Laser Scanning with AAD's LiDAR system
Table of Contents
- Background
- The Laser Scanner
- The Inertial Measurement Unit (IMU)
- Setup
- Data collection
- Flight track pattern
- Data processing
- Data samples
- Contact
A more technical version of this document is available as a single page PDF, 2MB download.
Image by Jan L Lieser (ACE CRC) & Adam Steer (AAD)
Background
The term laser scanner refers to a system using an opto-mechnical scanning mechanism to measure the range between the sensor and the illuminated spot on the object surface. In case of airborne laser scanners the laser beam is generally directed down so that this spot is on the ground. In contrast to multi-spectral scanners like the space borne Landsat Multispectral Scanner MSS or Thematic Mapper TM, laser scanners are active systems and use a laser beam as the sensing carrier. There are two common acronyms for such systems: LiDAR (Light Detection And Ranging) and LaDAR (Laser Detection And Ranging). While LiDAR systems might be built up with any kind of light source (for example xenon or flash lights), LaDAR specifically refers to laser sources.
The LaDAR system described in the following is a new asset of the sea ice group within the glaciology program at the Australian Antarctic Division (AAD) to undertake large-scale sea ice (plus snow cover) surface elevation and roughness surveys for the validation of satellite-based measurements. The system is aircraft-independent, which means that it can be used from helicopter and fixed-wing aircraft. It will play a crucial role in validation studies of laser and radar altimetry programs, operated by both NASA and ESA, including the existing ICESat (a laser altimeter satellite, operated by NASA) and ENVISAT (which has a radar altimeter and is operated by ESA) missions, as well as planned ICESat II and CryoSat II (a dedicated cryosphere radar altimeter, operated by ESA) missions.
Making sufficient surface measurements for satellite validation is a challenging proposition, given the coarse nature of satellite data compared with ground truth point measurements on single ice floes. The spatial differences in the two measurement techniques make it essential to conduct some form of aircraft-based measurements, as an intermediate layer between highly-detailed surface observations and low-resolution satellite measurements. Space-based instrumentation offers the only means by which sea ice and snow cover thickness can be monitored globally, making field validation efforts over the next decade crucially important to the development and improvement of algorithms using data from these instruments.
Deriving ice freeboard height F (or any other kind of surface elevation) from laser scanner data is based on the following relationship:
where F is the ice freeboard height, h is the height of the aircraft, H is the measured range to the ice surface, N is the geoid, and Δh includes measurement errors, geoid errors, and ocean dynamic topography (tides and mean). These freeboard estimates can then be converted into ice thickness estimates assuming isostatic equilibrium between sea ice and water. In situ measurements of snow thickness and density from ice stations will provide the necessary in situ data to help with the interpretation of the aircraft-based measurements.
The system consists of three components: the laser scanner, an inertial measurement unit, and a logging (and processing) PC. It will not only be used for marine cryosphere work but also provides capabilities for high precision range measurements for digital elevation mapping of glaciers, ice shelves, icebergs, and islands. In combination with coincident digital aerial photographs high resolution 3D digital elevation models can be produced.
The Laser Scanner
The Riegl LMS-Q240i-60 2D Laser Scanner makes use of the pulsed time-of-flight range measurement principle and beam scanning by means of an opto-mechanical scan mechanism, providing fully linear, unidirectional and parallel scan lines. The scanner has two major components: 1) range finder electronics, and 2) a rotating mirror.
The range finder electronics consists of a fast repetition rate laser, signal processing electronics and high speed data interface. The angular deflection of the laser beam is realized by a rotating polygon with four reflective surfaces. It rotates continuously at an adjustable speed to provide a unidirectional scan within an angle of q = 60°. For every measurement Range, Scan Angle, Signal Amplitude, and optionally a Timestamp are provided via a TCP/IP Ethernet interface. The LMS-Q240i accepts a TTL-signal (i.e., 1 pulse per second) from a GPS receiver, to reset an internal timer, which is used to timestamp every measurement.
The laser has a nominal measurement range of up to 650m (2100ft) for natural targets with r³ = 80%. The accuracy and precision of the measurements are 20mm and 15mm, respectively. The pulse repetition rate is fixed at 30kHz giving an effective measurement rate of about 8kHz (taking the rotation of the polygon into account, when the laser beam is not directed through the scanner window - at the edges of the polygon). The laser wavelength is at near IR (905nm). The beam has a divergence of 2.7mrad which results in a laser beam foot print at typical survey altitudes over snow covered surfaces of 36cm at 450ft (137m), and 119cm at 1500ft (457m). The scanning rate is adjustable between 6 and 80 scans per second.
The Inertial Measurement Unit (IMU)
An Inertial Measurement Unit (IMU) (or Position and Orientation System, POS) is mandatory for airborne LiDAR surveys. In conjunction with the AAD's laser scanner an Inertial and GPS Navigation System (INS/GPS) manufactured by Oxford Technical Solutions is used: RT4003. This IMU is a six-axis inertial navigation system that incorporates an L1/L2 Real Time Kinematic GPS receiver for position and a second GPS receiver for accurate heading measurements. It delivers better than 0.02m positioning under dynamic conditions using differential corrections and 0.1° heading using a 2m separation between the GPS antennas.
The RT4003 INS/GPS includes three angular rate sensors (gyros), three servo-grade accelerometers, the GPS receivers and all the required processing in one compact box. It has a fast update rate (250Hz) and a wide bandwidth. All outputs are computed in real-time with a very low latency and broadcast over LAN.
The internal processing includes the strapdown algorithms (using a WGS-84 earth model), Kalman filtering and in-flight alignment algorithms. The internal Pentium-class processor runs QNX real-time operating. The Kalman filter monitors the performance of the system and updates the measurements using GPS. The second GPS receiver measures the difference in position compared to the primary GPS receiver using the carrier-phase observations from both. The use of two GPS receivers delivers accurate measurements of heading even when the vehicle's dynamics are low.
The setup
As it stands at the moment the system can be used from helicopters to provide long uninterrupted transects across the pack ice. The laser instrument is currently fitted in a custom made shock mount that holds both the laser scanner and the IMU, as well as a pyrometer (which can be run simultaneously, but is not scope of this document). This mount fits in a Eurocopter AS 350 B 'Squirrel' helicopter in the front left floor window with a cowling to be slipped on to the mount instead of the window.
The laser points down through the outside part of the mount platform in a way that it produces a scan pattern across flight direction. A mechanical shutter protects the laser window (and pyrometer lens) against dispersing dust during take off and landing. The IMU is mounted as recommended by the manufacturer so that the x-axis of the instrument points in forward vehicle (helicopter) axis, the y-axis of the instrument points to the right, and the z-axis points down.
Two GPS antennae (Antcom 42GO1215A4-XT-1) are mounted in the skylight (roof) windows of the helicopter and connect to the IMU directly. Power and data distribution is done through a central rack installed on the floor of the helicopter in front of the back bench.
Data collection
The laser scanner data are collected using a set of purpose built Pascal routines by Wolfgang Lieff of Airborne Research Australia (ARA) at Flinders University, Adelaide. The scanner control software 'ricontrol' provides all necessary commands for normal operation. All parameter needed to be set for a specific scan pattern (lines per second, start angle of polygon for measurements, number of measurements per line, trigger mode) are controlled via the software. The trigger mode can be set to either internally (following the built-in clock) or externally (forced by a TTL signal). The laser scanner produces approximately 250MB of raw data per scanning hour.
To obtain measurements accurate to 0.02m, the RT4003 requires differential corrections from a suitable base station. During surveys commencing from RV Aurora Australis L1/L2 GPS base stations are required on board the ship and the ship should preferably be as stationary as possible. Data collection should be also requested from other suitable base stations (for example, land bases in Tasmania and/or Antarctica).
The IMU stores raw data on an internal hard drive (500 MB, approx. 15 hrs. data). These data files can be transferred through the Local Area Network. Once the hard drive reaches its capacity, the oldest data are being overwritten. The IMU is configured using manufacturer supplied software RT Configuration Wizard.
Flight track pattern
After the operating system is airborne, it is recommended to perform a figure Eight pattern (or a race track pattern) over any kind of know structure, for example the airport's runway or a ship's deck. This will help to rectify the good accuracy during data processing. If the known pattern appears congruent regardless of flight direction this is an easy test of the performance of the system.
The typical survey usually consists of a series of cross sections over specific regions, transects hereafter. The following are three types of transects that are flown commonly over sea ice:
- Ice station transects are repeated flights at low altitude across sea-ice floes adjacent to the ship during ice stations to validate the airborne data with the surface measurements.
- Longer transects flown on a specific heading are to collect large-scale information on the sea ice. This data in turn is used to validate satellite information. High accuracy of these transects (matching temporal and geographic coverage of satellite overpass) is imperative to validate the satellite data for example from ICESat, which has a footprint of 70 metres at 150 metre intervals.
- The buoy array is a third type of transect where repeated flights allow for quantifying dynamic processes in the sea ice (convergence and divergence) and lateral movement.
Data processing
The processing of laser data requires the following steps:
- Extract time, mirror angle, range and returned power from the raw measurements.
- Determine the orientation of the laser scanner relative to the aircraft.
- Obtain position and attitude for each measurement from combined GPS/INS solution.
- Ground reference each measurement using 3D geometry.
Contact
PI: Jan L. Lieser (Please replace '[at]' with '@' before clicking 'send'.)
Antarctic Climate & Ecosystems Cooperative Research Centre
+61 (0)3 6226 7899
Tony Worby (Please replace '[at]' with '@' before clicking 'send'.)
Australian Antarctic Division
+61 (0)3 6226 2985