GEOMORPHIC MAPPING FROM LIDAR TOPOGRAPHY

Geomorphic mapping of the glaciated lowlands of northwest Washington (e.g., http://pubs.usgs.gov/of/2009/1033/) documents a rich suite of subglacial and periglacial landforms, extensive ice-stagnation features, and a time sequence largely established by cross-cutting relations that reflect changes in base level. Geomorphic mapping allows rapid inventory of critical areas (wetlands, landslides, hillslopes) and provides a useful framework for neotectonic studies.

Map units are surfaces (2D), not materials (3D). Units are landscape primitives (e.g., alluvial flat, hillslope formed by mass wasting), locally aggregated into more complex units (e.g., landslide, ice-molded ground, kame-kettle) to facilitate mappability. Units are defined on the basis of form, position, and (in some cases) inferred genesis, and are surface types (e.g., beach face) rather than discrete entities (e.g., spits). We map on-screen in a GIS using a variety of DEM-derived images as backdrops. Most contacts are slope breaks. Overprinting (e.g., dunes over ice-modified ground, glacially-modified alluvial flats) presents few challenges; more difficult are errors in the lidar topography and human modification of the landscape. Field checking is largely unnecessary as (a) interpretations are not based on proxy data (e.g., multi-spectral reflectivity as a proxy for lithology) but on the same data (topography) as would be obtained in the field and (b) lidar generally provides topographic data that are more complete and easier to comprehend than those obtained by direct observation in the field. In this wooded terrain, where less than 1/3 of lidar pulses produce ground reflections, 1 pulse/m² data are barely adequate; 8 pulse/m² data are much better.

The presence of a recognizable time sequence and widespread preservation of 15,000-year-old glacial landforms are strong evidence that, at least in this terrain and at scales of ~10³ m and ~10⁴ years, landscape development is fundamentally punctuated.