The ATM surface elevation measurements have been re-sampled into an icessn format which smoothes the data and reduces the data volume. Users desiring unsampled data should use the IceBridge ATM L1B Elevation and Return Strength data.

The fundamental form of ATM topography data is a sequence of laser footprint locations acquired in a swath along the aircraft flight track. The icessn program condenses the ATM surface elevation measurements by fitting a plane to blocks of points selected at regular intervals along track and several across track. The block size and spacing can be specified, but a few typical values are used. The along-track distance smoothed is the distance which the aircraft moves in a fixed interval, 0.5 seconds for P-3 and DC8 aircraft, and 1.0 seconds for DHC-6 Twin-Otter. The data output interval is half of the smoothing interval so that there is 50% overlap between successively smoothed blocks. For each along-track position/time, there are multiple blocks spaced evenly across-track to span the swath width. Typically the number of blocks is five for the T3 scanner and three for the T2 scanner. There is an additional block located at aircraft nadir with a width typically set to 80 m. If a single profile is desired, the nadir profile can be selected from the full data set.

The two slopes estimated are used to estimate surface elevations at points other than the center point through the use of the following algorithm shown in Equation 1 and described in Table 3:

Note: The multiple across-track planes at a given along-track position will have the same time tag.

Processing Steps

The following processing steps are performed by the data provider.

1. Preliminary processing of ATM lidar data through the cvalid program, applying calibration factors to convert time of flight to range, scan pointing angles and interpolated attitude to each lidar measurement.
2. Processing of GPS data into aircraft trajectory files using double-differenced dual-frequency carrier phase-tracking.
3. Processing of the cvalid program output combined with the GPS trajectory data through the qfit program, resulting in an output file containing a surface elevation (ellipsoid height) and a geographic location in latitude and east longitude with other ancillary parameters.
4. Processing of the qfit output through the icessn program which averages the qfit surface elevation data into a small number of blocks or surface planes.

Version History

Beginning with the 2013 Arctic campaign, all new data are provided in Version 2 CSV format. Data from 2012 and earlier campaigns remained in the previously-used Fixed-width, space-delimited ASCII format. On 10 August 2015, Version 1 fixed-width space-delimited ASCII text files for 31 March 2009 to 08 November 2012 were replaced with CSV files and were included with Version 2 data.

A header was added to the Version 2 data format. In Version 1, the data set consisted of file pairs for each data segment: one data file, ILATM2_YYYYMMDD_HHMMSS_50pt_smooth_nadir3seg, and one summary file, ILATM2_YYYYMMDD_HHMMSS_50pt_icessn_summary. The new header of the Version 2 data file contains information that had appeared in the summary file, making the summary file redundant.

Error and Limitations

Uncertainty in slope estimates:
The uncertainty of the WGS84_Ellipsoid_Height, South-to-North_Slope, and West-to-East_Slope can be attributed to both physical topographical features and measurement error. The uncertainty in the WGS84_Ellipsoid_Height can be reasonably approximated with the RMS_Fit value. The uncertainty in the South-to-North_Slope and West-to-East_Slope can both be reasonably approximated with the equation:

Where:

Note: The RMS_Fit value is in centimeters and must be converted to meters to be used in the equation above.

Attenuated ATM signal 12 and 13 April 2010: 
During collection of IceBridge ATM Greenland data on 12 and 13 April 2010, hydraulic oil progressively leaked from the forward landing gear on the DC8 aircraft. The oil was blown back along the bottom of the fuselage and across the nadir window through which the ATM was transmitting and receiving the laser signal. The ATM signal was attenuated, and data in part of the scan is missing as a result. The problem developed during the flight and worsened through time. The ATM still acquired more than half of the shots throughout the scan. The net effect of this problem is to decrease the number of shot returns logged, the same as if the laser power was reduced. To the user this will appear as a reduced point density on the ground. This issue will not affect the accuracy of the data. In the Antarctic 2010 campaign, fuel leakage degraded the signal in a similar fashion.

Skipped glaciers and supplemented data for April 28, 2012: 
The flight on 28 April 2012 traversed the notoriously turbulent regions over Greenland's southeast glaciers. During the flight, two planned glaciers were skipped due to concern about expected severe turbulence. The survey data spans roughly 11:15 to 18:20 UTC. On the approach to Ikerssuaq glacier at 16:56:19.5 (GPST=60994.5 secs), both the ATM T3 and T4 instruments quit recording data within 0.1 second of each other. T3 resumed at 16:57:05.3 whereas T4 did not resume for the rest of that day's flight. Following this event, the flight followed the Ikerssuaq flow line, then traversed straight west across the icesheet back to the Kangerlussuaq airport. The data gap spans 46 seconds, from the fjord up to about 500 m elevation on the Ikerssuaq glacier. The T4 data quit during the creation of the file 20120428_165532.ATM4BT4.F1.qi. The Level-1B T4 data were supplemented by the these Level-1B narrow swath files of T3 data from the latter part of the survey:

20120428_165449.atm4cT3.F1.qi
20120428_170030.atm4cT3.F1.qi
20120428_170524.atm4cT3.F1.qi
20120428_171019.atm4cT3.F1.qi
20120428_171514.atm4cT3.F1.qi
20120428_172008.atm4cT3.F1.qi
20120428_172503.atm4cT3.F1.qi
20120428_172957.atm4cT3.F1.qi
20120428_173452.atm4cT3.F1.qi
20120428_173947.atm4cT3.F1.qi
20120428_180815.atm4cT3.F1.qi
20120428_181219.atm4cT3.F1.qi
20120428_181619.atm4cT3.F1.qi

The above files can be found with the IceBridge ATM L1B Qfit Elevation and Return Strength data in the /2012.04.28/ folder. For details on the ATM 4BT3 and 4BT4 instruments, see the Sensor Or Instrument Description section below, and the IceBridge Narrow Swath ATM L1B Qfit Elevation and Return Strength data set documentation.

ATM wide scan instrument failure 14 April 2015 - ATM Narrow Scan instrument fills data gap: 
On 14 April 2015, the ATM wide-scan instrument suffered a failure in the scanning mechanism and did not collect data for that mission. However, the ATM narrow-scan instrument was operating normally and has been used to fill in the gap in ILATM2 data for this mission.

The ILATM2 averaged product is derived from the ILATM1B wide-scan elevation point cloud, when possible. The ILATM2 data splits the swath into three segments relative to the direction of flight: port-side, center, and starboard-side. From these strips, the data is further separated into 0.5-second-long tiles. With a nominal Above Ground Level (AGL) of 450 m, a swath width of 240 m, and a groundspeed of 100 m/s; this creates tiles that are 80 m wide and 50 m long. There is also a fourth tile created, called the nadir file, which is positioned directly below the aircraft. This tile is also nominally 80 m wide and 50 m long.

The ATM narrow-scan swath width is only about 40 m wide, so generating tiles equally-sized to the ATM wide-swath is not possible. For best results, it was decided to separate the lidar swath in half: port-side and starboard-side. For consistency, the along-track distance of the tiles is still 0.5 seconds. Therefore, for this delivery, the 14 April 2015 tiles are roughly 20 m wide and 50 m long. The nadir tile is also included, which is still nominally 80 m wide and 50 m long like the wide-swath ATM, and for straight-and-level flight includes the entire swath.

Fall 2015 Campaign 
For the Fall 2015 ATM data, some ATM elevations are adjusted slightly (~10 cm) to compensate for a systematic anomaly related to the ATM scanner azimuth. The overall mean elevation is not changed, but some elevations around the scan are adjusted upward or downward as a function of scanner azimuth. For further details on the adjustment method, see Yi et al. (2015).

Fall 2016 Campaign
On 07 November 2016, the ATM wide-scan lidar (ATM6aT6) suffered a laser system failure and did not collect any usable ILATM1B data for that flight. Consequently, there are no ILATM2 data for this date.