The community should act now in preparation to generate rich content from massive data
One of the richest yet most under-utilized sources of geospatial intelligence (GEOINT) is about to become more widely available. Earlier this year, ICEYE launched the first microsatellite carrying a synthetic aperture radar (SAR) sensor. The Finnish company expects to have a constellation of 18 SAR platforms in place by 2020. And they won’t be alone. Capella Space and other small satellite startups are planning SAR missions of their own.
Satellite-based SAR technology isn’t unique. MDA, Airbus Defence & Space, the European Space Agency, and e-GEOS have operated traditional-sized radar satellites for decades, and the Swiss company sarmap has been developing technologies to process and analyze SAR data for nearly 20 years. But the operation of SAR sensors aboard relatively inexpensive constellations of small orbiting platforms is new, and ensures the volume, frequency, and timeliness of data will increase dramatically.
These developments present tremendous opportunity—and challenge—for the GEOINT Community.
The Unique Benefits of SAR
SAR technology is best known for utilizing an active sensor that emits radar signals toward the Earth. These signals interact with the surface—whether land, snow, ice or water—and reflect back to the instrument, carrying rich information about size, orientation, composition, condition and texture of the features encountered. SAR data can be captured day or night, at any sun angle, and in all weather conditions, including rain and clouds.
In comparison to passive electro-optical (EO) sensors, this all-weather, day-and-night collection ability alone makes SAR a uniquely useful situational awareness tool for many defense and intelligence applications when EO data is not available.
A SAR sensor essentially creates a two-dimensional radar reflectivity image of the Earth. The SAR instrument sends out radar echoes in the cross-track (range) direction and measures the time (distance) it takes for the pulse to return to the antenna. The motion of the aircraft/spacecraft in the azimuth (along-track) direction allows the SAR to synthetically create a larger antenna than would be physically feasible. In general, the larger synthetic antenna yields finer resolution.
Many space-borne radar systems also have the ability to operate in multiple power modes, collecting high-resolution data of small geographic areas or coarse data over wide regions, and many variations in between. Some radar wavelengths can penetrate dense vegetation and image features below the canopy.
This guaranteed collection of data on a specific revisit schedule without disruption caused by weather significantly enhances the ability to monitor change continuously at a specific point on the Earth’s surface over long periods. The enhanced temporal coverage introduces the fourth dimension—time—to remote monitoring. A series of multiple archived and newly collected SAR data sets can be processed to not only identify change, but also to determine the rate of that change over time.
The addition of dozens of small satellites increasing the frequency of these data collects will substantially improve monitoring and predictive capabilities.
The GEOINT Community already understands the good news/bad news aspect of small satellite constellations. More remote sensing satellites means more rapid revisit and global coverage. On the downside, however, the enormous volumes of data pose immense challenges to archiving, processing, and analysis.
For SAR data, this challenge is magnified because the information contained within the data requires complex processing methodologies. Individual SAR data files can be quite large. Some analysis workflows require several multi-temporal data sets to achieve accurate results. Much of the valuable information in radar data is contained in the intensity and phase of the radar signal that reflects off the surface and returns to the sensor.
Interpreting the signal intensity and phase difference is vital to exploiting SAR data, but it requires complex processing techniques, which now can be largely automated. Automation of EO imagery is also available, thus when both SAR and EO imagery are processed over the same area, users can benefit from the strength of both data types.
- Among the most critical situations for which temporal monitoring can be applied are natural disasters such as floods, wildfires, and volcanic eruptions. Before the clouds and smoke have cleared, SAR data acquisitions can occur that enable users to produce preliminary products and information for first responders.
- Coherent change detection workflows using SAR data provide the ability to measure terrain elevations to an accuracy of a few millimeters. Continuous monitoring over time may reveal subtle ground changes that could indicate clandestine underground tunneling and excavation or potentially dangerous subsidence around oil wells, urban areas, and mine operations. Shifting of terrain on slopes or near dams can also be detected. In desert environments, SAR can distinguish surface deformations as small as tire marks left behind by vehicles crossing the sand.
- SAR signal intensity can be interpreted to differentiate water, snow, and ice. More importantly, ice thickness can be calculated to determine if a waterway is safe for ship navigation, as Canada and other high-latitude nations have done with airborne and space-borne SAR for decades.
- Satellite-based SAR is also ideal for tracking ships in open seas. Radar signals are sensitive to the roughness of water surfaces, allowing the wakes of vessels to be detected easily. In addition, the geometric shapes of the ship itself deliver a strong signature against the background of the water.
- Many areas of interest to NATO are frequently cloudy such as Arctic shipping routes in Asia.
- SAR’s sensitivity to moisture and guaranteed monitoring capabilities can track agricultural growth to assess crop health and forecast harvest yields—vital aspects of food security monitoring.
The increased availability of SAR data from small satellites and other platforms is positive news for the GEOINT Community. However, availability does not equal accessibility, at least not in terms of the rich information locked in the data. Cloud computing and proven imaging science tools will grow in tandem with data to provide automated answers to quickly gain insights.
Defense and intelligence organizations must take steps now to leverage the processing and analytics capabilities needed to exploit the coming torrent of enormous and complex SAR data sets. Enterprise computing can accelerate processing large data sets, scale with increased data volume, and deploy commonly used SAR analysis tools that can break down the barrier of SAR data and expose the unique information it holds.
Headline Image: Sentinel-1 data, processed with ENVI SARscape, is used to monitor weather impacts such as flooding and wind on rice production in Vietnam’s Mekong river delta region. Image courtesy of Harris.