Radar Instrumentation for Operation Ice Bridge

Image of ICESat data collected over Greenland and Antarctica

Figure 1. A map created from ICESat data demonstrating the extent of ice sheet thinning in Greenland and Antarctica. Credit: British Antarctic Survey/NASA


Jenna Collins, Theresa Stumpf and Ashley Thompson, CReSIS, the University of Kansas, Lawrence, Kansas.

Background

The Earth’s fresh water ice sheets are rapidly – and unexpectedly – changing. Data collected by NASA’s Grace Satellite reveal that the ice sheets of Greenland and Antarctica are losing mass much more quickly than predicted, and subsequent melt from these polar masses is now contributing to sea level rise and altering ocean currents. These changes put millions of people around the world at risk [1]. ICESat data collected over Greenland and Antarctica, shown in Figure 1, illustrate a marked decrease in ice thickness, particularly at the margins, after just four years.

In 2007, the International Panel on Climate Change (IPCC) released its fourth assessment report, meant to serve as a comprehensive and up-to-date reference on the state of climate change and its potential environmental and socio-economic impacts. The report estimated an increase in sea level over a range of 18-59 cm by the end of the century. The IPCC acknowledged the uncertainty of these estimates (in particular the upper bound), which were generated using models that did not account for rising sea levels associated with rapid accelerations of outlet glaciers in Greenland and Antarctica. Consequently, highly-accurate and significantly-improved ice sheet models are urgently needed to address these challenges [2]. If we succeed in developing these models, we can make more precise predictions about sea level rise. If we can accurately predict sea level rise, we can then create realistic and effective climate change policies that will not only help to lessen the effect humans are having on the climate, but also prepare us for the changes that are already coming. Using the latest technologies to take highly accurate measurements of the Earth’s ice sheets, the scientists with NASA’s Operation Ice Bridge are currently working to improve the models polar scientists around the world rely on for predictions about the future of sea level rise. After first collecting information and data from expeditions to Greenland and Antarctica, they will then create detailed maps of both the shape of the bedrock beneath the ice and the surface texture of the ice itself, enabling significantly more accurate predictions about ice sheet melt and its contribution to future sea level rise.

An artist’s rendition of ICESat-I, launched in 2003. Credit: NASA

Figure 2. An artist’s rendition of ICESat-I, launched in 2003. Credit: NASA

Operation Ice Bridge

Operation Ice Bridge, thus far the largest airborne survey of Earth’s polar ice ever flown, is a NASA-sponsored program with an international impact [3]. OIB bridges the gap between ICESat-I, which had been in orbit since 2003, and ICESat-II, which isn’t expected to be launched until around 2015 [4]. ICESat-I, a NASA satellite used to measure the mass balance of the ice sheets, cloud and aerosol heights, and land topography and vegetation characteristics, played a crucial role in monitoring the state of the world’s ice sheets and can be seen in Figure 2. Observation has shown that ice sheet mass is decreasing, particularly around the edges, and that the last five years have seen some of the most drastic changes yet. Additionally, IPCC estimates for Arctic sea ice reduction have unfortunately been shown to be inaccurate – and overly optimistic [5]. According to the National Snow and Ice Data Center, arctic sea ice extent has been decreasing faster than any model in the IPCC AR4 has predicted.

ICESat-I operated by bouncing laser-beams off of the Earth’s surface and recording the received returns, which eventually were used to produce 3-D maps of the ice sheets. Its main instrument, the Geoscience Laser Altimeter System (GLAS) used three lasers to perform this function, the last of which failed in October of 2009 [6], and ICESat-I reentered the Earth’s atmosphere in early August of 2010 [7]. Data gathered from the ICESat-I satellite have already been successfully used to ascertain the thickness of sea ice, as well as to observe regional and inter-annual variability and map ice sheets and glaciers, but there are several years before the launch of the improved ICESat-II.

image of CReSIS graduate student William Blake operates a radar system on board the NASA DC-8 during the fall 2009 OIB field campaign in Antarctica. Credit: CReSIS

Figure 3. CReSIS graduate student William Blake operates a radar system on board the NASA DC-8 during the fall 2009 OIB field campaign in Antarctica. Credit: CReSIS

That’s where Operation Ice Bridge comes in. The OIB program, whose primary goals are to retrieve sea ice thickness, study regional and inter-annual variability, monitor the state of ice sheets, and collect data needed to understand and model why certain parts of ice sheets are changing rapidly, is designed to continue monitoring critical regions of the Arctic and Antarctic until the launch of ICESsat-II. The aircraft used for OIB utilize airborne LIDAR (LIght Detection And Ranging) technology and are also equipped with other instrumentation, such as a suite of radars built by the Center for Remote Sensing of Ice Sheets at the University of Kansas, a gravimeter run by Sanders Geophysics, and a fine-resolution digital mapping camera owned by Cirrus Digital Systems. Data collected using the gravimeter are processed by the University of Washington and the Lamont-Doherty Earth Observatory at Columbia University [8].

The radar suite includes a multi-channel radar depth sounder/imager to measure ice thickness and image the ice-bed interface, which is buried hundreds of meters blow the ice surface, and an ultra-wideband radar to measure both snow thickness over sea ice and internal layers in polar firn. It can do this with a very fine resolution of about five cm. Measurements of snow thickness are required to invert satellite laser and radar altimeter sea ice surface heights (free-board measurements) into sea ice thickness, and fine-resolution mapping of internal layers is essential to producing estimates of annual snow accumulation over ice sheets and interpreting satellite altimeter data sets. NASA, together with university scientists and researchers, has already begun to launch a series of airborne flights using the NASA DC-8 and P-3B aircrafts for this extensive campaign.

Image of The NASA DC-8 aircraft after installation of the radar package. Credit: CReSIS

Figure 4. The NASA DC-8 aircraft after installation of the radar package. Credit: CReSIS

Official field research for Operation Ice Bridge began in the spring of 2009, when the first campaign with LIDARs and radars on the P-3 aircraft was carried out over the Greenland ice sheet and Arctic sea ice. A second deployment with a more complete instrumentation package on the DC-8 to Antarctica followed in the fall of 2009. This campaign flew over different parts of the continent, focusing on the ice sheet, glaciers, and sea ice in West Antarctica. A total of 21 flight missions (four more than the projected 17) covered almost 100,000 miles from October to November of 2009. These missions were conducted from Punta Arenas, Chile, and lasted up to 12 hours per day [9]. The missions were very successful and resulted in precise measurements of ice surface elevation and detailed maps of glacier ice and snow cover thicknesses. Some of these data have already been processed and used to generate data products, which are available through NSIDC. Operation Ice Bridge data summaries can be found here: http://nsidc.org/data/icebridge/data_summaries.html.

The spring 2010 campaign in Greenland included flights over Greenland and Arctic sea ice using both NASA DC-8 and P-3B aircraft. These flights, which took place between March and May of 2010, collected detailed measurements over areas where the glaciers and the ice sheets have been experiencing rapid changes [3]. A large 15-element array was developed to sound fast-flowing glaciers with very rough surfaces and to generate 3-D topography of the ice bed covered with as much as 3 km of ice in the interior of the ice sheet. A digital imaging system collects the data used to create 3-D images of ice surface.

Image of an Echogram from the depth sounder radar on the DC-8 aircraft. Data were collected from an altitude of approximately 9,000 m to a maximum depth of approximately 3 km. This is the first successful depth-sounding using a jet aircraft at an altitude of 9,000 m. Credit: CReSIS

Figure 5. Echogram from the depth sounder radar on the DC-8 aircraft. Data were collected from an altitude of approximately 9,000 m to a maximum depth of approximately 3 km. This is the first successful depth-sounding using a jet aircraft at an altitude of 9,000 m. Credit: CReSIS

The Center for Remote Sensing of Ice Sheets

The radar instrumentation package used to collect data during both the fall 2009 campaign and the spring 2010 campaign was developed by the Center for Remote Sensing of Ice Sheets (CReSIS) at the University of Kansas. CReSIS, an NSF Science and Technology Center (STC), was established in June 2005 for the purpose of studying the present and future contributions of the Greenland and Antarctic ice sheets to sea level rise. The NASA P-3B and DC-8 aircraft are used here as platforms for data collection. For flights on NASA’s DC-8 aircraft (seen in Figure 4 below), the ICE Bridge radar instrumentation package consisted of three radar systems: a Ku-band radar altimeter, an ultra-wideband microwave radar called the Snow Radar, and a Multichannel Coherent Depth Sounder/Imager (MCoRDS/I). Initial results from the MCoRDS/I radar, collected from an altitude of about 9,000 m above the surface from the DC-8 aircraft in November 2010, can be seen in Figure 5. The maximum ice thickness is about 3 km for the flight line segment shown in the image, and internal layers can be seen to a depth of about 2,700 m. The layers below this depth are masked by surface clutter and range sidelobes of strong surface returns. The results shown in the figure are generated with a quick-look processor in about 24 hours, after returning to Punta Arenas, Chile. Additional processing of data with optimized space and time adaptive filters will reduce clutter and range sidelobes and bring out internal layers below the depth of 2,700 m. For flights on the P-3, the instrumentation radar suite, summarized in Table 1, also included the UHF Accumulation Radar and provided Operation Ice Bridge with improved surface-to-bedrock profiling capability; sample results from the large antenna array on the P-3 can be seen in Figure 7.

Table 1. Customized Radar Suite
Instrument Measurement Frequency Platform Resolution
MCoRDS/I Bed Topography,
Bed Imaging,
Internal Layering
195 MHz DC-8
P-3
4 m
Accumulation Radar Internal Layering 750 MHz P-3 40 cm
Snow Radar Snow Thickness,
Internal Layering,
Topography
2-7 GHz DC-8
P-3
5 cm
Ku-Band Altimeter Topography 14 GHz DC-8
P-3
5 cm

Once equipped, the aircraft are flown over glaciers, ice sheets, and sea ice to collect data. Finally, the data are processed using advanced algorithms and distributed to the scientific community through NSIDC.

The Customized CReSIS Radar Suite

The radar instrumentation outlined in Table 1 can be used to profile an ice sheet from the surface to the bedrock and to generate 3-D topography. The major innovations include operating a radar depth sounder/imager for the first time on a four-engine turboprop-powered airliner and the use of a large 15-element antenna array for the radar sounder/imager.

The 15-element antenna array is one of the largest arrays ever flown on the NASA P-3 aircraft for civilian applications, and it was designed, built and tested in only four months. This design required strong interaction between CReSIS electrical and aerospace engineers to create an array that had desirable radiation properties and was also aerodynamically feasible. The array boasts several desirable properties in this type of remote sensing application, including improved signal-to-clutter ratio, increased directivity of the transmission beam, and cross-track array progressing algorithms.

Image of The rough surface of the ice in Antarctica as seen from the DC-8 aircraft in fall 2009. Credit: CReSIS

Figure 6. The rough surface of the ice in Antarctica as seen from the DC-8 aircraft in fall 2009. Credit: CReSIS

This array is crucial to the OIB mission for two reasons. First, in order to attain more accurate future sea level rise estimates, data, particularly on the bed topography and basal conditions, are needed to understand the processes causing rapid changes and improve ice sheet models. Ice thickness data from which bed topography is determined and basal properties for areas undergoing rapid changes are, unfortunately, difficult to measure for a number of reasons. Outlet glaciers are characterized by two glaciological features that typically pose a challenge to airborne radar surveys: rough surface topography (seen in Figure 6) and lossier ice than that in the ice sheet interior. Ice at temperatures near melting has different electromagnetic properties than cold ice and attenuates a propagating signal more than dry, colder ice. Rough surface topography scatters radar transmissions back to its antennas. The scattered signal, referred to as surface clutter, in conjunction with the weakened signal from the lossy ice, makes it difficult for radars to detect bed echoes. Also, outlet glaciers are enclosed by steep valley walls (channels). Weak bed echoes are also often masked by signals reflected by the valley walls, and these reflected signals can be misinterpreted as ice-bed echoes.

Table 2. Geophysical Challenges to Radar Sensitivity and 15-Element Array Solutions
Outlet Glacier Geophysical Feature Challenge to Radar Sensitivity Increase in Measurement Accuracy by Improvement of Radar Sensitivity
Narrow Channels Backscatter off sides of the channel, or clutter, may be brighter than signal returning from the bed. Clutter contributes to noise energy received by the radar. Clutter degrades the signal-to-noise ratio (SNR) and radar sensitivity may be compromised when the signal energy level is low relative to the noise. In this case, radar is unable to detect bed return. Narrow physical antenna beam by increasing the number of elements in the array to focus beam on desired target and decrease illumination of undesired targets.

Apply processing techniques to digitally steer antenna beam in the data to illuminate desired target and/or reject the energy from the clutter.

Rough Surface Topography Surface clutter that scatters back to radar and compromises radar sensitivity to weak echoes from the bed. Improve signal to noise ratio by increasing number of elements in array.

Apply processing techniques to digitally steer antenna beam in the data to illuminate desired target and/or reject the energy contributed by surface clutter.

Temperate Ice Temperate ice near glacier margins contains non-ice inclusions (like rock or sediment) and channels of melt water that introduce larger amounts of volume clutter, which is absent from the cold, dry ice in the interior. Like surface clutter, volume clutter scatters back to radar, degrading the signal to noise ratio and masking returns from the bedrock. The effect of volume scatter can be reduced by reducing antenna beam width in the cross-track direction with the large antenna array in conjunction with along-track synthesized array.

Second, synthetic Aperture Radar (SAR) processing techniques can be used to reduce along-track surface clutter, but we need a large cross-track array to reduce cross-track surface clutter. The angular resolution in the cross-direction is inversely proportional to the array size. For this reason, CReSIS designed and developed a large cross-track array to perform and apply advanced array processing techniques to improve angular resolution and therefore reduce clutter. The P-3 provided the ideal platform for the large array because of its ability to fly at low altitudes over an extended period of time, as well as its bomb bay underbelly, on which seven of the antennas were mounted.

Data from each element of the array are digitized and stored for optimum processing to extract weak bed echoes buried in clutter and noise. Advanced array processing algorithms such as Minimum Variance Distortionless Response (MVDR) are applied to the data to suppress surface clutter [10]. Each mission results in about 2 terabytes of data from the radar system. These data are backed up and processed to produce quick-look data products in 24-48 hours after a mission is completed.

Figure 7 shows preliminary results from data collected with the large 15-element array over the Greenland ice sheet during May 2010. The top left image is an echogram that clearly shows bed returns with a signal-to-noise ratio of more than 40 dB for about 2.5 km thick ice. The top middle figure is a radar echogram generated from data collected with the snow radar over the ice sheet. A range window of 30 m extends 15 m above the surface and 15 m below into the firn. These data are collected with the aircraft flying at an altitude of 500 m above the surface. However, results are displayed over a small range window to clearly show layers. The image at the top right is a radar echogram generated from data collected with the snow radar over sea ice. Again, a small range window featuring the snow-air interface is shown in the image. The bottom left image shows sample results obtained with the accumulation radar over a range window of 300 m. The internal layers can be mapped with this radar to a depth of about 150 m. The bottom right image shows sample Ku-band radar results over the ice sheet.

Summary

Image of In-the-field data collected as part of the NASA OIB campaign, May 2010. Each echogram was taken aboard the P-3 aircraft; antenna configuration is shown in the lower central image. More advanced processing of MCoRDS data will be available in the next two months at www.cresis.ku.edu/data. Snow radar data provided by (11).

Figure 7. In-the-field data collected as part of the NASA OIB campaign, May 2010. Each echogram was taken aboard the P-3 aircraft; antenna configuration is shown in the lower central image. More advanced processing of MCoRDS data will be available in the next two months at www.cresis.ku.edu/data. Snow radar data provided by (11).

Though much work remains to be done, without the measurements and analyses performed by CReSIS and others, the future of sea level rise will remain dangerously uncertain. This uncertainty presents an unimaginable risk to not only the millions of people living in coastal regions, but also to the nearby populations that would be charged with resolving the issue of “climate refugees”. Sea level rise is not an isolated issue, and it will not disappear with time. NASA’s Operation Ice Bridge is currently poised to continue monitoring polar regions using airborne LIDARs between ICESat-I and ICESat-II and collecting critical data needed to improve ice sheet models; this data can only be collected on airborne platforms. The prediction of the future state of the Earth’s cryosphere depends on these measurements and the models that result.

Acknowledgements

The authors would like to thank the CReSIS OIB team for their support in providing data and figures for this article. We would also like to acknowledge NASA for their contribution of the P-3 and DC-8 aircrafts.

References

1. Douglas, Bruce C., Michael S. Kearney, and Stephen P. Leatherman. Sea Level Rise: History and Consequences. San Diego: Academic, 2001.

2. Alley, Richard B., Mark Fahnestock, and Ian Joughin. “Understanding Glacier Flow in Changing Times.” Science 332 (2008): 1061-062. Print.

3. “IceBridge.” NASA – IceBridge Mission. NASA. Web. 08 Sept. 2010. <http://www.nasa.gov/mission_pages/icebridge/index.html>.

4. Martin, Seeley. “The Road to Antarctic Science.” Web log post. Operation Ice Bridge. NASA Blogs, 28 Sept. 2009. Web. 31 Aug. 2010. <http://blogs.nasa.gov/>.

5. Markus, Thorsten. “The ICESat-2 Mission and Its Science.” CReSIS All-Hands Presentation. Nichols Hall, Lawrence. 26 Aug. 2010. Lecture.

6. Clark, Stephen. “ICESat Mission Complete after Seven Years in Orbit.” Spaceflight Now. 25 Feb. 2010. Web. 26 Aug. 2010. <http://spaceflightnow.com/news/n1002/25icesat/>.

7. “ICESat: Cryospheric Sciences Branch, Code 614.1.” NASA: ICESat. Web. 16 Aug. 2010. <http://icesat.gsfc.nasa.gov/icesat/>.

8. Watts, Anthony. “Operation Ice Bridge in Antarctica” Web log post. Watts Up With That? Blogpress, 8 Oct. 2009. Web. 31 Aug. 2010. <http://wattsupwiththat.com/>.

9. NASA. Goddard Space Flight Center. Critical Polar Data Flows Briskly to Researchers. IceBridge. NASA, 1 Sept. 2010. Web. 8 Sept. 2010. <http://www.nasa.gov/mission_pages/icebridge/news/10-075.html>.

10. Paden, J., T. Akins, D. Dunson, C. Allen, and P. Gogineni, “Ice-sheet bed 3-D tomography,” Journal of Glaciology, vol. 56, no. 195, pp. 3-11, Jan. 2010.

11. Panzer, B., C. Leuschen, A. Patel, T. Markus, and P. Gogineni, “Ultra-wideband radar measurements of snow thickness over sea ice,” 2010 IEEE IGARSS Conference, Honolulu, HI, 29 July 2010.

Topic: