How satellite remote sensing is enhancing our understanding of Earth

Although Shawn Kefauver had directly seen oil companies do ‘shady stuff’ in Ecuador, it was the analysis of data recorded from space he contributed to that bore witness in court, not him. The researcher from the University of Barcelona in Spain and his colleagues wanted to use the light spectrum recorded by satellites to study changes in the Amazon rainforest near Ecuador’s oil facilities. They turned to data from the series of sensors on the Landsat satellites jointly operated by Nasa and the US Geological Survey, first launched in 1972. To establish that the method correctly detected that forest was being cut down, the scientists had to go to sites in Ecuador to establish the ‘ground truth’, Kefauver explains.

Landsat instruments are well suited to monitoring deforestation because they offered high spatial resolution, with down to 30m length, roughly as long as two shipping containers. Their long history of available data also helped. The study confirmed that roads the now defunct US oil company Maxus built to access wells in the early 1990s were the primary sources of environmental damage. However, campaigners found other severe problems, such as another company digging holes in the ground, filling them with waste and leaving.

‘I was just a volunteer translator fresh out of college on that expedition,’ Kefauver says. ‘It was my first introduction to remote sensing.’ But it was a small role in an important project: prosecutors cited the satellite study in the Ecuadorean supreme court as part of the overall case. The verdict went against the oil company this time, Kefauver recalls, including an award of millions of dollars in damages. Yet other related cases have found in favour of the companies, in a controversy that still rages.

Satellites like Landsat detect light interacting with substances in Earth’s atmosphere and on its surface. But it’s not as simple as just launching and pointing the instruments and reading out a conclusion. Instead, scientists must work hard to devise metrics, known as products and indices, to glean meaning from the data. They must also ensure the measurements are accurate, involving not just laboratory validation but also experiments that take scientists to more unusual settings in the atmosphere and ocean. And as new instruments head to orbit, covering more of the electromagnetic spectrum, scientists are accessing ever more information to help do everything from feeding the world to dealing with toxic algae.

An early episode in satellite data helping scientists track substances in the atmosphere came in 1966, from studying US meteorological images. Steven Platnick from the Nasa Goddard Space Flight Centre in Maryland, US says that in them a scientist saw unusually bright lines of low cloud over the sea. He thought that they might originate from fine particles in the form of aerosols coming from the fuels that ships burned, and scientists eventually called the lines ship tracks. In 1994 Platnick and colleagues used Advanced Very High Resolution Radiometer (AVHRR) instruments on the US National Oceanic and Atmospheric Administration’s Polar Orbiting Environmental Satellites to better understand the link. The study provided a clear fingerprint showing how ship emissions create brighter clouds, says Platnick. Yet showing that was rather different to taking a person’s fingerprint.

The power of satellite analysis

Remote sensing instruments detect bands that cover narrow ranges of the electromagnetic spectrum, usually in the visible, infrared and ultraviolet range. AVHRR covered five visible and infrared bands, with the ship track experiment using them to measure reflections from the cloud. Measurements showed that reflectance increased due to aerosols helping increase cloud droplet concentration but reducing their size. ‘Ship tracks are a microcosm for how humans can change climate,’ Platnick adds.

However, AVHRR’s reflectance calibration drifted dramatically in space, creating challenges for quantitative science, Platnick explains. The scientists had to measure reflectance for a place on the ground using other means, and correct their data based on that information. ‘AVHRR had no onboard way to independently derive its calibration,’ Platnick says.

The Modis instruments created a revolution in remote sensing

In the ship track study, Platnick’s team also ‘flew aircraft through the ship plumes’ to measure the particles that formed tracks. One aircraft carried an airborne simulator of what would later become the twin satellite-borne moderate resolution imaging spectroradiometer (Modis) instruments. Today, Platnick leads the Modis atmosphere team and cloud optical and microphysical product algorithm. Modis flies on Nasa’s Terra and Aqua satellites, which have been in orbit for 24 and 22 years, respectively. They view the whole Earth’s surface every one to two days, acquiring data in 36 bands of wavelengths from blue light at 405nm to 14.4μm infrared.

Platnick and colleagues mainly study light absorbed in reflections around 1.6 and 2.1μm to infer the size of water droplets and ice particles. The amount of light in these wavelengths that the droplets and particles absorb relate to their size. Having learned lessons from AVHRR, Modis has been ‘a pathfinder for calibration stability’, according to Platnick. The instrument contains four calibrators, with one example being a solar diffuser that reflects the Sun in ways that scientists can easily calculate. Pointing the spectroradiometer at the diffuser therefore enables them to recalibrate it.

When the Modis instruments launched, they ‘created a revolution in remote sensing’, says Lorraine Remer of University of Maryland, Baltimore County (UMBC) in the US. ‘The longevity of Modis might make it the most important’ of all currently used instruments in remote sensing, she says. Remer develops algorithms that translate measurements such as radiance – the energy carried by light being detected – into ‘products’ that characterise the airborne particulates. Devising ways to convert the data coming back from satellite instruments into such products and other similar ways to understand our planet, usually called indices, isn’t easy.

Clearing up dusty questions

Starting in 1991, Remer helped adapt a way for Modis to study aerosol particles. Called atmospheric optical depth (AOD), this measurement describes the proportion of light that aerosols in the atmosphere block. Remer and colleagues did field experiments using aircraft instruments to test the method’s assumptions and methods. Like Kefauver in the Amazon, once Modis launched Remer ‘helped to validate the products against ground truth’. Improvements in Modis meant AOD could work over the entire planet, including land and ocean, for the first time. It helped answer ‘several aerosol questions as people worked to estimate the role of aerosols in climate change’, says Remer.

One of the most important Modis studies that Remer participated in was a study led by Hongbin Yu of Nasa, showing that half of the aerosols above North America arrived there from other continents, and most of that aerosol was dust. That limits the potential to reduce the amounts of aerosols, which are important contributors to air quality and climate change, using regulations within the US. ‘There is a lot of dust transport from Asia to the Americas,’ Remer explains. ‘Tons and tons. There is a lot of intercontinental transport of materials all over the globe. None of this was imagined before remote sensing.’

Dust can even threaten life in the form of severe storms – something Steve Palmer’s team from the University of Exeter, UK, used Modis to show it might be possible to predict and mitigate. Doing that involved exploiting the fact that Modis’s bands span part of the thermal infrared region, covering medium and long wavelengths from 3.7–14.2µm. In this range, the instrument primarily senses photons emitted from Earth’s atmosphere, clouds, land surface and water surface. These bands are referred to as emissive bands, whereas other satellite-based instruments focus on reflective bands. His team has compared four existing indices that measure dust in the atmosphere, some of which relied on reflective bands, and some the emissive bands. ‘We wanted to quantify how well different approaches worked in our study area, which happened to be Saudi Arabia,’ Palmer explains.

Palmer and colleagues looked at dust storms in the Arabian peninsula between 2000 and 2015. They identified environmental patterns preceding these events, such as wind speed, temperature and rainfall. They showed that it could be possible to develop an early warning system for dust storms, alerting people to imminent threats. ‘There’s economic costs to when dust storms happen, but there’s also a big health impact in terms of respiratory diseases,’ Palmer says.

Now the Exeter team hopes to combine data from Modis with the European Space Agency’s (ESA’s) three Sentinel-2 satellites and Landsat to identify threats to life at sea. They’re looking at rapid and toxic growths of marine microorganisms known as algal blooms. Modis will bring daily observations with a broad field of view and spatial resolution down to 250m. Meanwhile Sentinel-2 offers multispectral imaging across 13 bands with 10m spatial resolution, enabling detailed examination of specific events, such as patterns in algal blooms.

From orbit to the ocean

‘The beauty of Modis is that you get the daily observations, then you can zoom into some events in more detail using Sentinel-2,’ says Palmer. And although the first Landsat only recorded light from four visible light bands, over time new instruments have also expanded into the infrared spectrum. This approach will allow the team to analyse the impacts of conditions like ocean currents on algal blooms, and how they affect industries like tourism and fishing.

Harmful algal blooms can cause illness in humans, and massive die-offs of fish, turtles, manatees and birds, because algal species like Karenia brevis produce potent neurotoxins called brevetoxins. Such blooms have affected large stretches of the Florida coast, explains Kelly Luis from Nasa’s Jet Propulsion Laboratory in Pasadena, US. She’s therefore studying them with the Tropospheric Monitoring Instrument (Tropomi) on ESA’s Sentinel-5 Precursor satellite, launched in 2017. Luis notes that the instrument spans the UV, visible and infrared spectrum. Spectroscopy in the near infrared region from 700–750nm is especially useful for studying algal blooms and water quality, she says.

Scientists can study chlorophyll in algae from space because not all the light it absorbs gets used for photosynthesis. Instead, some gets reemitted, forming two emission peaks from 650–850 nm. As both Tropomi and Modis cover this range, both can analyse this fluorescence to detect algal blooms. Luis’s team has shown that the two instruments produce consistent results when studying ocean biogeochemistry. ‘You had two instruments designed for very different purposes looking at the same window in the electromagnetic spectrum,’ she explains.

That’s surprising because Tropomi is primarily designed for atmospheric monitoring yet offers new possibilities for monitoring algal blooms when other instruments can’t. Unlike traditional instruments requiring clear skies, Tropomi can detect blooms even with atmospheric interference, improving forecasting capabilities.

This is one example of how Luis likes using multiple instruments, while another involves two on board the International Space Station. One, Ecostress, looks at thermal infrared wavelengths like Modis, in this case in the range from 8–12µm. It measures plant temperatures and can detect warmer waters where phytoplankton grow faster. The other instrument, Emit, observes visible to shortwave infrared from 380–2500nm, including the range that’s useful for studying blooms, explains Luis, who is Emit’s aquatic applications lead.

Sometimes Luis’s research takes her into the ocean to confirm the measurements. For chlorophyll fluorescence, she ventures among the algae with a device which emits light and measures resulting fluorescence signals. She can also filter water, collecting algae, and then later separating pigments chromatographically to understand which are responsible for the fluorescence.

Such well-validated plant pigment detection methods are among the best-established and most versatile remote sensing approaches, which other researchers also use to study Earth’s land surface.

New satellites, new perspectives

At the same time as studying Amazon deforestation, Kefauver was among the first to use remote plant growth data to help manage crops precisely. This ‘precision agriculture’ involves identifying how different areas should be treated, for example applying more or less fertiliser, Kefauver explains. This approach is an active and growing discipline, with many companies processing satellite data and providing prescription maps to farmers. Precision agriculture can increase yield and profitability while promoting sustainability by ensuring crops receive only what they need.

Originally the key metric that Kefauver and colleagues used in precision agriculture was the normalised difference vegetation index (NDVI). While many different instruments can provide this information, his team mostly used Landsat. The NDVI assesses how green and healthy plants are by combining two spectral bands, Kefauver explains. ‘One tracks chlorophyll, the other one tracks biomass,’ he says.

Today, Kefauver and his colleagues primarily uses the Sentinel-2 satellites. Compared to Landsat, Sentinel-2 has a multispectral imager that captures many extra spectral bands, which help calculate the content of interesting pigments such as carotenoids, chlorophyll and xanthophylls separately, Kefauver explains.

In a recent study with the United Nations Food and Agriculture Organisation, Kefauver’s team used the NDVI to monitor the spread of the invasive fall armyworm pest as it devours entire crops before turning into a moth and traveling 100km to continue the cycle. ‘It’s like a really hungry worm, kind of like The Very Hungry Caterpillar,’ says Kefauver. By analysing NDVI time series data, the researchers identified negative growth patterns in crops, indicating the pest’s presence and impact. And just as progress from Landsat to Sentinel-2 has broadened Kefauver’s work, new satellites are set to further enhance remote sensing capabilities.

You can start to zero in to what type of phytoplankton might be in the water

According to Remer, the February 2024 launch of Nasa’s Plankton, Aerosol, Cloud Ocean Ecosystem (Pace) satellite is an exciting development. She is Pace’s science and applications deputy team chair and works with two of its instruments: a polarimeter and an optical spectrometer. The former is an advanced multi-angle, multi-wavelength imaging polarimeter, offering much more data than Modis. It retrieves various aerosol properties, including complex refractive index and particle shape and will reduce the amount of uncertainty seen with Modis, Remer explains.

The spectrometer measures across a broad spectral range from the ultraviolet at 350nm to the short-wave infrared up to 2.25µm, a broader range than Modis. ‘That extra reach into the UV provides extremely important information about the aerosol that we could not reach with Modis,’ says Remer. She has helped produce an algorithm to derive aerosol properties from the instrument. Overall, she is generally encouraged by the way that advances such algorithms ‘combine information from different sensors to retrieve very detailed optical, physical and maybe chemical properties of the particles’.

The spectrometer allows detailed analysis of phytoplankton types and their fluorescence, Luis adds. ‘You can start to zero in to what type of phytoplankton might be in the water,’ she says. Luis is also excited about the impending launch of the Geosynchronous Littoral Imaging and Monitoring Radiometer in 2026 or 2027, which will focus on US coasts and the Gulf of Mexico. It will provide multiple daily measurements of phytoplankton fluorescence allowing scientists to understand their dynamics and composition more comprehensively, she says.

Luis emphasises that new missions sense light from across much more of the electromagnetic spectrum. ‘I think as these new missions are going online, we’re going to see new perspectives of spectroscopy from satellite sensing,’ she says.

Andy Extance is a science writer based in Exeter, UK