Ensuring Data Quality for Alaska’s North Slope

A 1.342Mb PDF of this article as it appeared in the magazine complete with images is available by clicking HERE

The Alaska Department of Natural Resources, Division of Geological & Geophysical Surveys (DGGS) collected high-resolution LiDAR topographic data for swaths of land that would likely be used if new natural gas pipelines were constructed in Alaska. The pipelines would be designed to deliver Alaska North Slope natural gas to out-of-state and Alaska customers. Collection of LiDAR data was supported by the State of Alaska Gas Pipeline Project Office, the Office of the Federal Coordinator, and the Alaska Gasline Development Corporation.

The data–covering approximately 3,000 square miles along the proposed pipeline routes–were acquired to facilitate analysis during the design, permitting, and construction of the pipelines, including evaluation of active faulting, slope instability, thaw settlement, erosion, and other engineering constraints along the proposed pipeline routes. The data will also provide an Esri ArcGIS base layer for the statefederal GIS database and will be used to evaluate permit applications and construction plans.

LiDAR was chosen as the preferred data format because it has proven to be one of the most useful forms of remotely sensed data for identification and characterization of potentially active faults and many other surficial-geologic landforms and hazards, especially in areas of heavy vegetation where access may be difficult and other forms of remotely sensed data are ineffective.

Acquiring LiDAR data
LiDAR data were acquired and processed by Watershed Sciences, Inc. (WSI), based in Corvallis, Oregon. The data consist of continuous 1-mile-wide minimum coverage of existing infrastructure along the entire length of the proposed natural gas pipeline corridors from Prudhoe Bay, Alaska, to the Canada border along the Trans-Alaska Pipeline System (TAPS) and Alaska Highway; Delta Junction to Valdez, Alaska along the TAPS corridor, and Livengood to Point McKenzie, Alaska along the George Parks Highway. LiDAR data were collected between September 21 and October 1, 2010, for an area that included portions of the Alaska Highway east of Johnson River to Robertson River, and included 117,357 acres within the requested area. This area was expanded to include a 100-meter buffer to make sure coverage was complete and there were adequate point densities around survey boundaries, resulting in 121,746 acres of delivered data.

The LiDAR survey data were collected with Leica ALS60 sensors in a Cessna Caravan 208B and Partenavia P-68 aircraft. The Leica systems were set to acquire 200,000 laser pulses per second and flown at 900 or 1,300 meters above ground level, depending on cloud ceiling and terrain, capturing a scan angle of plus/minus 14 degrees from nadir. The system achieved a minimum survey density of 8 pulses per square meter and up to four returns per pulse. The data included minimum 1-mile-wide corridors covering gas-pipeline routes being considered by applicants, and half-mile-wide coverage of existing primary pipeline-support roads outside the main corridor. Other expanded coverage included areas where data are needed for evaluation of active faults, slope instability, and other geologic hazards.

To facilitate processing and product delivery, WSI grouped the data into delivery areas (subsets of the entire data collection region) in the order in which they were collected and processed. Files for each delivery area were organized by 1:63,360-scale quadrangle. Next, WSI sent the data for each delivery area to the State of Oregon Department of Geology and Mineral Industries (DOGAMI) for independent quality-control analysis under separate contract with DGGS. DOGAMI is a proven leader in the collection and use of LiDAR data, with vast experience in these types of analyses.

After correcting any errors identified by DOGAMI, WSI sent the revised data set to DGGS along with a delivery report describing details about LiDAR acquisition, accuracy, and quality for the delivery area. DOGAMI also provided a separate report for each delivery area, summarizing their methodologies and the results of quality control checks.

LiDAR points were classified as ground, low and medium vegetation, buildings or manmade structures, noise/ error points, water, snow, and pipeline. Contour lines were derived at 1 meter intervals from ground-classified LiDAR point data using TerraSolid processing software and Microstation. For the derived DEMs, lakes and other closed water bodies with surface areas of greater than 150 square meters were flattened to a consistent water level using a hydroflattening procedure to ensure stream channels break at culvert locations and the sinks and pits in the data are removed.

Ensuring Quality Using ArcGIS
Bare-earth and highest-hit grids were delivered in ArcGIS grid format with 1 meter cell size. LiDAR point data were delivered in LAS binary format for ground classified returns as well as the entire LiDAR point cloud. Georeferenced intensity images of 1 meter cell size were supplied in geoTIF format. Supplementary data include 1-meter-cellsize vegetation rasters displaying canopy and other vegetation metrics. Real-time kinematic ground survey data that are used for absolute vertical adjustment were supplied in shapefile format.

All data associated with the delivery were loaded and viewed to ensure completeness. Raster imagery such as elevation grids and intensity geoTIFFs were viewed in ArcGIS and cross referenced with the delivery area. These grids were loaded into ArcGIS software for visual analysis. Data were examined through slope and hillshade models of bare earth returns. Hillshades of the highest-hit models were used to identify areas of missing ground. Both bare-earth and highest-hit models were examined for calibration offsets, tiling artifacts, seam-line offsets, pits, and birds (both real and anomalous).

Errors noted during visual analysis were digitized for spatial reference and stored in Esri shapefile format. Each feature was assigned an ID value and commented to describe the nature of the observed error. The shapefile was delivered to WSI for locating and fixing apparent errors. Upon receiving the observed error locations, the vendor performed an analysis to conclude whether the error was valid, and reprocessed if necessary.

Delivering the goods
DGGS has made this LiDAR data available free to the public at its website (www.dggs.alaska.gov/pubs/id/22722). The data are arranged by U.S. Geological Survey (USGS) quadrangle in the order that delivery areas were received from WSI. A single data delivery from WSI typically covered portions of several 1:250,000-scale quadrangles. Individual delivery reports from WSI were combined into a comprehensive report organized by delivery area. Section 1 of this delivery report contains information about the first delivery received by DGGS; Section 2 contains information about the second delivery received, and so on. A similarly organized quality control report contains information provided by DOGAMI. Both reports are available via the DGGS website.

The DGGS data release includes bare-earth digital elevation models (DEMs), LiDAR intensity images, bare-earth DEM hillshade images, water-body polygons, canopy cover digital surface models (DSMs), normalized DSMs, vegetation DSMs, mean vegetation elevations, highest-hit DSMs, coefficient of variation DSMs, and point-cloud data.

Processing three-dimensional data such as LiDAR requires software that can accurately manage, represent, and analyze its feature information. Alaska DGGS found that ArcGIS provided the platform needed to ensure the data are ready for analyzing proposed pipeline routes by providing a base layer of data for the GIS database that will be used by both the state and federal governments for evaluating permit applications and construction plans. For more information on LiDAR and 3D mapping with ArcGIS visit esri.com/lidar.

James Weakland, GISP, has been with DGGS since 2011 and serves as a GIS analyst in the Geologic Communications Section, updating legacy geologic mapping techniques including instituting division standards for GIS data, and creating web mapping applications for customers without access to GIS software. Weakland holds a degree in Geography from Excelsior College and is currently studying at The Pennsylvania State University for his Post Baccalaureate Certification in GIS. He can be reached at james.weakland@alaska.gov or (907) 451-5029.

A 1.342Mb PDF of this article as it appeared in the magazine complete with images is available by clicking HERE