Two Weeks and 4,700 Lakes: Lidar Does the Job in Alaska’s North Slope

Author: Kutalmis Saylam, Research Scientist

Bureau of Economic Geology, The University of Texas at Austin


In July 2014, the Bureau of Economic Geology, a research unit at The University of Texas at Austin, was contracted to conduct an airborne survey in the Alaskan North Slope. The purpose of the project was to further determine, understand, and map the local landscape and thaw-lake attributes of an area west of the Dalton Highway and Sagavanirktok (“Sag”) River, approximately 30 km southwest of Deadhorse, Alaska. Researchers from the Bureau had visited the area in 2012 and found bathymetric Light Detection and Ranging (Lidar) to be an effective tool for measuring the area’s lakes [1]. When we returned in July 2014, we conducted numerous flights using the Chiroptera airborne Lidar system manufactured by Airborne Hydrography AB (AHAB) of Sweden. Chiroptera has a near-infrared red wavelength of 1.064 µm for topographic (“topo”) data collection and a green wavelength of 0.515 µm for bathymetric (“hydro”) data collection. The system also encompasses a high-resolution Hasselblad camera that collects color-infrared or natural-color (RGB) imagery.

Shallow thaw lakes, less than 2 m deep in general, are a major component of the tundra landscape of the Alaskan North Slope, where they compose 20% of the total area [2]. They are completely ice-free only a few weeks in a calendar year, so we scheduled the field work accordingly. The lakes’ depth, ice growth, and decay determine whether the lakes are suitable as habitats for wildlife and aquatic fauna, as well as for industrial development [3]. Winter ice is assumed to be 1.5 to 2 m thick in this area, and any liquid water most likely lies below winter ice and talik in the central basins of these lakes if the water is deeper than 2 meters [4]. Survey findings were particularly important for the project sponsor because they could improve geophysics surveys, reveal lakes deeper than 2 m (suitable for building ice-roads), and serve other environmental and industrial purposes.

A de Havilland Canada DHC-6 Twin Otter aircraft was assigned to the project, including two pilots and a mechanic. Researchers from the Bureau travelled from Austin, Texas to Grand Junction, Colorado, where the aircraft was staged. Equipment was shipped to the base location at Grand Junction, where system installation and local testing procedures were completed. Test and calibration flights were completed in Grand Junction to ensure full system functionality before the production phase of the survey began in Alaska.

In Deadhorse, a total of 95 lines were flown to cover the survey area, with line numbers increasing westerly. Survey lines were designed and flown from north to south and south to north. The average flight line was approximately 50 km long. To ensure complete coverage, the flight line spacing was set at 160 to 180 m, where the ground laser swath footprint was calculated to be 280 to 290 m wide. To compensate for the changing ground elevation (30 m in the north, 95 m in the south), atmospheric pressure was monitored during flights to maintain a constant altitude and a uniform swath width on the ground.

Three Trimble 5800 series GNSS receivers were set up at different locations to maximize geometric data quality. Data acquired from each station were corrected and results were averaged with solutions from the Canadian Spatial Reference System (CSRS), University of New Brunswick GAPS, and Trimble CenterPoint RTX post-processing services. In addition, data from two (PU01 and PB0C) Continuously Operating Reference Stations were downloaded for confirmation and/or backup purposes. For Lidar system calibration purposes, ground truth data were acquired over the taxiway at Prudhoe Bay Airport using a Trimble R8 GNSS system. For both scanners, the average vertical offset was measured at less than 1 cm, while the standard deviation was calculated at approximately 3 cm.

A total of 21 billion points were recorded with the topo scanner, and the hydro scanner registered 3 billion points, for a total of 4 terabytes of raw data. A total of 823 RGB images were collected at 1,700 m. The resolution of these images was 22 cm/pixel—highly effective for distinguishing land and water characteristics. These images were tiled and geo-referenced using 50-kHz topographical Lidar data acquired during the same mission. We overlaid these images on Lidar data to enhance the visibility of land features (figs. 3 through and 4).

Figure3_lowResData were organized in 827 individual 1 x 1 km tiles with a 20-m buffer to create a seamless digital elevation model (DEM).  We rasterized data classes water surface, shallow bottom, and water bottom to generate the final DEM, which combines returns from these classes. All 827 raster grids were merged into a single raster data set. The raster grid was then imported into a program written in-house to calculate the bathymetric statistics and extract the minimum and maximum elevation values for each water body. Because of possible anomalies in the elevation values, we conducted a visual inspection to determine which of the highs and lows were out of range. To ensure that these data points were excluded from our computations, we discarded the upper and lower 0.5% of values for each water body.
For bathymetric calculations, we excluded water bodies that were shallower than 0.1 m and smaller than 1,000 m2 in surface area. The minimum volume calculation was set at 34 m3. Every water body was labelled with its latitude and longitude information, in Universal Transverse Mercator coordinates, based on the mid-point of the water bodies determined from the Lidar data set.

A total of 4,697 distinct water bodies were identified in the survey area. The deepest water was calculated at 3.5 m. Only 4.6% (216 total) of all lakes had depths that exceeded 2.0 m. The average depth of all water bodies was calculated at 0.67 m. Shallow lakes were mostly clustered in the northeast area of the project location, and deep water bodies were mostly clustered by the river in the northwest and southwest areas (fig. 5). We found that 64% of all lakes contained less than 1,000 m3 water volume, where the average volume was 12,771 m3.

Once again, airborne Lidar technology provided accurate, detailed, and cost-effective results that permitted analysis of micro-topographic and bathymetric features in a remote location of the world. Water bodies of all shapes and sizes—riverine environments, wetlands and uplands, hills and flat areas, and all other terrain features—were mapped and analyzed rapidly and accurately. The Lidar survey produced detailed and precise topographic and bathymetric data in areas where traditional survey methods would not have been feasible.

The Bureau, is interested in Lidar and hyper spectral imaging acquisition, connected to science-driven outcomes. We are open to collaboration with other academic institutions, State agencies, and government organizations.


J. G. Paine, J. R. Andrews, K. Saylam and T. A. Tremblay, “Airborne LiDAR-based wetland and permafrost-feature mapping on an Arctic coastal plain, North Slope, Alaska,” in Remote Sensing of Wetlands, London, CRC Press, 2015, pp. 413-434.
P. V. Selmann, J. Brown, R. I. Lewelen, H. McKim and C. Merry, “The Classification and Geomorphic Implications of Thaw Lakes on the Arctic Coastal Plain, Alaska,” National Technical Information Service, US Department of Commerce, Hanover, NH, 1975.
M. O. Jeffries, K. Morris and G. E. Liston, “A Method to Determine Lake Depth and Water Availability on the North Slope of Alaska With Spaceborne Imaging Radar and Numerical Ice Growth Modeling,” Arctic, vol. 49, no. 4, pp. 367-374, 1996.
K. M. Hinkel, Y. Sheng, J. D. Lenters, E. A. Lyons, R. A. Beck, W. R. Eisner and J. Wang, “Thermokarst Lakes on the Arctic Coastal Plain of Alaska: Geomorphic controls on Bathymetry,” Permafrost and Periglac, vol. 23, pp. 218-230, 2012.

One comment

  • Although gaining pourtapily this is not new.Combining LiDAR and Sonar has been common practice by hydrographic firms for several years now (four years for our firm). QINSy and HYPACK software packages both provide mature drivers for at least Reigl and MDL lasers.It is an effective method of providing detail on shoreside features both built and natural and combines well with airborne LiDAR.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.