement saying that "the retrorockets were briefly fired up, however it appears that they may have shut off earlier than anticipated, at a height that still has to be specified. This is the result of the partial, ongoing analyses of the data that Schiaparelli sent to the mother ship during the descent stage," the company said.
Since the module’s descent trajectory was observed from three different locations, the teams are confident that they will be able to reconstruct the chain of events with great accuracy, ESA said.
According to the agency, the teams continue to decode the data extracted from the recording of Schiaparelli descent signals recorded by the ExoMars TGO in order to establish correlations with the measurements made with the Giant Metrewave Radio Telescope, GMRT, an experimental telescope array located near Pune, India, and with ESA’s Mars Express from orbit.
By October 24, engineers narrowed down a possible culprit to an error in the software of the Schiaparelli's Doppler radar altimeter, which misled the main computer into thinking that the spacecraft had already reached the landing altitude. The radar software apparently froze and failed to respond to the queries from the guidance and navigation computer, leading to the premature release of the parachute. Then, the operation system decided that the lander had reached the surface and switched off the engines.
On October 27, Rolf Densing, the Head of the European mission control center in Darmstadt, Germany, was quoted in German media as saying that some unexpected movement of the parachute could confuse the flight control system into thinking that the landing had already taken place. This phenomenon apparently manifested itself during only a few out of the thousands of tests, but it was considered so unlikely that no action was deemed necessary to negate it.
On November 23, 2016, ESA announced the preliminary results of the investigation into the Schiaparelli crash on the Martian surface on October 19.
ESA said that good progress has been made in investigating the accident. According to the agency, the large volume of data recovered from the Mars lander shows that the atmospheric entry and associated braking occurred exactly as expected. The parachute deployed normally at an altitude of 12 kilometers and a speed of 1,730 kilometers per hour. The vehicle’s heat shield, having served its purpose, was released at an altitude of 7.8 kilometers.
As Schiaparelli descended under its parachute, its radar Doppler altimeter functioned correctly and the measurements were included in the guidance, navigation and control system. However, saturation – maximum measurement – of the Inertial Measurement Unit, IMU, occurred shortly after the parachute deployment. The IMU measures the rotation rates of the vehicle. Its output was generally as predicted except for this event, which persisted for about one second – longer than would be expected, the agency said.
ESA provided no explanation or any hypothesis at the time what could cause the "saturation" of the IMU.
The ESA statement continued: when merged into the navigation system, the erroneous information generated an estimated altitude that was negative – that is, below ground level. This in turn successively triggered a premature release of the parachute and the back shell, a brief firing of the braking thrusters and finally activation of the on-ground systems as if Schiaparelli had already landed. In reality, the vehicle was still at an altitude of around 3.7 kilometers.
This behavior has been clearly reproduced in computer simulations of the control system’s response to the erroneous information.
“This is still a very preliminary conclusion of our technical investigations,” says David Parker, ESA’s Director of Human Spaceflight and Robotic Exploration. “The full picture will be provided in early 2017 by the future report of an external independent inquiry board, which is now being set up, as requested by ESA’s Director General, under the chairmanship of ESA’s Inspector General.
Meanwhile, scientific data from the instruments aboard Schiaparelli during the entry, plus tracking data from the ExoMars Trace Gas Orbiter, Mars Express and India’s Giant Metre Wave Radio Telescope India have been passed to the science teams. These data will contribute to understanding of the Red Planet and especially its atmosphere, ESA said.
Schiaparelli was landing in familiar neighborhood
Schiaparelli landing site on the global map of Mars relative to previous successful landers. Credit: ESA
Schiaparelli headed to the Meridiani Planum, the same region where NASA's Opportunity rover landed on Jan. 25, 2004. The EDM was designed to land on a terrain with rocks as high as 40 centimeters and slopes as steep as 12.5 degrees, however the lander had no obstacle avoidance capability, so it needed some luck.
The egg-shaped probable landing zone was centered around 6.15 degrees West longitude and just 1.9 degrees South of the Martian equator. The ellipse extended around 100 kilometers from East to West and is 15 kilometers wide from North to South.
The Schiaparelli lander will aim at the center of the 100 by 15-kilometer ellipse on the Meridiani Planum. Credit: ESA
The mission planners favored this location partially because it is low enough to provide some extra depth for braking using the planet's thin atmosphere. The site was around one kilometer below Mars' de-facto equivalent of the sea level known as MOLA. MOLA stands for the Mars Orbiter Laser Altimeter, which was installed on NASA's Mars Global Surveyor spacecraft and was used to create elevation maps of the planet.
This area interests scientists because it contains an ancient layer of hematite, an iron oxide that, on Earth, almost always forms in an environment containing liquid water. It is also relatively flat and considered generally safe for landing.
Like some of its unfortunate predecessors, Schiaparelli landed at the end of the Martian summer, known for its vicious dust storms. The bad weather was predicted to peak on October 29, or 10 days after the landing. However, European engineers were confident that the state-of-the-art spacecraft had been well designed to survive the experience. In fact, the probe's DREAMS instrument could make more scientifically interesting measurements thanks to the increased dust content in the atmosphere, ESA officials said.
Planned surface operations
Following the touchdown, Schiaparelli was programmed to continue transmitting signals for around 15 minutes, before switching to sleep mode to conserve battery power. It was expected to wake up on a programmed schedule as orbiters pass overhead to receive and relay its data back to Earth. These relay slots included 18 provided by NASA’s Mars Reconnaissance Orbiter, MRO, eight windows provided by Mars Odyssey and six by Maven. ESA’s Mars Express was to conduct a planned series of 14 overflights.
During Schiaparelli's operation on the surface, every sol, the lander had two opportunities to talk to MRO as it flew overhead. The windows via the European Mars Express spacecraft were more limited due to its orbit.
Once on the surface, the DREAMS set of science instruments was to be activated for operation during at least two martian days. DREAMS activities were also optimized to make the most of the limited energy available, so they had to be performed in predefined windows rather than operating continuously. Typically, DREAMS was expected to function for six hours every sol (martian day lasting 24 hours and 37 minutes).
The most critical engineering data gathered during the descent and landing of Schiaparelli should be transmitted back home within two martian days. The descent images from the lander are expected to be uplinked via MRO on October 20.
It would take eight sols to relay the complete set of data acquired during the mission.
The hope was to complete all the important transmissions by the time the batteries onboard Schiaparelli would go dead around October 23, 2016.
Read (and see) much more about the history of the Russian space program in a richly illustrated, large-format glossy edition:
Page author: Anatoly Zak; Last update: June 28, 2019
Page editor: Alain Chabot; Last edit: November 25, 2016
All rights reserved
Schiaparelli will be enveloped by plasma during its initial entry into the atmosphere of Mars at 14:42 GMT on October 19. Click to enlarge. Credit: ESA
An artist rendering of the Schiaparelli lander at the end of aerodynamic braking. Click to enlarge. Credit: ESA
Click to enlarge. Credit: ESA
14:45 GMT: Parachute release on Schiaparelli. Click to enlarge. Credit: ESA
An artist rendering of the Schiaparelli lander during a parachute descent. Click to enlarge. Credit: ESA
14:46 GMT: Front shield is jettisoned at altitude of seven kilometers and descent speed of 320 kilometers per hour. Click to enlarge. Credit: ESA
Click to enlarge. Credit: ESA
In this simulation, the ‘footprints’ of Schiaparelli’s 15 descent camera images follow a roughly clockwise order, as seen in this view from larger to smaller fields-of-view. In this simulated view, the first image is taken at about three-kilometer altitude and covers roughly 17 square kilometers, while the final image is taken at an altitude of about 1.5 kilometers and covers 4.6 square kilometers. The images are taken every 1.5 seconds. The scene captured by each image during real landing will depend on the actual altitude, but also on the movement and rotation of Schiaparelli during the descent through the atmosphere. The view is generated from a NASA MRO/CTX image of the center of the Schiaparelli landing ellipse. Click to enlarge. Credit: ESA
14:47 GMT: Back shield is jettisoned at an altitude of 1.3 kilometers at descent speed of 240 kilometers per hour. Click to enlarge. Credit: ESA
Click to enlarge. Credit: ESA
Schiaparelli's final descent under rocket power. Click to enlarge. Credit: ESA
An artist rendering of the Schiaparelli lander on the surface of Mars. Click to enlarge. Credit: ESA
The landing site on Meridiani Planum is a relatively smooth, flat region on Mars. The largest crater to the right (East) of the image is Endeavor, which is about 22 kilometers in diameter. NASA's Opportunity rover has been studying its western rim since 2011. Opportunity has discovered the presence of rocks that were altered by contact with liquid water, billions of years ago. The minerals that formed in Meridiani indicate that the region was once covered by mildly acidic groundwater or freshwater lakes that could have been suitable for the evolution of primitive life. Some of the impact craters in this image are surrounded by muddy material that flowed out of them during their formation. Click to enlarge. Credit: ESA
The Schiaparelli landing site in context. Click to enlarge. Credit: ESA
The GMRT antenna array in India. Click to enlarge. Credit: ESA
A mockup of the Schiaparelli lander is being prepared for a post-landing event at the European control center in Darmstadt, Gemany. Click to enlarge. Credit: ESA