This paper presents the motivation for, and initial results from, the Multiple Altimeter Beam Experimental lidar (MABEL) instrument. The MABEL instrument provides a new capability for airborne altimetry measurements and serves as a prototype and simulator for the upcoming NASA second-generation Ice, Cloud, and Land Elevation Satellite (ICESat-2) mission. Designed to be highly flexible in measurement capability, MABEL serves as both a demonstration of measurement capability and a science tool for cryospheric and biospheric remote sensing. It is important to document the instrument specifications and essential background information to provide a suitable reference for the detailed MABEL results and science investigation publications that will be forthcoming.
Quantifying ongoing ice sheet and sea level changes remains an earth science priority. The second-generation Ice, Cloud, and Land Elevation Satellite (ICESat-2) mission is the second-generation laser altimeter designed for cryospheric science to continue measuring elevation changes of the Greenland and Antarctic ice sheets, to document changes in sea ice thickness distribution, and to provide crucial information about the mass and energy balance between ice sheets, sea ice, and the climate system.
Following recommendations from the National Research Council for an ICESat follow-on mission (National Research Council 2007), the National Aeronautics and Space Administration (NASA) initiated the ICESat-2 mission, which is now under development for launch in 2016. The primary goals of the ICESat-2 mission are to continue measurements of sea ice thickness and volume change and ice sheet elevation changes at scales from outlet glaciers (~100 km2) to the entire ice sheet (~106 km2; Abdalati et al. 2010). Unlike the original ICESat, which used a laser operating at 40 Hz and an analog detection system to record reflected laser energy in the infrared as a waveform (Abshire et al. 2005; Schutz et al. 2005), ICESat-2 will employ a photon-counting approach to obtain better measurement sensitivity with overall lower resource requirements on the satellite platform. The current concept uses a high-repetition-rate (10 kHz), low pulse energy laser in conjunction with single-photon-sensitive detectors to measure the range to the Earth’s surface using 532-nm light. The ICESat-2 concept also makes use of multiple beams to allow unambiguous separation of ice sheet slope and elevation change (Zwally et al. 2011), while the higher repetition rate generates dense along-track sampling. Combining ICESat-2 data with existing and forthcoming altimetry data will yield a 15-plus-year record of elevation change.
Given the substantially different design of ICESat-2 compared with ICESat, the ICESat-2 project elected to develop an airborne ICESat-2 simulator to verify ICESat-2 instrument models, to simulate ICESat-2 data over the cryospheric targets that are the focus of ICESat-2, and to generate ICESat-2-like data for algorithm development. Consequently, the Multiple Altimeter Beam Experimental lidar (MABEL) was developed to meet the project needs. This paper outlines the science objectives for the ICESat-2 mission, the current measurement concept for ICESat-2, and the instrument concept and preliminary data from MABEL. Subsequent papers detailing MABEL science results and ICESat-2 comparisons will be able to reference this paper as background on the MABEL instrument.
2. ICESat-2 science objectives
The primary goals of the ICESat-2 mission are consistent with the direction outlined by the National Research Council (2007), including deploying a spaceborne sensor designed to collect altimetry measurements of the earth’s surface, optimized to measure ice sheet elevation change, sea ice thickness, and global vegetation biomass. The ICESat-2 ad hoc science team developed science objectives in each of these measurement areas, which include the following:
Quantify polar ice sheet contributions to current and recent sea level change and the linkages to climate conditions.
Quantify regional signatures of ice sheet changes to assess the mechanisms driving those changes and to improve predictive ice sheet models; this includes quantifying the regional evolution of ice sheet change, such as how changes at outlet glacier termini propagate inward.
Estimate sea ice thickness from freeboard measurements to examine ice/ocean/atmosphere exchanges of energy, mass, and moisture.
Measure vegetation canopy height as a basis for estimating large-scale biomass and biomass change.
In particular, ICESat-2 has a requirement to produce an ice surface elevation product that enables the determination of whole ice sheet elevation changes to 0.4 cm yr−1 accuracy on an annual basis. This is a demanding requirement that drives pointing knowledge, measurement signal-to-noise ratio, and orbital considerations.
3. ICESat-2 measurement concept
The measurement approach selected for ICESat-2 is a photon-counting system with a high-repetition-rate, low pulse energy laser and single-photon-sensitive detectors. The outgoing laser beam is split into six transmitted beams to enable measurement of cross-track surface slope.
The ICESat-2 instrument is designed for operation at 532 nm. Arguments for science benefits of 1064 versus 532 nm are manifold, with pros and cons for each wavelength. However, the reality is that spaceflight qualifiable solid state (e.g., not analog detection) single-photon-sensitive detectors at 532 nm are readily obtainable, whereas none are currently available in the 1-μm wavelength region.
The concept of photon-counting lidar is not new, having been employed for many years in atmospheric (e.g., cloud and aerosol) profiling (McGill et al. 2002). The photon-counting concept is, however, a relatively new concept for altimetry. Traditionally, altimeters have used analog detection with fast digitization of the return signal. The photon-counting concept permits use of less powerful lasers, which lessens the thermal and power requirements for a satellite-based instrument. However, to avoid being overwhelmed by solar background at 532 nm requires a small instrument field of view to limit the solar signal. In the case of ICESat-2, the diameter of each laser footprint on the surface is ~10 m (from a nominal 500-km orbit) with a receiver field of view of ~40-m diameter. For ICESat-2, the plan is to use six transmit beams in three pairs, in a rectangular configuration. A yaw angle from the satellite track then sets the specific spacing of the footprints.
Once a decision was made to pursue the photon-counting approach for ICESat-2, the need for an airborne demonstrator instrument became apparent. Although not required to be an exact scaled-down duplicate of the ICESat-2 instrument, the demonstrator instrument would have to be configured to permit the evaluation of science and engineering trade spaces. The following were the primary flexibilities required in the instrument design:
Ability to vary the laser repetition rate. This is necessary to permit the evaluation of multiple-pulse ambiguity. From the nominal 500-km orbit, ICESat-2 will have ~35 pulses in flight at any one time. At the 10-kHz repetition rate of ICESat-2, signal photons from the surface as well as clouds at 15-km height arrive at ICESat-2 at the same time, leading to a range ambiguity of 15 km. The MABEL laser is capable of operation in discrete steps from 5 kHz (no ambiguity from 20-km flight altitude) up to 25 kHz.
Ability to time tag every photon event. This is the very heart of the ICESat-2 concept, so demonstration is critical.
Ability to oversample the expected spaceborne resolution. Demonstration and validation measurements should always be obtained at a higher resolution than at which the instrument is being validated, and the inherent difference in satellite and aircraft speeds satisfies this objective. Oversampling is also a critical component of developing ICESat-2-like data from MABEL data.
Ability to vary the viewing geometry/footprint pattern. This flexibility permits the determination of optimal geometry and validation of the spaceborne instrument design. This also allows great flexibility for future validation of the spaceborne measurements.
Ability to operate at both 532 and 1064 nm. Although ICESat-2 is a single wavelength (532 nm), initially there was a desire to use either 1064 nm or both wavelengths. Even so, operating the airborne instrument at both wavelengths permits better validation measurements and allows for quantification of differences in reflectivity and water or snow penetration, for example. Dual-wavelength data will also aid in relating ICESat data (at 1064 nm) to ICESat-2 data (at 532 nm).
Ability to vary the energy level for each footprint. This capability is needed to explore the dynamic range trade space and to inform the design of the satellite instrument, and to verify the link model for the satellite instrument.
Ability to capture data for use in algorithm development and testing. This means acquiring data with clouds and other environmental effects that will be encountered by the satellite-based instrument. Of all the instrument design goals, acquiring data for algorithm development is perhaps the most important contribution MABEL has made to the ICESat-2 development process.
Designing one instrument capable of exploring the multiple trade spaces was challenging. Based on previous experience and desiring to mimic the satellite observation as closely as possible led to the selection of the NASA ER-2 aircraft as the target platform. The ER-2 flies at ~65 000 ft (~20 km) and thus is above ~95% of the earth’s atmosphere. More importantly, flying at high altitude allows for data collection over a variety of cloud types, and allows for examination of the range ambiguity that results from having multiple pulses simultaneously propagating through the atmosphere. These are important considerations for the ICESat-2 mission, and they cannot be adequately addressed by flying instruments at low altitudes. Developing an instrument for the ER-2 platform also permits an easier transition to the unmanned Global Hawk, or most any other platform, for future science and validation applications.
MABEL is not intended as, and need not be, an exact duplicate of the ICESat-2 instrument. The primary purpose of MABEL is to verify the measurement concept and to provide data for algorithm development, rather than attempting to achieve the best possible range precision to the surface. Consequently, the aircraft instrument does not need to have the same level of geolocation, pointing accuracy, or range precision as the spaceborne instrument. MABEL is meant to obtain data similar to ICESat-2 but with flexibility to explore science and engineering trade spaces.
4. MABEL instrument design
The MABEL instrument uses a high-repetition-rate pulsed laser fabricated by Fibertek, Inc. The laser repetition rate is variable from 5 to 25 kHz, although typical operations will use 10 kHz to mimic the on-orbit behavior of ICESat-2. The laser pulse length is 2 ns. If the laser is operated in 10-kHz mode, then a pulse is emitted every 2 cm along track given the nominal speed of 200 m s−1 of the ER-2.
The laser generates both 1064- and 532-nm output. However, because only one laser is used, a means to divide the output energy into multiple transmit paths is required. To accomplish this feat, the transmitter fiber splitter box was developed. Refer to Fig. 1 for an overall optical layout of MABEL and to Table 1 for a compilation of the primary instrument parameters. Essentially a series of cascading beam splitters, the transmitter fiber splitter box separates the laser output into the two wavelengths and then divides the output pulse energy down to the 5–7-nJ level. The output of the transmitter fiber splitter box is a series of eight 1064-nm output beams and sixteen 532-nm output beams. Each output is focused into a 50-μm-diameter fiber. In addition, just prior to coupling into the fibers, each beam path can accommodate a small neutral density filter to be used to vary the energy level for each footprint.
The real heart of the MABEL design, and the aspect that permits great flexibility in selecting and changing the instrument viewing geometry, is the transmitter fiber select box and the corresponding transmitter fiber array. Basically, an array of 215 fibers (107 fibers for each wavelength, plus one center fiber) defines the instrument transmit/receive geometry. The fiber arrays were custom fabricated by Fiberguide, Inc. The instrument geometry as defined by the fiber arrays is illustrated in Fig. 2. At any given time, 16 of the 107 532-nm fibers can be coupled to the 532-nm transmitter fibers and 8 of the 107 1064-nm fibers can be coupled to the 1064-nm transmitter fibers. Selection of a set of transmitter fibers thus defines the instrument viewing geometry. Because each wavelength has a separate set of transmitter fibers, it is possible to have both wavelengths interrogate the same surface area (e.g., the footprints for each wavelength can overlap). While not changeable during flight, the geometry can be changed between flights to permit evaluation of instrument geometry or to optimize validation capability.
The transmitter fiber array is positioned at the focus of a 12.7-cm-diameter telescope. The telescope is f/4.17, resulting in a 100-μrad field of view for the transmitter. From a nominal 20-km operating altitude, this corresponds to a laser footprint on the ground of 2-m diameter. Two matched telescopes, both custom fabricated by Special Optics, Inc., are used in the system. The maximum view angle for the system is ±3° (±52 mrad), or about ±1.05-km cross track.
The receiver is similar, except the fibers are 105-μm diameter, resulting in a 210-μrad receiver field of view. The receiver fiber array, positioned at the focus of the matched receiver telescope, is precisely aligned to the transmitter array. The same fiber positions illuminated on the transmitter fiber select box are selected on the receiver fiber select box. The sixteen 532-nm receiver fibers are routed to a 16-channel Hamamatsu model H7260 photomultiplier tube (PMT) detector. The eight 1064-nm receiver fibers are routed to eight individual Excelitas single-photon-counting modules (SPCMs). The detector dead time (e.g., the time required for the detector to recover from a photon event before the next photon can be detected) on the SPCMs is not ideal for precision ranging, but choices of photon-counting detectors able to operate in the near-IR region are limited.
The signals collected by the detectors are time tagged by custom electronics developed by Sigma Space Corporation. As a prototype for the ICESat-2 mission, demonstrating the electronics was an important step. The time-tagging electronics have a measured resolution of 83 ps (about 12 mm), which is smaller than the ICESat-2 requirement of 150 ps. The data system used to operate the instrument and to collect data is based largely on heritage from the cloud physics lidar (CPL) instrument (McGill et al. 2002). Operating it on the ER-2 platform requires fully autonomous capability and rugged design. Data are stored to a solid-state hard drive that is removed after flight for data retrieval. A Novatel model HG1700 inertial measurement unit (IMU) is mounted directly to the telescope assembly to permit accurate determination of instrument pointing.
The overall schematic of the MABEL instrument is shown in Fig. 3, and a photo of the completed instrument ready for flight is shown in Fig. 4. As built the instrument weighs approximately 250 kg and is accommodated in the nose of the ER-2 aircraft.
The MABEL viewing geometry can be changed by rearranging the connections on the receiver and transmitter fiber select boxes, as described above. For the initial flights in December 2010, we used a symmetric narrow measurement swath, with the 532-nm channels illuminating the 0-, ±0.1-, ±0.2-, ±1.0-, ±1.9-, ±2.0-, ±3.0-, ±4.0-, and +5.0-μrad angles. The 1064-nm channels were set at ±0.1, ±0.2, ±1.9, and ±2.0 μrad. From the nominal 20-km measurement altitude, this resulted in a cross-track width of ±100 m.
To provide precise ranging accuracy, small differences in the optical transmit/receive path must be characterized. Path differences in the transceiver path will correspond to error in ranging. Thus, it is important to properly characterize the instrument offsets. The optical path differences, known from design and from measurement of the optical path, are summarized in Table 2.
5. Initial flights and data results
To meet the needs of the ICESat-2 project, MABEL development was on a highly accelerated schedule. Instrument development began in October 2009, and initial demonstration flights were expected by November 2010. By the target date, the instrument was completed and was shipped to the aircraft for integration. First demonstration flights occurred 1–7 December 2010, with the completion of five flights on consecutive days.
For operation during daytime, the photon-counting system requires stringent filtering to reduce solar background levels. In many lidar and altimeter instruments, interference filters coupled with Fabry–Pérot interferometers (termed etalons if the plate spacing is fixed) are used to minimize sunlight. The MABEL receiver was designed to include an etalon filter, but for the initial flights, the etalon filter was not yet in place. Thus, the December 2010 flights were conducted late in the day and into evening, to minimize the effects of solar background. The very first flight produced satisfactory results, demonstrating that all subsystems functioned properly. The second flight resulted in acceptable demonstration data, and subsequent flights were used to optimize system performance and to sample over differing terrain types.
After the initial flights, the etalon filter was received and installed. With the etalon filter in place, a series of daytime flights was planned for March 2011 to conclusively demonstrate the expected count rates during daytime operation. For expediency and cost effectiveness, the demonstration flights were flown out of Dryden Flight Research Center in California, with emphasis on obtaining useful surface returns rather than obtaining measurements over specific cryospheric targets. Future flights will focus on measurements over ice and snow cover.
An example of the first data collected with MABEL is shown in Fig. 5, from 9 December 2010. The image shows a single MABEL channel (the 532-nm, nadir-pointing channel). The data shown are raw, with no corrections for aircraft pointing or other offsets. Of immediate notice is the multilayered structure in the 5–10-km-altitude region. For reference, the CPL instrument was flown with MABEL, and the CPL quick-look image immediately reveals the presence of multiple cloud layers over the surface (in fact, the clouds are dense enough on the left side of the image that the ground is obscured). These initial data were a surprise to some, as there was a belief that the small pulse energy being transmitted would not permit detection of diffuse targets (i.e., clouds). In fact, as verified by the CPL measurements, MABEL—and therefore the ICESat-2 instrument—are quite capable of detecting cloud layers. Because these data were being plotted in field with minimal processing, display capabilities were limited to one channel at a time. Work is now underway to develop contouring graphics that will utilize the full multichannel capability of the MABEL data.
One early flight of MABEL flew over Lake Mead, east of Las Vegas, Nevada, on 22 March 2011. On this flight, it happened that circumstances allowed a good demonstration of the cross-track measurement capability. As shown in Fig. 6, the aircraft skirted the lake shore, with part of the MABEL swath over the open water of Lake Mead and part of the swath over land and then going out over water. In this image, two of the 1064-nm channels are shown. One channel, 44, is pointed −21 mrad off nadir. This channel remains over water for the duration of the image. Because water is a dark surface, the background signal is low when over water. The second channel, 47, is pointed +47 mrad off nadir. This channel sees an increase in background over the brighter land surface, with the background dropping once that channel moves out over the water. As noted on the figure, the separation between the two channels is about 1.32 km. Prior to the correction for angle offset, the data also showed the proper 17-m-altitude offset between the two channels, which is correct for the effective 26-mrad pointing difference from nadir from the aircraft altitude.
The initial ICESat mission, launched in 2003, demonstrated the unique capability of spaceborne laser altimetry for a variety of advances in the cryospheric and earth sciences. The second-generation ICESat mission (ICESat-2) will represent a substantial advance in terms of ice sheet coverage, measurement precision, and change detection capability over steeply sloping outlet glaciers, along with improved lead detection for sea ice freeboard measurements.
Moreover, the ICESat-2 measurement concept of photon-counting detection is a major shift in measurement approach from previous NASA altimetry missions. The laser technology required to support the photon-counting approach is considerably different from that required for analog altimetry. The optical coalignment and pointing requirements are also more stringent than in previous NASA applications. The success of the technical approach to the ICESat-2 mission will enable a new era in precision laser altimetry measurements for science applications.
To verify and validate the ICESat-2 concept, an airborne simulator was developed. Dubbed MABEL, the airborne instrument was developed in a mere 12 months and was first deployed for engineering test flights in December 2010. A subsequent series of flights was conduced in March 2011. Initial results from MABEL conclusively demonstrate that the concept of photon-counting detection for surface ranging is a viable approach for the ICESat-2 mission. This paper details the MABEL instrument design, as a reference for forthcoming papers that will describe in detail the surface measurements, measurement errors, and other results from upcoming MABEL science flights.
The MABEL development was supported by the ICESat-2 project, managed by Douglas McLennan. In addition to specific vendors acknowledged in the text, the systems engineering, mechanical, and electronics support were provided by Sigma Space Corporation.