Deriving Rich Coastal Morphology and Shore Zone Classification from LIDAR Terrain Models

Abstract

Comprehensive mapping of shore-zone morphology supports evaluation of shore habitat, monitoring of environmental hazards, and characterization of the transfer of nutrients between marine and terrestrial environments. This article shows how rich shore-zone morphological metrics can be derived from LIDAR terrain models and evaluates the application of LIDAR to classify shore-zone substrates. The utility of LIDAR methods was tested in comparison with the current best-practice method of photo interpretation (i.e. The BC ShoreZone system) on Calvert Island, British Columbia, Canada. Wider applications are considered. Indicators of shore-zone morphology (i.e. slope, width, roughness, backshore elevation) were calculated from LIDAR terrain models for regularly spaced transects perpendicular to the coastline. A combination of boosted regression-Tree modeling and direct-rule application was used to classify the shore-zone morphology according to the British Columbia (BC) ShoreZone system. Classification accuracy was assessed against existing ShoreZone classification data. Shore-zone substrate was classified from LIDAR-derived morphometric indicators with 90% accuracy (five classes). A full classification, which combined substrate with shore width and slope, results in lower correspondence (40%; 25 classes) when compared with ShoreZone classes. Differences can likely be attributed, in part, to variation in spatial resolution of elevation-based methods and photo interpretation. It is concluded that LIDAR data can be used to support characterization of shore-zone morphology. Differences in processing and interpretation cause a low direct correspondence with the current image-based classification system, but LIDAR has the advantage of higher resolution, rich terrain information, speed, and an objective and repeatable method for monitoring future change in coastal environments.