Maintaining the health of an aging Mars orbiter

The Mars 2020 rover will land in February 2021. Like its predecessors Curiosity, Opportunity, and Spirit, it will need to return most of its data through orbital relays. So it's kind of a problem that there aren't currently any new NASA Mars orbiters planned. The most recently-arrived orbiter is MAVEN, and it's expected to do a great deal of data relay for Mars 2020. But its orbit can't produce the repetitive afternoon communication session that Mars orbiters have come to rely on. My colleagues Casey Dreier and Jason Callahan wrote an article about this problem last year.

The Opportunity and Curiosity rovers have been sending data to Earth via the Odyssey and Mars Reconnaissance Orbiters. Both of these workhorse spacecraft are far, far past their warrantied lives (as is Opportunity and the one other orbiter capable of data relay, ESA's Mars Express). Every new year, I predict that we'll finish out the year having seen at least one of these aging spacecraft go silent, and every new year's eve I'm delighted to find out I've been wrong. Through the tender care of their Earth-based engineering teams, these elderly spacecraft have kept going.

On February 9, NASA announced some changes to how the engineers are operating Mars Reconnaissance Orbiter in order to prolong its life as long as possible, hopefully long enough to support the Mars 2020 rover's operations "through the mid-2020s." There are two planned changes, one to navigation and one to its orbit. The navigation change -- a switch from the use of aging inertial measurement units to maintain its knowledge of its orientation -- will be implemented in March. The orbit change won't be implemented until after Mars 2020 lands.

The press release, "Mars Reconnaissance Orbiter Preparing for Years Ahead", explains the changes in excellent detail so I'll just quote it here, and discuss some further implications for MRO after:

NASA’s Mars Reconnaissance Orbiter (MRO) has begun extra stargazing to help the space agency accomplish advances in Mars exploration over the next decade....

In early February, MRO completed its final full-swapover test using only stellar navigation to sense and maintain the spacecraft’s orientation, without gyroscopes or accelerometers. The project is evaluating the recent test and planning to shift indefinitely to this “all-stellar” mode in March.

From MRO’s 2005 launch until the “all-stellar” capability was uploaded as a software patch last year, the spacecraft always used an inertial measurement unit -- containing gyros and accelerometers -- for attitude control. At Mars, the orbiter’s attitude changes almost continuously, with relation to the Sun and other stars, as it rotates once per orbit to keep its science instruments pointed downward at Mars.

The spacecraft carries a spare inertial measurement unit. The mission switched from the primary unit to the spare after about 58,000 hours of use, when the primary began showing signs of limited life several years ago. The spare shows normal life progression after 52,000 hours, but now needs to be conserved for when it will be most needed, while the star tracker handles attitude determination for routine operations.

The star tracker, which also has a backup on board, uses a camera to image the sky and pattern-recognition software to discern which bright stars are in the field of view. This allows the system to identify the spacecraft’s orientation at that moment. Repeating the observations up to several times per second very accurately provides the rate and direction of attitude change.

“In all-stellar mode, we can do normal science and normal relay,” [project manager Dan] Johnston said. “The inertial measurement unit powers back on only when it’s needed, such as during safe mode, orbital trim maneuvers, or communications coverage during critical events around a Mars landing.” Safe mode is a precautionary status the spacecraft enters when it senses unexpected conditions. Precise attitude control is then essential for maintaining communications with Earth and keeping the solar array facing the Sun for power.

To prolong battery life, the project is conditioning the two batteries to hold more charge, reducing demand on the batteries, and is planning to reduce the time the orbiter spends in Mars’ shadow, when sunlight can’t reach the solar arrays. The spacecraft uses its batteries only when it is in shadow, currently for about 40 minutes of every two-hour orbit.

The batteries are recharged by the orbiter’s two large solar arrays. The mission now charges the batteries higher than before, to increase their capacity and lifespan. It has reduced the draw on them, in part by adjusting heater temperatures before the spacecraft enters shadow. The adjustment preheats vital parts while solar power is available so the heaters’ drain on the batteries, while in shadow, can be reduced.

The near-circle of MRO’s orbit stays at nearly the same angle to the Sun, as Mars orbits the Sun and rotates beneath the spacecraft. By design, as the orbiter passes over the sunlit side of the planet during each orbit, the ground beneath it is about halfway between noon and sunset. By shifting the orbit to later in the afternoon, mission managers could reduce the amount of time the spacecraft spends in Mars’ shadow each orbit. NASA’s Mars Odyssey spacecraft, older than MRO, successfully did this a few years ago. This option to extend battery life would not be used until after MRO has supported new Mars mission landings in 2018 and 2021 by receiving transmissions during the landers’ critical arrival events.

I'd gotten wind of the proposed change to the orbit via Twitter, when some people who work on HiRISE targeting observations mentioned it. Changing the orbit will have an effect on HiRISE science because when its sun-synchronous orbit shifts to a later time of day, the shadows cast by topography will stretch longer across the surface.

Much of HiRISE's science these days is focused on repeat imaging of places on the Martian surface where change happens: recurring slope lineae, slope streaks, and polar ices, to name a few. The orbit change won't make this work impossible, but will complicate it. People who care about polar processes will be especially hard hit, because moving the orbit an hour later in the day will dramatically reduce the period of the Mars year that some parts of the poles are visible. That being said, if the orbit is shifted late enough in the day (4:30 is what's currently being discussed), it's possible that HiRISE will, for the first time, be able to see some polar terrain very early in the morning as well as in the afternoon, during polar summer.

HiRISE also puts a lot of effort into repeat imaging of interesting topography for stereo (3D) views, and that, too, will be less effective if the shadows are notably longer in the second image than the first, taken before an orbit change. Even if both images are taken after the orbit change, a later orbit means that there will be much more seasonal change in shadow positions, so the HiRISE team will have to sequence both images for the stereo observation as close to each other in time as possible.

The good news is, the HiRISE team has a few years to wrap up their best-quality data sets for stereo views and active surface processes. Except there's one other problem with HiRISE, mentioned briefly in the press release:

[S]ome HiRISE images taken in 2017 and early 2018 show slight blurring not seen earlier in the mission. The cause is under investigation. The percentage of full-resolution images with blurring peaked at 70 percent last October, at about the time when Mars was at the point in its orbit farthest from the Sun. The percentage has since declined to less than 20 percent. Even before the first blurred images were seen, observations with HiRISE commonly used a technique that covers more ground area at half the resolution. This still provides higher resolution than any other camera orbiting Mars -- about 2 feet (60 centimeters) per pixel -- and little blurring has appeared in the resulting images.

The problem is discussed in much greater detail in a 2018 Lunar and Planetary Science Conference abstract by HiRISE principal investigator Alfred McEwen (PDF). For multiple reasons, HiRISE is now acquiring more of its images in a mode where it averages four adjacent pixels together, producing images with coarser resolution (about 60 centimeters per pixel instead of 30 centimeters per pixel). Still, 60 centimeters per pixel is better than any other orbiter can get. A plus side of operating HiRISE in this mode is that they can get four times longer swaths for the same cost in terms of data volume. [EDIT: Just after publishing this post, McEwen emailed me to tell me they may have a fix for the blurring problem. Stay tuned.]

There's one other instrument on the orbiter that's showing its age, but the rest are fine:

Using two spectrometers, CRISM can detect a wide range of minerals on Mars. The longer-wavelength spectrometer requires cooling to detect signatures of many minerals, including some associated with water, such as carbonates. To do this during the two-year prime science mission, CRISM used three cryocoolers, one at a time, to keep detectors at minus 235 Fahrenheit (minus 148 Celsius) or colder. A decade later, two of the cryocoolers no longer work. The last has become unreliable, but is still under evaluation after 34,000 hours of operation. Without a cryocooler, CRISM can still observe some near-infrared light at wavelengths valuable for detecting iron oxide and sulfate minerals that indicate past wet environments on Mars.

The Context Camera (CTX) continues as it has throughout the mission, adding to near-global coverage and searching for changes on the surface. The Shallow Radar (SHARAD) continues to probe the subsurface of Mars, looking for layering and ice. Two instruments for studying the atmosphere -- the Mars Color Imager (MARCI) and Mars Climate Sounder (MCS) -- continue to build on nearly six Mars years (about 12 Earth years) of recording weather and climate.

The people who operate Mars Reconnaissance Orbiter are caring for it as best they can. But it won't last forever. I talked with a lot of people about this at JPL at the end of the Cassini mission. In a way, the Cassini end-of-mission event was a happy one because we were able to come together and honor its life at an emotional event marking its end. Mars Reconnaissance Orbiter is very unlikely to end the same way. One day, it'll just stop talking to Earth because some crucial component has failed. NASA will spend time searching for it and trying to regain contact. Maybe they'll command another spacecraft to try to shoot a photo of it, though nobody else really has the resolution right now. (Possibly ExoMars CaSSIS.) After weeks or months of no response, they'll give up, and declare the mission over.

In the meantime, let's enjoy Mars Reconnaissance Orbiter for however long it lasts, and advocate for the development of a new orbiter for lander relay. There isn't a direct opportunity to do that right now, but if you sign up for our email updates, you can be notified when such an opportunity exists.