Stochastic Uncertainty Analysis for Lidar-Derived Shoreline and Comparison with New Experimental Results

Abstract

NOAA’s National Geodetic Survey (NGS) is responsible for mapping the legally-recognized national shoreline depicted on NOAA nautical charts. The primary intended use of this shoreline is in support of safe navigation; however, both historical and current shoreline data are also being used in an increasingly wide range of coastal science and management applications. In particular, NGS’ shoreline is proving beneficial in understanding and responding to climate change threats, such as sea-level rise and shoreline erosion. Other uses within the coastal research community include habitat mapping, monitoring, and restoration, as well as emergency planning and response. Over the past half century, NGS’ principal method for shoreline mapping has involved stereo compilation from tide-coordinated, near-infrared aerial imagery. While these photogrammetric procedures continue to be NGS’ primary shoreline mapping methodology, in collaboration with numerous partners over the past decade, NGS has conducted several phases of research into lidar shoreline mapping. The lidar work has culminated in a set of production procedures for utilizing lidar in mapping NOAA’s shoreline. Proven or anticipated benefits of the lidar-based procedures include increased automation, reduced subjectivity inherent in the manual compilation process, and provision of multi-use coastal elevation datasets. Because NGS’ lidar shoreline mapping procedures are beginning to be used for coastal mapping production projects, empirical accuracy assessments are needed to generate metadata. Additionally, stochastic uncertainty analysis is needed for a variety of purposes, including: 1) conforming to International Hydrographic Organization (IHO) standards that mandate uncertainty analysis for charted “coastline” (shoreline); 2) informing policy decisions related to accuracy and efficiency in shoreline mapping; and 3) performing sensitivity analysis. Furthermore, the uncertainty analysis is also required by external partners within the coastal research community to determine the suitability of the output product for other applications—including those mentioned above—and analyzing the propagated uncertainty in downstream products, such as shoreline change rate estimates. To address these needs, we conducted a study in the North Carolina Outer Banks using lidar data and ground truth collected during an Integrated Ocean and Coastal Mapping (IOCM) pilot project. The first research objective was to develop and test a new empirical accuracy assessment procedure for NGS’ lidar-derived shoreline. Ground truth consisted of shoreline transects surveyed with an integrated laser-level and real-time kinematic (RTK) GPS, and tied vertically to tidal benchmarks in the project area. The second objective was to perform a stochastic uncertainty analysis using Monte Carlo simulation and estimates of the uncertainty in each observable in the lidar system. The results of the two parts of the study will be described here, followed by a comparison of both and recommendations for follow-on research.

Department

Center for Coastal and Ocean Mapping

Publication Date

2-2010

Volume

91, Issue 26

Journal Title

Ocean Sciences Meeting

Conference Date

Feb 22 - Feb 26, 2010

Publisher Place

Portland, OR, USA

Publisher

American Geophysical Union Publications

Document Type

Conference Proceeding

Link to Full Text

Share

COinS