The Future of Digital Elevation Models


Until recently, the world knew more about the terrain of the moon and Mars than it did about Earth’s polar regions. But today, thanks to groundbreaking collaboration among geospatial experts across United States government, industry, and academia, the Arctic is now one of the best-mapped places in the world.

According to nearly all scientific accounts, the Arctic is also one of the fastest naturally changing places on Earth and is experiencing global warming at extreme rates. In May 2013, the White House published the country’s first National Strategy for the Arctic Region, followed in November of that year by the Department of Defense’s (DoD) release of its Arctic Strategy. Both documents acknowledge heightened U.S. national security interests in the Arctic as ice melt causes it to transition from a state of isolation to one of increasing access.

“The [DoD strategy] recognizes the role that the Arctic region will play in shaping the global security environment in the 21st century,” then Secretary of Defense Chuck Hagel wrote in the foreword. “As we monitor how changes in the Arctic influence geopolitical landscapes, we will balance our Arctic investments against the Department’s responsibilities and objectives around the world, while collaborating domestically and internationally to help develop effective solutions.”

The Arctic gained even more national attention in 2015, when the U.S. took the helm of the Arctic Council, an intergovernmental forum among eight stakeholder nations that rotate chairmanship every two years. In a related Executive Order titled “Enhancing Coordination of National Efforts in the Arctic,” the White House listed U.S. interests in the region as: “national defense; sovereign rights and responsibilities; maritime safety; energy and economic benefits; environmental stewardship; promotion of science and research; and preservation of the rights, freedoms, and uses of the sea as reflected in international law.”

In September 2015, in coincidence with President Obama’s tour of the Arctic, the National Geospatial- Intelligence Agency (NGA) released its unclassified Arctic website. That same day, the President announced the agency’s ambitious public-private partnership, formed to develop the first public, high-resolution, satellite-based digital elevation model (DEM) of Alaska in 2016, and of the entire Arctic by 2017.

“The public website sat there for a year with all of these great products, but the DEM piece was missing,” said Brian Bates, a data scientist with NGA’s Office of Strategic Operations.

The resulting project, funded by the National Science Foundation (NSF) and known as Arctic DEM, exceeded expectations and brought together a powerful coalition to produce an unprecedented geospatial product that no member could have achieved alone.

“[This project] is extremely beneficial to our community that wants to do change studies and attribution,” said Dr. Kelly Falkner, director of NSF’s Office of Polar Programs, of the decision to fund Arctic DEM. “We could see the scientific benefit in addition to the operational benefit.”

SPARKING INNOVATION

The common standard for global DEMs are those generated from NASA’s Space Radar Topography Mission (SRTM) in 2000. The 30-meter dataset was made public in 2015, but no information was collected north or south of 60 degrees, i.e., Earth’s poles.

“Prior to 2010 and our partnership with the National Science Foundation, NGA did not have a lot of unclassified satellite imagery requirements in the polar regions, but in that same year, through our commercial partnership with DigitalGlobe, NGA’s capacity to collect large amounts of unclassified satellite data over these areas of the Earth had increased dramatically,” Bates said.

NGA began collecting stereo imagery from polar orbits in 2010 through its commercial imagery program, according to Bates. In 2015, in accordance with the President’s Executive Order and with plans for Arctic DEM in the early stages, the agency redoubled its efforts, aiming to collect as many stereo pairs of Arctic and Antarctic imagery as possible.

“Our colleagues [at Ohio State] came up with an algorithm that could take stereo images and create the z-axis mapping and get DEMs produced out of electro-optical satellite images at the unclassified level,” Bates said.

But at the 2015 release of NGA’s public Arctic website, this algorithm was still immature.

“We went through months of working closely with the scientists, making corrections, and holding test runs before we thought we could produce a product we would all be proud to put our names on,” Bates said.

Once the algorithm was ready for prime time, the team needed an unclassified, high-performance computer in order to apply it at scale. Paul Morin, director of the Polar Geospatial Center at the University of Minnesota, which served as the lead organization on the Arctic DEM project, contacted the National Center for Supercomputing Applications at the University of Illinois at Urbana-Champaign. The center’s Blue Waters supercomputer is one of the world’s fastest and largest computers, according to Bill Kramer, senior associate director of the Blue Waters Project Office.

The Blue Waters office only takes on partners with substantial missions that would be nearly impossible to achieve without the use of the supercomputer.

“By concentrating the computing and analysis features of Blue Waters plus applying our special support staff we make these projects productive and decrease their time to solution,” Kramer said. “We’re not wide breadth with thousands of people using Blue Waters. It’s the people who have a frontier scientific challenging in front of them, and in this case the DEM work far exceeded those expectations.”

Blue Waters’ vast computing, data storage, data movement, and wide area networking capabilities enabled the Arctic DEM team to deliver its first data release ahead of schedule, and to eventually go on to produce the Reference Elevation Model of Antarctica (REMA), a similar project for Earth’s southern pole.

THE REFERENCE ELEVATION MODEL OF ANTARCTICA

Fueled by the success of Arctic DEM, the same team—this time led by The Ohio State University—pressed on to produce the Reference Elevation Model of Antarctica (REMA) in September 2018. REMA boasts a resolution of 2 to 8 meters, compared to 1,000 meters, which was typical for previous maps of the continent. According to a study published in February by the National Academy of Sciences, Antarctica is losing six times as much ice as it was 40 years ago. And Antarctica has significantly more ice than the Arctic, at least on land, but is much flatter. Since elevation is what requires such extensive computing power to process, REMA was able to be produced in a matter of a few months instead of over a couple of years. (Photo credit: Polar Geospatial Center)

Once the DEMs were generated using Blue Waters, the next step was to make the massive data trove publicly available. Esri’s ArcGIS Image Server is designed to manage extensive volumes of imagery and to process the data on the fly, according to Peter Becker, the company’s senior product manager for imagery.  

“On the fly means the data is processed as you access it instead of pre-processing,” Becker said, adding that this means users can actually take measurements of the data and study change rather than simply panning and zooming across cached web tiles.

“Users can say they want to see the difference between two different dates,” Becker continued. “The server will find those dates, subtract between one and the other, and send that difference back to the user.”

Today, following its seventh data release in September 2018 at 2-meter resolution, visitors to the Arctic DEM website can draw a line to designate a desired area of study and view the DEMs covering that area—in some cases as many as 40 to 60. Users can calculate the volumetric loss of elevation (often ice) or conduct time series plotting with results that are accurate within a couple of meters.

As a result, environmentalists, policymakers, and national security leaders alike are able to observe quickly and in detail information related to changes in navigation routes, deforestation, glacial retreat, coastal erosion, the construction of new roads and buildings, and much more. The information has been a treasure trove for scientific research and is also highly useful for planning operations in the north without the need for expensive, risky, or near-impossible reconnaissance flights.

Scientists are using Arctic DEM data of the Lena Delta at the end of a remote river in Siberia to study river systems that are otherwise inaccessible. (Photo credit: NGA)

“This is a component of Arctic situational awareness,” Morin said. “We’re talking about denied airspace, in a form that it’s not denied by an air force or government, but simply because of geography.”

As the U.S., other Arctic nations, and interested observers work to create a global framework for the modernization and expansion of commercial and government activities in the Arctic, the ability to quickly and accurately monitor change will become increasingly useful.

“We’re talking about elevation as a standard derived product from remote sensing at this point—just something that’s produced as a matter of course,” Morin said.

“Before, people were doing change detection, but from pictures. That’s very different than change detection from surfaces.”

Falkner emphasized NSF’s long history with mapping agencies such as the U.S. Geological Survey (USGS) that have a vested interested in mapping the Arctic. The Arctic Spatial Data Infrastructure group, born under the Atlantic Council and led by USGS, takes a more GIS-based approach to understanding the region and unifies the mapping agencies of all Arctic nations for data sharing.

“USGS has very strong standards for mapping and there was some skepticism initially,” Falkner said in reference to Artic DEM. “But as they began to see the products coming out they evolved their thinking. I think we sparked the innovation and others are starting to pick up on it.”

A MULTI-DISCIPLINARY FEAT

The interdisciplinary, highly collaborative Arctic DEM project created an entirely new way of developing accurate surface maps—one that likely couldn’t be achieved by any other country, said Kramer, pointing to all of the physical and intellectual resources that came together to make this possible.

Falkner hailed the collaboration across government, industry, and academia as “extraordinary” and added, “I can’t tell you how beautiful this is if you’re a scientist.”

Morin noted the significance of taking a scientific approach to the daunting task.

“What we did was revolutionary—and when we did it there wasn’t a spec written,” Morin said. “We didn’t know what the accuracy, repeat, or precision would be. That was all written after it was produced. It wasn’t the government way, but the scientific way. In science, we write the paper after the dataset is produced. In government, it is the other way around—the government writes the [RFP] and then vendors have to meet that.”

Bates said by partnering with civil federal agencies and universities, NGA was able to achieve much more than its individual budget and capabilities would have allowed.

Now, the polar science and geospatial communities are rallying around how to analyze the DEM.

“Even here where we built this stuff we have a hard time dealing with how to analyze this much,” Morin said. “What do you do with 400 terabytes of data? And that number is never going to get any smaller. … We’ve already burned piles of computing time producing this, now we need more time to figure out what’s in it.”  

NGA is already experiencing a steady stream of requests for unclassified 2-meter data of Alaska, Greenland, and Russia, according to Chuck Crittenden, an applied scientist with the agency. In an effort to perform automated quality assurance work on Arctic DEM data and expand upon those efforts, NGA awarded a five-year, $15 million contract to GeoNorth Information Systems (GNIS) in August 2018.

Anchorage-based GNIS, in partnership with Lockheed Martin and the University of Alaska Fairbanks, will provide NGA with access to scalable geospatial data processing tools to produce specified foundation products and services.

“NGA has extensive capabilities and very refined and detailed data products,” said Jonathan Heinsius, general manager and director of geospatial programs at GNIS. “They are very interested in having data products and services that are readily shareable and open for a place like the Arctic, which is very much about international collaboration.”

Part of this effort will include scouring Arctic DEM to automatically correct for any data artifacts, according to Crittenden.

“There’s a lot of land to go through,” Crittenden said.

“For us to do that in our standard production would take a long time.”

The innovative Arctic DEM project has shown the value and potential of time-dependent terrain, and many geospatial experts are now interested in applying the method for other areas of the globe.

Crittenden added he would like to see a process similar to Arctic DEM used for coastal areas around the world.

“In the event of a disaster or emergency we would have unclassified data we could share immediately to help first responders,” he said.

Morin concluded that Arctic DEM has ushered in a “golden age of topography,” and said, “Between this and LiDAR we can know the structure of the Earth’s surface like we’ve never known before and we can watch it change.”

 


Guarding the Arctic Coasts


As the physical and geopolitical landscape in the Arctic continues to evolve, the U.S. Coast Guard is going to be present in the region more often and in increasing numbers, according to two Coast Guard analysts working at the National Geospatial-Intelligence Agency: Jason Tucker, a civilian, and IS1 Rob Wright, who is active duty.

“GEOINT is of utmost importance to the Arctic,” Wright said.

Tucker and Wright both said they are experiencing significantly more requests for Arctic-related work compared with four or five years ago. They provide the Coast Guard’s Intelligence Coordination Center with visual representations of changing ice coverage, increasing vessel activity, regional infrastructure, and predictions such as the potential for mineral consumption or anticipated ice melt.

In addition to illustrating polar ice transformation, the analysts are also able to reveal navigable waters and how vessels transit these waters at times they haven’t been able to in the past, according to Tucker.

Doing so helps the Coast Guard gauge its preparedness for more frequent search-and-rescue missions in the Arctic.

“The Crystal Serenity cruise ship went through in 2016 with thousands of people on it and there’s really not any search-and-rescue infrastructure up there,” Wright said. “What if the ship had hit a rock? We’re showing there’s a lot more people in the Arctic, a lot more vessel activity.”  

Recently, the duo’s analysis helped inform Congress’ decision to fund six Polar Security Cutters to replace the Coast Guard’s two legacy ships. At press time in April, the service was said to be close to awarding a contract for the first icebreaker.

Wright and Tucker created a graphic that visualized Arctic maritime domain awareness, including SAR agreements and available infrastructure such as airfields, airports, and port areas.

“[Decision-makers] could see all at once, this is what the operating environment looks like currently, this is what we think the operating environment is going to look like in the next 20 to 50 years, and take that environment into account to realize there is a need for a new Polar Security Cutter,” Wright said.

Looking ahead, Wright and Tucker predict the demand for Arctic GEOINT products will continue to grow as new cutter capabilities enable the Coast Guard to have a stronger presence in the far north.


The U.S. Coast Guard Cutter Healy carries around 100 crew members and 30 scientists to deploy sensors and semi-autonomous submarines to study how environmental factors affect water below the ice surface for the Office of Naval Research. (Photo credit: Nyxolyno Cangemi/U.S. Coast Guard)

CUBESATS FOR SEARCH & RESCUE

The Coast Guard Research, Development, Test, and Evaluation (RDT&E) Program is testing two cubesats with the Department of Homeland Security (DHS) Science & Technology Directorate to evaluate the use of space-based sensors in support of Arctic search-and-rescue missions.

In December two cubesats, dubbed Yukon and Kodiak, were launched into low-Earth polar orbit on a rideshare mission from Vandenberg Air Force Base in California. The Polar Scout project will be used to inform satellite technology recommendations for many potential applications within the Coast Guard and across DHS.

In addition to the cubesats, the program includes two Mobile CubeSat Command and Control (MC3) ground stations, one at the Coast Guard Academy in New London, Conn., and one at University of Alaska Fairbanks. The team will use a civilian icebreaker to take out Emergency Position Indicating Radio Beacons (EPIRBs) for testing this summer.

“The [cubesat sensors] will be able to detect the [EPIRB] stress beacon, geo-locate it, and send that information to the MC3 ground station,” said Holly Wendlin, C4ISR Domain Lead for RDT&E.

Wendlin said the Polar Scout team aims to conclude its final report by the end of the year.


Eyes on the Ice


The latest generation of remote sensing satellite technology offers more ways than ever for humans to learn about the Earth’s changing poles.

NASA’S NEWEST ICESAT

In September 2018, NASA launched its ICESat-2 photon-counting LiDAR satellite to characterize Earth’s polar ice sheets and sea ice.

ICESat-2’s laser pulses 10,000 times per second, sending light to the ground, collecting the returning photon in its telescope, and recording the photon travel time. Because the speed of light is constant, travel time can be converted to distance traveled. But doing so requires precise knowledge of the satellite’s location using GPS and star trackers.

According to Tom Wagner, NASA’s ICESat-2 program scientist, the original ICESat, launched in 2003 and de-orbited in 2010, had 170 meters between footprints (spots illuminated), but ICESat-2 has only 70 centimeters in between. Wagner used a football field analogy to quantify the advance.

“The original ICESat put one footprint in each end zone, and ICESat-2 gives us a footprint on every yard line,” he said.

To “bridge the gap” between the two ICESats, NASA’s IceBridge airborne survey mission mapped each polar region once a year, flying over the Arctic from March to May and over Antarctica from October to November. The new satellite mission will decrease the need for costly, risky, and technologically limited flights, according to Wagner. And so far, ICESat-2 is exceeding expectations.

“We’re getting a far greater signal from the ground than expected, and with the photon-counting mission, we’re going to be able to see through the clouds in most areas,” Wagner said. “ICESat-2, we’re hoping, is going to give us a new look at what goes on, especially during the warmest periods.”

This will lead to new insights, considering planes aren’t able to fly below cloud cover, and clouds are more frequent during warmer seasons. The advanced ICESat-2 technique also appears to be able to see through water, mapping heights in streams or reservoirs and in some cases seeing depths down to 30 meters.

“A lot of people will be interested in this data,” Wagner said, adding that groups such as NOAA and the Army Corps of Engineers have already taken notice.

The first ICESat-2 data release is expected to take place in the first half of 2019.

FINLAND’S ICEYE STARTUP

One of the first use cases Finland-based ICEYE explored for its commercial synthetic aperture radar (SAR) satellite constellation was monitoring Arctic conditions in real time, according to company chief scientific officer and co-founder Pekka Laurila.

“A key thing in the Arctic is literally half of the year it’s dark and a majority of the time it’s cloudy,” Laurila said, highlighting why Arctic surveillance is an ideal application for SAR technology, which can “see” through clouds and darkness.

Now that ICEYE has successfully demonstrated the size, cost, and operational formats of its technology with its ICEYE-X1 and ICEYE-X2 satellites, both launched in 2018, the company has set the ambitious goal to multiply its fleet and create “the world’s largest constellation of SAR satellites,” according to Laurila.

Currently, many operations in the Arctic, such as those underway by the oil and gas industry, require 24/7 aircraft monitoring to track sea ice and ensure other safety measures—but flying the aircraft itself can be dangerous.

As the company’s constellation expands and its revisit rates increase, Laurila predicts ICEYE’s technology can begin to replace aircraft-level monitoring within a couple of years.

“Getting the revisit rates to increase to more than once per day starts to bring the majority of these operational use cases into the commercially interesting range,” Laurila said, adding the company aims to “very quickly get to a point where we are talking about mere hours between imaging opportunities.”

This ICEYE-X2 radar satellite image from Prince William Sound in Alaska shows ice flowing to the sea. (Image credit: ICEYE)

CANADA’S GRAY JAY PATHFINDER PROJECT

In February, Canada’s Department of National Defence (DND) awarded a C$15 million contract to Space Flight Laboratory (SFL) at the University of Toronto Institute for Aerospace Studies to develop multipurpose microsatellites in support of Arctic surveillance.

As ice melts in the Arctic, new routes are being created that could enable encroachment by maritime vessels, said Dr. Robert Zee, director of SFL.

“Maintaining Canadian sovereignty and continental security has become increasingly more challenging in the 21st century,” Zee said.

DND’s “Strong, Secure, Engaged” policy outlines various initiatives aimed at enhancing the Canadian Armed Forces’ ability to operate in the Arctic and adapt to a changed security environment, including the development of new technologies to improve surveillance and control.

Following testing of a prototype, SFL will build two additional microsatellites to create a small formation. The microsat constellation will include multiple sensors—primarily imaging and radio frequency—operating in close formation in low-Earth orbit to allow for quick detection and identification of surface or airborne targets.

According to Zee, the Gray Jay Pathfinder project, named after the unofficial bird of Canada, will track commercial airliners and vessels in the far north to demonstrate basic hardware and algorithms that if successful could be enhanced for future operational purposes.

“An operational mission would require bigger, better sensors, equipped to detect high-speed, highly maneuverable weapons,” Zee said, adding that increased intercontinental ballistic missile range and cruise missile maneuverability are growing potential threats.

In May 2018, DND expanded the Canadian Air Defence Identification Zone to cover “the entirety of Canada’s Arctic Archipelago and its approaches,” according to a DND spokesperson.

“Countries like Russia and China may be developing defense systems and technology that will challenge those of NORAD,” Zee continued. “Newer technologies are needed to counter those threats and provide sufficient deterrence.”

Related


Posted by Kristin Quinn