Ice-Spy: Declassified Satellite Images Measure Glacial Loss

U.S. spy pilot, Gary Powers (RIAN/Creative Commons).
U.S. spy pilot, Gary Powers (RIAN/Creative Commons).

Since the 1960s, images from spy satellites have been replacing the use of planes for reconnaissance intelligence missions. Making the transition from planes to satellites was prompted by an infamous U-2 incident during the Cold War when U.S. pilot Francis Gary Powers’ U-2 spy plane was shot down in Soviet air space. Five days later, after considerable embarrassment and controversy, President Eisenhower approved the first launch of an intelligence satellite, part of a new scientific electronic intelligence system termed ELINT. Today, declassified images from satellites have resurfaced to support scientific research on glaciers and climate change.

Scientists from Columbia University and the University of Utah created 3-D images of glaciers across the Himalayas, and Bhutan specifically, by using satellite imagery to track glacial retreat related to climate change. Joshua Maurer et al. published the results of their Bhutan study in The Cryosphere to help fill in the gaps of “a severe lack of field data” for Eastern Himalayan glaciers.

Looking down the valley from a glacier the team visited in Bhutan in 2012 (Source: Joshua Maurer).
Looking down the valley from a glacier the team visited in Bhutan in 2012 (Source: Joshua Maurer).

Being able to understand and quantify ice loss trends in isolated mountain areas like Bhutan requires physical measurements that are currently not available due to complex politics and rugged terrain. Luckily, the scientists found an alternative route to reach their measurement goals by comparing declassified spy satellite images from 1974 with images taken in 2006 using the ASTER, Advanced Spaceborne Thermal Emission and Reflection Radiometer, a spaceborne imaging instrument aboard NASA’s earth-observing Terra satellite.

Bhutan has hundreds of glaciers and glacial lakes. Physical data collection can be a daunting process in such a region considering the vast quantity of glaciers in combination with freezing weather conditions and high winds. The lead researcher of the Bhutan study, Joshua Maurer from Columbia University, experienced firsthand the logistical challenges associated with directly measuring changes in glacial ice density when conducting research on glacial change in the remote and high-altitude regions of Bhutan. Inspired by this difficult experience, Maurer collaborated with other scientists from the University of Utah to find alternative methods for quantifying trends in glacial ice density.

Camera system for a Discoverer-Corona spy satellite (Tim Evanson/Creative Commons).
Camera system for a Discoverer-Corona spy satellite (Tim Evanson/Creative Commons).

Maurer and the team of researchers devised a strategy to use declassified satellite images to collect data by a process of photogrammetry, the use of photographs to survey and measure distances. More than 800,000 images from the CORONA Satellite program, taken in the 1970s and 1980s, have been sent to the U.S. Geological Survey from the Central Intelligence Agency (CIA), and made available to the public.

Several advanced mathematical tools are necessary for making measurements from raw image files. For this particular study, the team used the declassified photos from the 1970s to track changes in glacial ice coverage over time when compared to more recent images from the Hexagon Imagery Program database taken by the Swiss-based Leica Geosystems’ airborne sensors in 2006. Once a timeline was created from the pictures, measurements were made using NASA’s space tool ASTER. This method, Maurer argues, is the solution for measuring massive amounts of hard-to-access data.

Landsat 8 satellite image, with studied glaciers outlined in white (Source: Joshua Maurer).
Landsat 8 satellite image, with studied glaciers outlined in white (Source: Joshua Maurer).

But making precise measurements integrating several sets of images from different periods of time is no simple task. Pixel blocks, minute areas of illuminations from which images are composed, were processed to correspond with regions designated on the film. The blocks of pixels were then selected to maximize coverage of glaciers and avoid regions with cloud cover. Computer-generated algorithms transform these blocks of image into measurements using automated point detectors and descriptors.

Images from the declassified satellite database may suffer from a lack of clarity, so it was also important for the researchers to address these issues. For example, debris-covered glaciers are difficult to distinguish from surrounding terrain using visible imagery only. Furthermore, loud cover and poor radiometric sensing data in remote areas can prevent complete observation. In order to address challenges like these, images were analysed by a computer and then manually edited to more accurately match glacial extent in the year that the image was taken. In order to prevent statistical errors, the research team focused on a select sample size of glaciers representative of the area being studied.

Satellite image analysis like that performed in Bhutan has become increasingly important in the study of climate change. In terms of glaciers, these analyses have proven valuable to scientists in reaching otherwise hard to access data. The main findings of the study were that glacial retreat in the last fifty years is significantly contributing to the creation of glacial lakes in the East Himalayan region and associated flood outbursts. A glacial lake outburst flood is a type of flood that occurs when the dam containing a glacial lake fails due to a buildup of water pressure. Bhutan has low lying river planes that are vulnerable to such floods, so measuring ice loss can help scientists identify which dams are at risk of bursting. This can further help policy makers take appropriate action to mitigate potential disaster.

Declassified Corona spy satellite image from the year 1974 showing the glaciers in Bhutan (Source:Joshua Maurer).
Declassified Corona spy satellite image from the year 1974 showing the glaciers in Bhutan (Source: Joshua Maurer).

Following the successful completion of the Bhutan study, Maurer and his team were granted additional funding from a NASA Earth and Space Science fellowship to expand the same methodology to other regions of the Himalayas. Understanding ice loss is important, and the effort to overcome logistical barriers is worthwhile.

“Ice loss will impact hydropower, agriculture, and ecosystems in the region,” Maurer told GlacierHub. Understanding the glacial ice balance in the Himalayan region and the rates of ice loss assists adaption plans for building strategic dams and reservoirs for seasonal water storage. These actions could result in more people being better off, more people receiving reliable electricity, and a reduced risk of moraine dam outbursts.

While observation of changing trends in glacier mass may not be complete, the information that is available due to declassified spy satellite imagery positively contributes to the Himalayan people’s capabilities regarding future impacts linked to ice loss, according to Maurer et al. Overall, results from spy satellite images have enhanced the understanding of potential glacier contribution to sea-level rise, impacts on water resources, and hazard potential for high mountain regions and downstream populations in Asia.

 

 

Please follow, share and like us:
error

Tracking Glaciers From Space: GLIMS

Picture of GLIMS book coverIn 1994, an international group of scientists came together to form GLIMS (Global Land Ice Measurements from Space), a worldwide initiative to monitor and study glaciers using satellite data. For at least one hundred years, scientists had primarily used traditional field measurements to track glacier dynamics, but field data are by necessity limited in scope, and can be expensive and laborious to obtain.

The GLIMS team ultimately chose to use an imaging system called Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), jointly managed by NASA and Japan, for their glacier measurements. ASTER is installed aboard Terra, the flagship satellite of NASA’s Earth Observing System (EOS), which was launched in December 1999. ASTER data can be used to map land surface temperature, reflectance, and elevation, which allows the scientists to distinguish between glacier ice and snow and to measure changes in glacier volume.

Glacier retreat lines at the Brøggerhalvøya Glacier between 1936 and 2007. Chapter, 10, p. 234, Figure 10.3.
Glacier retreat lines at the Brøggerhalvøya Glacier between 1936 and 2007. Chapter, 10, p. 234, Figure 10.3.

Using digital images and data provided by ASTER, GLIMS created an up-to-the-minute database of the world’s glaciers, which includes ID, name, cross-references, and analysis of the state and dynamics of individual glaciers. In August 2014, GLIMS published their findings in book form: Global Land Ice Measurements from Space compiles these glacier profiles, provides a review of analysis methodologies for measuring changes in glacier volume, and offers predictions for future glacier change as well as some interpretations of potential impacts for policymakers in the context of climate change. The GLIMS scientists provide firm evidence that glaciers are shrinking worldwide, and they believe the cause is global warming.

The GLIMS book offers a basic theoretical background in glacier monitoring and mapping as well as remote sensing techniques. It also discusses measurements of glacier thinning from digital elevation models (DEMs), and calculation of surface flow velocities from satellite images. DEMs can provide specific data for every pixel in a satellite image, with a margin of error at 0.5m/year. Although cloud cover can interfere with accurate satellite data on glaciers, scientists are able to identify and discard this faulty data.

As described in the book, GLIMS scientists Siri Jodha Singh Khalsa and his colleagues have been able to assess the mass balance of alpine mountain glaciers by comparing historical topographic maps and DEMs derived from ASTER. For instance, they built a model and limited the error in the computation of mass balance from field measurements of China’s Sarytor glacier to less than 150mm/year.

Tropical glaciers in the northern Andes. Chapter 26, page 614, Figure 26.1.
Tropical glaciers in the northern Andes. Chapter 26, page 614, Figure 26.1.

Similarly, using techniques established by Dr. Todd Albert,who is also a member of GLIMS, a set of images of the Quelccaya Ice Cap spanning four decades was analyzed to create a history of ice surface area. Overall, Albert found that the ice cap has retreated from 58.9 km2 in 1975 to 40.8 km2 in 2010, with a loss of surface area of 31%. This history matches what has been observed in the field by glaciologists Lonnie Thompson and Henry Brecher since the 1970s.

Thanks to GLIMS, the rate of glacier melting can be measured and documented more precisely, providing readers with potential evidence of climate change. The GLIMS data provides solid support for future scientific research and planning in the face of climate change.

For other stories on the measurement of glaciers, look here.

Please follow, share and like us:
error