The development of new satellite monitoring technology has accelerated in recent years. New types of optical sensors, as well as the active introduction of synthetic aperture radars (SAR), have significantly expanded the “field of vision” offered by surveillance satellites. These new advantages have been evident in several domains ranging from environmental protection to military affairs.
Today, we will examine how satellite observation of the Earth has evolved in recent years, including new types of ultra-sensitive sensors for environmental monitoring, innovative AI-enhanced image processing algorithms, and the introduction of new services for space situational awareness (SSA).
New environmental monitoring sensors
The American National Oceanic and Atmospheric Administration (NOAA) will soon be employing new Atmospheric Composition (ACX) satellite sensors after NASA awarded the British BAE Systems a $365 million contract this year.
The new sensor will be a hypersensitive hyperspectral camera capable of observing several spectral ranges at once, from ultraviolet to visible light. The introduction of this sensor will significantly expand NOAA’s capabilities for the detection of ozone, nitrogen dioxide, formaldehyde, glyoxal, and sulfur dioxide emissions, making it an important addition to the agency’s existing tools for environmental monitoring.
The ACX sensor is designed for use in new satellites that will be added to NOAA’s Geostationary Extended Observations (GeoXO) constellation. This small constellation currently consists of only three satellites in geostationary orbits. Other BAE Systems technology will also be featured on the new satellites, including the GeoXO Hyperspectral IR Sounder, which is a device for measuring atmospheric moisture content, tracking hurricanes, and weather forecasting.
The functionality of the constellation will be complemented by other sensors as well, including the GXI Vis/IR Imager (for real-time infrared spectrum observation), the LMX Lighting Mapper (lightning mapping through an optical sensor), and the OCX Ocean Colour Instrument (a hyperspectral sensor for ultraviolet measurement using an integrated near-infrared passive radiometer).
Source: nesdis.noaa.gov
New types of sensitive hyperspectral sensors for high-resolution monitoring of greenhouse gas emissions will also be introduced in the GHGSat satellite constellation. The development of these sensors is being handled by ABB, which previously designed seven satellites for this constellation.
It is expected that these new types of hyperspectral sensors will be installed on four satellites, which are scheduled for launch at the end of 2024. In August 2024, GHGSat, the world leader in greenhouse gas emissions monitoring, signed a contract with Space Flight Laboratory to build two additional satellites to monitor air pollution, meaning that the total number of GHGSat satellites is set to increase to 13.
These efforts have had a real impact: GHGSat’s space-based analysis of greenhouse gas emissions helped reduce harmful CO2 emissions by 5.6 million tons between 2021 and 2024. The expansion of the GHGSat constellation is aimed at further decreasing CO2 emissions from large industrial enterprises. The biggest advantage of satellite monitoring of CO2 emissions is observational objectivity: the satellite either detects dangerous emissions or records clean air. This eliminates the possibility of industrial giants concealing or downplaying data about emissions from their facilities.
Implementation of AI algorithms in satellite image recognition
New types of sensors are not the only factor enhancing the capabilities of Earth observation satellites. At this year’s World Economic Forum, the authors of a report called “The Catalytic Potential of Artificial Intelligence for Earth Observation” concluded that new machine learning algorithms now being used during the image post-processing phase significantly enhance object recognition and greatly reduce the human resources previously required for such tasks. A human operator now only needs to verify the data provided by AI, removing incorrectly identified objects.
The use of AI in satellite monitoring processes increasingly includes the combination of disparate data sets into a single overall picture. This allows for a more holistic and unbiased analysis and enables a more accurate assessment of actual conditions. Among the other benefits of introducing neural networks into the recognition process is the use of machine learning (ML). As the number of image processing cycles increases, the neural network gradually reduces the percentage of incorrectly identified objects. Over time, the AI becomes capable not only of distinguishing objects in satellite images but also of identifying their origins. This is especially important during environmental monitoring, identifying the causes of methane emissions, monitoring the formation of landfills, etc. Machine learning has also significantly improved the work of climate and weather observation satellites.
Integrating machine learning with a wide range of satellite data also allows the creation of basic models (also known as fundamental models) that allow the analysis of large volumes of data. Such models help avoid investing months of work in creating and training a neural network for recognition “from scratch.”
It is also now much easier to work with satellite data using simplified user interfaces (UI) that allow not only trained technical specialists but also ordinary businessmen to work with satellite observation data using only their smartphones or PCs.
Source: eos.com
This approach provides a significant boost to the creation of business models, which is why the satellite monitoring and data processing market is continuously attracting new players. The scalability offered by new client-oriented services and the implementation of AI in the recognition process are among the major factors driving the growth of the Earth observation satellite sector.
The use of AI in satellite image processing can also make the recognition process more transparent and unbiased, in turn encouraging new clients and creating a market for new companies that aim to provide similar services.
The military also benefits from these kinds of developments. Last year, the U.S. Defense Advanced Research Projects Agency (DARPA) requested the integration of AI into the data recognition process used by American military satellites, signing a $7 million contract with BAE Systems to implement the plan. Two other companies specializing in AI system development, OmniTeq and Apogee Research, along with Systems & Technology Research, will also be involved in this project.
The introduction of AI automation into the post-processing of satellite images is especially important when it comes to deploying large military satellite constellations. Previously, processing such vast amounts of data required a wide range of ground operators, who will now function largely as overseers in the analysis process. The new software will be integrated with various sensors that monitor everything from optical data to radio frequencies.
The first phase of DARPA’s program will last 15 months, during which time the companies are expected to develop new software and test it under terrestrial conditions. The second phase will involve the deployment of new satellites in orbit. In particular, sensors using machine learning algorithms will be installed on Space Development Agency (SDA) satellites and will be focused on tracking and monitoring new types of maneuverable hypersonic missiles.
The U.S. Space Force (USSF) also continues to work with companies on improving the accuracy of existing sensors. One potential solution in this area could be using neural networks to artificially enhance the resolution of previously obtained images using post-processing. However, this technique requires further refinement since the technology is still in the early stages of implementation and the AI sometimes introduces distortions during image enhancement.
Laser inter-satellite communication: new opportunities in data transfer
Another major development in the satellite monitoring industry is linked to the emergence of free-space optical communication (FSO), also known as laser satellite communication.
The new signal relay technology, which can be used in satellite-to-satellite or satellite-to-ground communications, has positively impacted both the speed and the security of large-scale data transmission due to the use of narrow (compared to radio frequency) and focused laser beams. Lasers are also less susceptible to existing electromagnetic and radio frequency signal interception systems.
In addition to enhanced data transmission speed and signal security, FSO has also significantly reduced the energy consumption requirements for data transfer on spacecraft. This, in turn, is expected to directly influence the size of future satellites, making them more compact and cheaper to manufacture.
Laser inter-satellite data transmission technology has already been demonstrated in space missions, including on the Japanese NICT’s SOTA CubeSats (which achieved a maximum data transfer rate of 10 Mbps), the German DLR’s OSIRISv2 CubeSats (which achieved 1 Gbps), and NASA’s OCSD (which achieved the first high-speed laser downlink communication channel).
Source: dlr.de
In just a few years, we should see the implementation of laser communication technology on the vast majority of both civilian and military near-Earth monitoring satellites. New SDA satellites will be able to transmit data via 3-5 laser communication channels, with the help of which military satellites can communicate not only with ground stations but also with aircraft, ships, and other satellites. This will ultimately facilitate the creation of an extensive multi-level satellite telecom architecture.
At its current stage of development, laser communication technology still faces some difficulties. These include susceptibility to weather conditions (atmospheric attenuation can deflect or suppress optical waves), problems with guidance, reception, and tracking (since the technology requires high accuracy to target the laser on the receiver), and relatively high initial investment costs.
New developments in space situational awareness (SSA) services
Customer-oriented space situational awareness (SSA) services are used to expand and accelerate the systematization of information received from satellites connected to a single network, offering clients tracking, Earth monitoring, orbital control, and other services. As part of a unified architecture, SSA satellites can quickly provide necessary data upon request. This is especially useful for organizing emergency responses to natural and technological disasters or for taking swift action on national and global security issues.
SSA services give users insights into the movement and capabilities of thousands of spacecraft passing overhead daily. Consequently, most space awareness services function as a sort of space radar, capable of providing clients with highly sensitive operational information about conditions on Earth and in space. In recent years, the number of players in this field has increased.
One such company is Beyond Gravity (formerly RUAG Space), which offers clients access to a database containing information from over 10,000 satellites that can be accessed in real-time.
SSA is also proving to be extremely valuable for strategic decision-making within limited timeframes, and this is why such services are in increasingly high demand in the world’s militaries. For instance, the Italian National Armaments Directorate (NAD) has expressed interest in acquiring three new ground-based SSA sensors to enhance the country’s awareness of the space environment.
The system required by NAD will consist of two optical sensors complemented by a highly sensitive radar station. Together, they will provide the Italian military with complete information on space objects up to 15 cm in low Earth orbit (LEO) and objects smaller than 35 cm in medium Earth orbit (MEO) in the areas of the sky where the observation takes place. The new Italian SSA service will also be able to localize objects in higher orbits, such as GEO, but it is not yet known what size they must be to be detected by the new system.
The first optical sensor, called Flyeye, is designed to enhance Italy’s space awareness and is being developed by OHB Italia. This ground-based optical sensor, standing 6.5 meters tall and 4 meters wide, resembles a wide-field telescope. Flyeye’s capabilities will be complemented by a second, smaller, and less powerful ground-based optical telescope and a ground-based radar station that will capture radio signatures left by spacecraft in orbit.
It is anticipated that NAD’s new SSA system will be able to classify detected objects, distinguishing between satellites, small orbital debris, or minor asteroids orbiting the Earth.
Source: ohb-italia.it
Italy plans to deploy a trial version of this system by 2027, with full operational capability expected in 2029. The U.S. is also seeking to update its SSA resources. In the spring of 2024, the USSF announced plans to deploy a new SSA satellite constellation in geostationary orbit. Currently, the USSF has space situational awareness capabilities through its Geosynchronous Space Situational Awareness Program (GSSAP), whose satellites can not only monitor adversaries like Russia and China but also perform orbital rendezvous with unfriendly satellites and transmit data about them back to Earth.
Although the GSSAP constellation consists of only five satellites (a sixth having been deactivated in 2023), their orbital characteristics (at over 35,000 km) allow them to cover nearly the entire Earth. The main goal of augmenting the existing GSSAP constellation with new satellites is to extend their operational lifespan by equipping them with a port for orbital refueling, which current satellites lack. Given that orbital refueling technology is still in its infancy, GSSAP satellites with these capabilities will likely start appearing in orbit only toward the end of the 2020s.
Overall, SSA services are becoming increasingly effective and powerful, primarily due to advancements in computer recognition technologies, many of which utilize machine learning algorithms. Additionally, new types of optical sensors and high-power radar systems are improving the quality and speed of detection of even very small objects.
The latest major leap in SSA development involves the implementation of new software standardization methods, allowing space situational awareness systems from different countries to be integrated into a unified network. This significantly speeds up detection times and increases accuracy in observations by combining optical sensor and radar resources.
New possibilities in satellite spectrometry
One of the important satellite monitoring missions is observing biodiversity and ecosystems. In 2022, a scientific study was conducted by Anna Schweiger from the Remote Sensing Laboratory of the Geography Department at the University of Zurich (UZH) and Étienne Laliberté from the University of Montreal.
The aim of the study, which was called “Beta-Diversity of Plants in Biomes Using Imaging Spectroscopy,” was to determine the biodiversity of terrestrial flora from space using new types of satellite spectrometers. This research will help track the dynamics of plant biodiversity changes across the planet, from tropical rainforests to Arctic tundras.
The primary criterion for obtaining data using satellite spectrometry is the reflectance coefficient of various types of plants. Since each type of terrestrial flora has a different chemical structure and plant cover density, they reflect sunlight differently. This allows for the classification of plants by type and facilitates ongoing observations of their condition over many years.
Spectrometry measurements are conducted across different ranges of the electromagnetic spectrum: from short infrared wavelengths to longer wavelengths visible to the human eye. Measuring reflected light enables the construction of indices indicating plant type, current condition, regional moisture content, and more.
The beta-diversity study was conducted using satellite data provided by the National Ecological Observatory Network (NEON) with ultra-high spatial resolution (1×1 m per pixel). For other types of research, this resolution is typically 20×20 m per pixel. Recent advances by ESA and NASA suggest that plant ecosystem changes can be monitored even using satellites with a resolution of 30×30 m per pixel, provided data collection occurs at least every 16 days. Such systems are currently being developed by both agencies and are expected to be used in the future for detecting ecosystem change dynamics.
The EOS SAT satellites from EOS Data Analytics (EOSDA) operate using satellite spectrometry. This is the world’s first agricultural satellite constellation, primarily aimed at observing the croplands of large agricultural enterprises, as well as monitoring the condition of forests and other types of flora. Currently, only one satellite, EOS SAT-1, has been launched into orbit as part of this future constellation. The company plans to increase the number of satellites operating in low Earth orbit (LEO) to seven. The total daily coverage area of the constellation is expected to be 12 million square kilometers.
Growing cooperation with the commercial sector
Increasingly, military agencies are inviting the commercial aerospace sector to collaborate on the acquisition and processing of satellite imagery. The figures are striking: $4 billion out of the USSF’s $30 billion 2024 budget will be allocated specifically to developing commercial capabilities, which the military is increasingly integrating into intelligence and SSA processes.
Indeed, the establishment of a new Commercial Space Office within the USSF indicates that collaboration between the military and the private sector will only continue to strengthen. Such collaboration is not limited to involving private satellite companies in military operations; they can also provide satellite observation data for disaster localization, both technological and natural. The commercial sector may also be engaged in infrared observation missions and offer the military alternative positioning, navigation, and timing capabilities, significantly expanding the capabilities that the USSF possesses with its own fleet of satellites.
Over the past decade, new satellite sensors and space situational awareness services have exponentially increased the amount of commercially available Earth observation data. The entire commercial satellite observation sector is estimated to generate several hundred terabytes of data per day, a figure projected to grow substantially in coming years as the number of monitoring satellites in orbit, especially from the commercial sector, continues to grow.
Currently, the main obstacle is the lack of standardized software protocols that would allow American commercial satellites to directly connect with military systems for signal retransmission. This issue has significantly slowed integration between military agencies and the commercial sector. However, by mid-2024, it is clear that the USSF’s transition from a strategy of acquiring its own satellites to procuring services from commercial satellites is already a fait accompli.
