Tiny Satellites Big Picture

It’s 3:00 a.m. and Los Alamos engineer Tracy Gambill is trying to wake up her satellite. The satellite, called NACHOS (Nano-satellite Atmospheric Chemistry Hyperspectral Observation System), passes over the Los Alamos command station several times every 24 hours, but sometimes the best window for communication is ten minutes in the middle of the night, so that’s when she does it.

NACHOS is the first of its kind. It is a small satellite currently in orbit, purpose-built to “photograph” specific gases in Earth’s atmosphere. The satellite is tiny, with limited power, so it must operate efficiently. When NACHOS is not actively gathering data, it conserves its resources in a sleep mode. Gambill sends a few “wake” commands and waits for a reply so that she can establish a secure connection to the onboard computer. Once connected, Gambill uploads detailed instructions—prepared by her colleague, Lab electrical engineer Hannah Mohr—about exactly how the satellite should rotate to point toward the Mount Merapi volcano in Indonesia to take an image of the sulfur dioxide (SO₂) gas being emitted. 

The satellite maneuver must be exact. Many satellites observe large target areas—such as the whole continent of North America—sampled coarsely in chunks tens of kilometers across. NACHOS images areas about a hundred kilometers across and can resolve targets as small as a football stadium. Not only does NACHOS point at something relatively small, it is surprisingly small itself. NACHOS is built as a “CubeSat”, a class of satellites each about the size of a loaf of bread. The use of CubeSats is beginning to be an affordable alternative to large satellites as they can be designed to carry different types of instrumentation for significantly less cost to both build and launch. 

With a vantage point hundreds of miles above Earth, many satellites in low-earth orbit measure atmospheric gases—like NACHOS does—to help scientists understand terrestrial phenomena such as pollution or the climate. Most of these satellites, however, use large, power-hungry instruments that deliver only coarse-grain data. The Los Alamos team miniaturized these instruments into a novel system. NACHOS is more than fifty times smaller and lighter than satellites that collect similar data, uses significantly less power, and was designed to deliver high-resolution images. Now, two years into its first deployment, NACHOS is successfully sending home high-quality images of atmospheric gas distributions and demonstrating the possibilities of a whole new era of satellite technology. 

Just a trace

NACHOS began its orbit in July of 2022, but its journey to the launchpad began about twenty years ago. Physicist Steven Love, the principal investigator for NACHOS, has spent much of his Los Alamos career developing various types of instruments to detect and study chemicals—often making them smaller to be used for remote sensing. Love, who grew up in Washington state and remembers vividly the 1980 eruption of Mount St. Helens, always had an interest in volcanos and believed remote sensing could be useful for volcano science since many volcanologists make risky trips to get close to eruptions to study their activity. Furthermore, some volcanos are too remote to visit, but their plumes could be of concern when they reach populated areas.

Some satellites point at the whole continent of North America; NACHOS aims for targets the size of a football stadium.

While studying the Popocatépetl volcano near Mexico City—on the ground, a safe distance away—Love’s team detected a small but distinct change in the composition of gases just before the volcano erupted, including an increase in SO₂ and silicon tetrafluoride. These whiffs of gas piqued the scientists’ interest in studying gases that only occur in trace amounts. 

Earth’s atmosphere contains large amounts of ozone, carbon dioxide, and methane and many instruments can measure their abundance. But some gases that occur in relatively small, or trace, amounts are of interest to scientists because they contribute to air pollution and impact public health. Monitoring SO₂ from volcanos could lead to safety improvements, but SO₂ is also produced by burning sulfur-rich coal and is a major contributor to forest- and crop-damaging acid rain, a significant issue for the well-being of the planet. Another trace gas, nitrogen dioxide (NO₂), is a major constituent of smog. NO₂ can be created when oxygen, ozone, and sunlight react with nitric oxide (NO), which is produced by burning fossil fuels in cars or power plants. 

Chronic exposure to pollutants such as SO₂ and NO₂ gases can cause health problems such as asthma, chronic obstructive pulmonary disease, and cardiovascular disease. The World Health Organization estimates that millions of people die prematurely each year due to outdoor air pollution. To thoroughly understand these threats, scientists studying air pollution rely on elaborate data collection stations and direct-sampling point sensors around the world. However, to get detailed data on the atmospheric distribution and chemical evolution of gases such as SO₂ and NO₂, airborne or space-based monitoring is better because these methods can document the concentration of gases at their points of origin, and then trace how they spread through the air.

Around the time of the Popocatépetl volcano experiments, a technique called hyperspectral imaging was emerging. Hyperspectral imaging is a method that takes a photo of a scene while simultaneously breaking incoming light into its component wavelengths and recording a “picture” of the absorption patterns, so that spectral fingerprints can be used to identify which chemicals are present in the scene and where.

“Hyperspectral imaging is sort of like a super-detailed version of color imaging. It’s useful because every pixel of the image contains a high-resolution spectrum with hundreds of wavelength channels—more than just red, green, blue—so you can look at the detailed structure of the spectrum of gases and very reliably distinguish one gas from another,” says Love.

As the use of hyperspectral imaging became more widespread, Love recognized its value for detecting trace gases. However, the instruments were massive and required powerful computers that could process very large amounts of data. As a result, many ground-based instruments used hyperspectral imaging, but few satellites had the capability. Love knew if hyperspectral imagers could be smaller and more portable, they could be extremely useful. Love envisioned miniature hyperspectral imagers on drones or even satellites to monitor trace gases—and after years of dedication, he and his colleagues have made it happen.

Bloodhound in the sky 

CubeSats were first developed in the early 2000s as inexpensive, quick-to-build alternatives to large satellites that could be used for distinct, short missions. In 2008, Los Alamos began developing its own CubeSats and, once some reliable designs were established, began to consider what instrumentation could be included as a payload to collect scientific data. Most instruments robust enough to travel through space and sensitive enough to gather valuable data are large and require a lot of energy to function. Using a CubeSat to collect data would require significant miniaturization and adaptation of any existing technology. Around 2015, Love finally got the opportunity to make a hyperspectral imager small enough for a CubeSat.

To fit a hyperspectral imager in a satellite the size of a loaf of bread, Love assembled a team of engineers and software specialists. This project would be a challenge. The team needed to build a small hyperspectral imager but also consider every aspect, from power supplies to the physical robustness needed to survive space travel. The first internally funded project was called Targeted Atmospheric Chemistry Observations from Space (TACOS). The team built an extra-small spectrometer using two concave mirrors and a convex grating component. One mirror directs incoming light to the grating component, which spectrally disperses the light. The second mirror focuses the dispersed light onto a charge-coupled device (CCD) detector array to form the image. Once the TACOS prototype was proven successful, the team received funding from NASA to build it into a CubeSat and launch it into orbit —and the NACHOS project was born. 

Another major function that had to be miniaturized was data processing. By nature, hyperspectral imaging creates very large data files—hundreds of independently measured wavelength bands comprising each pixel of the image. But a CubeSat in orbit has very limited communications bandwidth, much like a frustratingly slow modem connection. Realizing that it would take far too long to transmit the instrument’s enormous raw data files from orbit back home for processing on the ground, the team developed a method for onboard data processing. Using custom algorithms, Los Alamos software engineers programmed NACHOS to be able to process raw data onboard; this approach reduces hundreds of megabytes of data into small sets of one-megabyte gas-density images for each target. The images are quickly downloaded to the team on Earth and the raw data can then be erased from the NACHOS hardware to make space for the next batch of data. 

Blastoff

In July 2022, the NACHOS CubeSat was launched into orbit, riding aboard the Virgin Orbit LauncherOne air-dropped vehicle along with six other CubeSats built by institutions around the country. Ride-sharing is one of the big advantages of CubeSats; larger satellites usually require their own rocket to carry them to space, but CubeSats can just hitch a ride with another mission. 

Since its deployment, the NACHOS CubeSat has been orbiting Earth at an altitude of 500km and will remain in orbit for roughly another year, at which point it will fall back through the atmosphere and burn up. During its time in orbit, the team is working fervently to collect as much data as they can—and they admit there were some challenges at first. 

CubeSats include hardware to help them orient themselves toward their targets back on Earth. This hardware, called an attitude determination and control system, fits within the infrastructure half of the loaf-of-bread sized CubeSat (the other half contains the hyperspectral imager) and is an engineering achievement in itself. NACHOS was built in a CubeSat design that had previously been used for other instruments that required less sophisticated attitude determination (they merely needed to aim a radio antenna in the right general direction). For NACHOS, the CubeSat needs to point and scan with far more precision than had ever been demanded of it before because the hyperspectral imager must be accurately aimed at the target and then smoothly scanned across it. To do this, the attitude system has sensors to detect the sun, magnetometers to detect the earth’s magnetic field, and gyroscopes to measure how the CubeSat is spinning. Each time the ground team prepares to take data from NACHOS, engineer Hannah Mohr calibrates the CubeSat’s position using the attitude instruments and calculates how to direct it to move into position to begin scanning the specific target.

Hyperspectral imaging allows scientists to reliably distinguish one gas from another.

“You can’t see if the satellite is responding to your instructions,” says Mohr. “The only knowledge we have comes from the sensors, and you can’t always trust the sensors since they might be influenced by the environment.” Mohr says that one time the satellite was meant to be recording a measurement from the sun, but she discovered that it was actually the sun’s reflection off a solar panel that was being recorded. 

Mohr spends a great deal of time choosing targets that will have clear and sunny weather at the time NACHOS is overhead, and comparing images to other photos from space—from other satellites—to help ensure that NACHOS can get a good direct view of the desired target. Once in position, NACHOS collects data for about 90 seconds, scanning over a specific target using a so-called push-broom technique. In this technique, NACHOS images a one-pixel-wide, 350-pixel-long stripe across the target, dispersing the wavelengths from each of the 350 pixels across its CCD detector array to produce a 2D frame containing spectral information along one dimension and spatial information along the other. NACHOS collects a rapid series of hundreds of these 2D data frames as the satellite sweeps, like a push-broom, across the target. This collection of 2D data frames, all stacked together, make up a 3D hyperspectral data cube comprising both spectral and spatial dimensions. 

With these data, Love and the other scientists can observe the concentration and location of various gas species as they change over the target area. Overall, this approach is more informative than a single-point detector on Earth. For instance, when evaluating smog in a given city, a single detector would just get an overall NO₂ reading for the area local to the detector, while NACHOS can show a higher concentration of NO₂ near a particular power plant and also show how far it disperses and in what direction. Early tests using a NACHOS hyperspectral imager set up in the back of a passenger car in Farmington, New Mexico, showed NO₂ both in the air and pooling near the ground by a power plant—the latter a key detail that was missed by other techniques. 

“The Environmental Protection Agency only has fixed-position sensors that don’t detect NO₂ production and pooling like we observed near the Farmington power plant,” says Los Alamos scientist Manvendra Dubey, who has helped analyze data from early NACHOS prototypes and the one in orbit. Dubey says that the NACHOS design could be equally useful for remote sensing on Earth as well as above.

So far, NACHOS has collected data from several sites worldwide, including the Mount Merapi volcano in Indonesia, where it caught the volcano in the act of spewing a plume of ash, and congested cities such as Tokyo and Naples, where it imaged the NO₂ smog generated by these cities. 

Constellation of CubeSats

Although the clock is ticking on how long NACHOS will be able to collect data before it falls out of orbit, each new target is teaching the Los Alamos team more and more about what their invention can do. Love says that the vision is to one day launch an entire “constellation” of NACHOS CubeSats. With this expanded capability, the team could collect data from multiple targets within a larger geographic area and create a more comprehensive analysis of the distribution and evolution of trace gases. Ultimately, dynamic and agile remote sensing could help lead to better prediction of when a volcano is going to erupt or more detailed smog warnings in city centers. 

“Another reason to study NO₂ is for climate monitoring. Manmade CO₂ is tricky to detect from space because it is difficult to distinguish manmade emissions from the complex variations of CO₂ naturally found in the atmosphere. Monitoring NO₂ is a useful substitute because it indicates fossil fuel use from human activity, and is much more easily detected from space,” says Love. A potential use in this respect would be to verify that cities (or specific power plants) are meeting their greenhouse gas reduction goals set forth in future climate treaties. 

In 2023, the NACHOS scientists won an award for their design and the team is excited that with every new data collect they are perfecting their technique so that a future constellation of remote sensors could be realized. In the meantime, the single orbiting NACHOS satellite is going strong. Gambill regularly checks on its system health in between passes, and carefully maintains its software so that it is ready to go whenever the team needs it—even if it that means getting up at 3:00 a.m. LDRD