This article originally appeared in the May 14, 2018 issue of SpaceNews magazine.
A new generation of radar satellites coupled with high-speed computer processing and machine learning could dramatically expand the use of radar data.
That conviction was palpable in April when leading radar satellite proponents gathered in Tampa, Florida, for a party at the 2018 GEOINT Symposium thrown by Ursa Space Systems, a geospatial data analytics company. One by one, radar satellite veterans and recent converts took the microphone to profess their faith in the technology and its promise.
“I am a card-carrying member of the radar mafia,” said Keith Masback, U.S. Geospatial Intelligence Foundation chief executive. “I was an infantry officer who needed to fight at night and in bad weather. That left me high and wet when it came to electro-optical. I, along with my special operations colleagues, absolutely appreciated the value of radar.”
Unlike electro-optical satellites, which require light and clear skies to take pictures, Synthetic Aperture Radar (SAR) satellites work in all weather conditions, through clouds, day and night, because they bounce microwave pulses off targets and measure the reflected energy.
That makes SAR satellites extremely versatile but also far more complicated to build than electro-optical spacecraft. They require significant onboard power and need large apertures to view a wide ground swath. In addition, SAR datasets require far more processing than electro-optical imagery.
“SAR is strange,” said Thomas Ager, former lead SAR engineer for the U.S. National Geospatial-Intelligence Agency’s Information Technology Directorate, who teaches a course called The Essentials of SAR. “It’s intimidation factor is off the charts. My mission is to try to replace the confusion and intimidation with clarity and interest.”
To date, radar satellites have been primarily designed, built and operated by prime contractors working with government space agencies. Government agencies, like the National Geospatial-Intelligence Agency (NGA), also have been the dominant customers for the data these satellites produce.
“The value of this supply and the value of the suppliers who can deliver it at the commercial level is absolutely critical,” said Justin Poole, NGA deputy director. “As NGA moves forward and we look to broker the types of supply we need to attack the challenges in this ever-changing world, commercial radar is absolutely critical.”
Startups backed primarily by private venture capital have announced plans to launch dozens of small SAR satellites within a decade to take advantage of the miniaturization of electronics. The new companies plan to lean heavily on advances in computer processing and machine learning to extract useful information from their data streams.
Meanwhile, industry veterans like Airbus Defence and Space, MDA and the Italian Space Agency are preparing to update their constellations of large radar satellites that produce high-value products, like maps that can pinpoint changes in elevation of millimeters. Digital Elevation Models created with SAR can show, for example, “if land is sinking or a building has moved or if somebody tunneling made the surface of the land above them shift,” said Mike Greenley, president of MDA, a Maxar Technologies company. “It’s very sensitive to elevation, which is important to a lot of customers.”
Airbus’ next generation of large radar satellites offering high resolution across a wide ground swath will “truly be a game changer for the business,” said Greg Buckman, managing director, president and chief executive for Airbus Defense and Space Intelligence North America. “We will complement that capability with many of the smaller radar satellites and partners we have in the audience. Couple that with machine learning and a radar image has more information than any electro-optical image,” he said at the party in Tampa.
In recent years, the commercial market for electro-optical imagery has grown rapidly as customers like hedge funds, insurance companies and multinational construction firms have adopted it to keep tabs on business activity around the world. Companies preparing to launch new SAR satellites anticipate similar strong demand.
Ursa’s staff has tripled since December to keep up with demand for its products, which often pair SAR with other types of information. For instance, Ursa and Clipper Data, a company based in New York that tracks crude oil and petroleum shipments, produce weekly reports on crude oil demand in China.
“We have finally peeled back enough layers of the onion to figure out what sells on the commercial side, what you have to do to go from a measurement to a product that answers the customer’s question,” said Adam Maher, Ursa chief executive.
SAR entrepreneurs are bullish on their business prospects but not everyone is convinced that commercial demand will grow quickly enough to justify all the new radar satellites. ICEYE of Finland, which produced SAR imagery with its first 70-kilogram satellite launched in January, plans to operate a constellation of 18 satellites. In early May, an investor revealed than Umbra Lab, a secretive startup based in Santa Barbara, California, plans to build 12 SAR satellites. XpressSar of Arlington, Virginia, received a NOAA license in 2015 for four SAR satellites. Capella Space of San Francisco won a NOAA license in March for two SAR satellites, although the company has talked about flying a constellation of 36 to revisit targets hourly.
Whether commercial market will be large enough support all the new entrants is an open question, Chirag Parikh, NGA deputy director for counter-proliferation, said during an April interview. “Are there enough mining companies, shipping companies, agricultural entities, land management companies who would use this?” It’s hard to say.
But the SAR mafia thinks so.