March 2014 Meeting - Denver, Colorado

At the APS March meeting FPS sponsored a session titled Keyhole to the World: Public Access to Satellite Data for Environmental, Security, and Social Ends. In this session we had earth scientists, environmental watchdogs, and nuclear arms control analysts who use publicly available satellite imagery in their work. Jeff Dozier from U.C. Santa Barbara presented “40 years of Landsat images: What we learned about science and politics,” in which he described the substance of a recent National Research Council study on the future of Landsat. John Amos, president and founder of SkyTruth, described the work of his independent, environmental nonprofit organization in “Bringing the Crowd to Environmental Investigation and Monitoring.” Irmgard Niemeyer gave a talk titled “Nuclear Verification from Space? Satellite Imagery in Support of Non-Proliferation and Arms Control,” which explained her research group’s work at Forschungszentrum Jülich GmbH. And Lawrence Friedl gave a survey of NASA’s Earth observation satellite capabilities in “Earth Science Serving Society: Using NASA Earth-observing Satellites for Policy, Management, and Capacity Building.” At the end of the session, the group held a panel discussion led by the session chair, Micah Lowenthal.

In addition to asking them to tell us about their work, we also asked them to provoke the audience into thinking about the problems the speakers work on by giving us some unsolved challenges in their work, a kind of wish list. We asked them to do this because physicists are a smart, knowledgeable, and creative bunch and it would be a wasted opportunity not to seek input from the physicists.

Dr. Dozier described the unique and extensive data set that Landsat provides dating back to 1972. The Landsat images have moderate spatial resolution (pixel size 10-100m), which provides useful environmental and other data while still being able to provide coverage of the whole earth. Higher resolution imagery takes much longer to cover the same area. The several Landsat sensors cover many of the wavelengths from blue to infrared that are transmissible through the atmosphere, and one in a range that is not. By fusing data from different sensors, analysts can obtain enhanced images to track regional carbon emissions, land-use patterns, and other applications. In 2008, the U.S. government made the Landsat data available for free and the number of images accessed per year rose from very few to approximately 2.6 million. But the future of the Landsat mission is in some doubt as Congress does not wish to spend the $1B it typically costs to design, construct, and launch new satellites for the mission (the Senate wished to limit costs to $650M and the House wished to spend nothing). Dozier noted several options to create a less costly, more robust program: Acquire satellites differently (committing to a series at once); integrate with other data sources; increase the swath width; or create a constellation of smaller satellites. Speaking for himself, Dozier underscored the value of the satellites and the importance of maintaining the Landsat mission.

John Amos began with the early history of earth observation from space, displaying the first photograph of earth taken from 65 miles above the surface by a tumbling V2 rocket captured from Germany shortly after the end of the war and launched by the United States. He then jumped to the current state of satellite affairs with over 1000 operational satellites giving us useful and unique information about the behavior of oceans and clouds, and how people are changing the landscapes, habitats and ecosystems through processes like mining, drilling, deforestation, pollution, and climate change. In 2001, he started the nonprofit SkyTruth to use satellite imagery and data to promote awareness about environmental issues. Within the first week of the BP/Deepwater Horizon blowout, SkyTruth used publicly available images to calculate that the spill was at least 20 times larger than official BP and Coast Guard estimates at the time. Later, the U.S. government determined that SkyTruth’s estimate was low by a factor of 3. Amos and his colleagues have combined satellite and ground-based radio-collar tracking data to assess disruption of Mule Deer habitat by gas development in Wyoming and have used satellite imagery to map flaring of natural gas at oil wells worldwide. They have assessed how much of the Appalachian landscape has been directly impacted by mountaintop removal coal mining, which supported evaluations of whether a tipping point has been reached in the environment’s capacity to accommodate the practice. Satellite data will increase dramatically in the coming years, which affords many opportunities to utilize those data in new ways. To take advantage of those opportunities, Amos argued that we need new techniques, not only in automated image analysis and information delivery, but in harnessing the power of the crowd to help us continuously monitor environmental changes all over the Earth. Amos described SkyTruth’s use of crowdsourcing to check on fracking in Pennsylvania. He closed by noting that automation and crowdsourcing are not alternative approaches, but reinforcing approaches for image analysis, because one ultimately needs human eyes to evaluate images and the automation can allow people to focus only on the ones that need human evaluation.

Irmgard Niemeyer described her work developing tools that can use satellite imagery to monitor and verify current and proposed nuclear nonproliferation and arms control agreements, such as the the Treaty on the Non-Proliferation of Nuclear Weapons (NPT), the Comprehensive Nuclear Test Ban Treaty (CTBT), or a Fissile Material Cut-Off Treaty (FMCT). If onsite verification is not feasible, then satellite imagery provides one of the few opportunities to gather direct (and indirect) evidence. Niemeyer’s group has partnered with other groups on projects for the European Union like G-SEXTANT, which uses Earth Observation (EO) products to address needs associated with humanitarian crises, natural resources, land conflict situation awareness, monitoring of nuclear sites and activities, illicit crops; and border surveillance. Automated change detection of optical imagery enables tools to identify new facilities and using synthetic aperture radar data, they can construct 3-D models of sites and facilities.

Satellite imagery analysis currently is not included in the CTBT verification regime, although it is considered as an additional technology whose verification potential should be examined (Article IV, paragraph 11). It is anticipated that satellite imagery would be used in the preparation of on-site inspections and when combined with in situ measurements, satellite imagery can contribute to better verification. Niemeyer closed with a wish list, noting that thermal infrared and hyperspectral sensors with improved spatial and temporal resolution; very high resolution optical sensors with enhanced spectral resolution, and high spatial resolution SAR sensors offering polarimetric data would all be valuable additions.

Lawrence Friedl described NASA’s Earth Science Division, which aims to characterize, understand, and improve predictions of the Earth, a complex, inter-related system of environmental phenomena, human activities, and ecological effects. As noted by other speakers, the sensors on the satellites utilize a broad range of the electromagnetic spectrum. The 14 current NASA satellite missions provide multiple approaches to measure phenomena and yield valuable analyses concerning air quality, water resources, agriculture, climate, weather, disasters, energy, and ocean phenomena. GRACE, a pair of satellites that uses ultra sensitive distance measurements between the two to observe gravitational fluctuations, measures groundwater abundance over large regions. SERVIR is a NASA-USAID partnership to improve environmental management and disaster resilience in countries that have little capacity for environmental data observation and forecasting. SERVIR linked JASON-2 altimetry data to flood forecasts, increasing the flood prediction times from three days to eight days in northern India and Bangladesh, which can save thousands of lives. The Terra and Aqua satellites possess sensors to identify thermal anomalies and active fire locations in remote African areas (among other places), and notices are sent via SMS to fire managers in areas. Also in Africa, satellite data are being used to identify the locations with the greatest risk of malaria transmission. NASA continues to seek new tools and better ways to use the increasingly extensive data the satellites generate.

These contributions have not been peer-refereed. They represent solely the view(s) of the author(s) and not necessarily the view of APS.