ICC Programming

Mapping landscapes

7110.3 - Patterns of terrain feature extraction from variable terrain using multiple methods

Friday, July 7
9:10 AM - 9:30 AM
Location: Coolidge

Extracting terrain surface features from a raster digital elevation model (DEM) provides information for topographical analysis, deriving feature-based information, and the analysis of pattern- and process-related effects. In addition, it allows for semantic reasoning involving relationships among the features. With such aims for future research, we are interested in identifying terrain surface features related to surface network theory: peak, pit, saddle, ridge, and course features.

Identifying such terrain features from DEMs, however, is not a trivial task. We evaluated the effects of multiple scales and multiple extraction workflows across various degrees of terrain variability. The workflows involved several free, open-source, and commercial software. Our empirical analysis varied the three factors of scale, workflow, and terrain variability, resulting in a sequential exploration of fuzzy membership values at several stages. We discuss multi-scale and multi-feature membership values for both single- and multi-method approaches. As the multi-scale fuzzy membership values for the five features can intersect spatially, a confusion index is employed as an uncertainty metric for the multi-feature analysis of dominant and second-ranked terrain features at each location. Multi-scale and multi-feature maps are generated for each stage of analysis involving both single- and multi-method approaches.

In general, we find that terrain feature uncertainty is related to terrain variability with smoother, less-variable terrain leading to poorly-identifiable features. Terrain with high variability results in crisp features, with regard to both high membership values and fine spatial extents. In addition, we find that locations of all surface network features are highly dependent upon the scale of analysis. Using a multi-scale approach reveals that any location has the potential to exhibit characteristics of any terrain feature. Although we find that features associated with natural terrain can only be described in terms of uncertainty, our overall conclusion is that a multi-method approach can provide a better understanding about where particular terrain features might exist.

Boleslo E. Romero

PhD Candidate
Department of Geography University of California, Santa Barbara

Boleslo E. Romero is a Ph.D. candidate in the Department of Geography at the University of California, Santa Barbara. His research interests are Geographic Information Science (GIScience), Geographic Information Systems (GIS), spatial analysis, remote sensing, terrestrial lidar, and geographical feature extraction. Recent research involves spatial outliers and terrain feature extraction.

Presentation(s):

Send Email for Boleslo Romero

Keith Clarke

Professor
Department of Geography, University of California, Santa Barbara

Dr. Keith C. Clarke is a research cartographer and professor with the M.A. and Ph.D. from the University of Michigan in Analytical Cartography. He is the former North American Editor of the International Journal of Geographical Information Systems and has authored about 250 book chapters, journal articles, and papers in the fields of cartography, remote sensing, and geographic information systems. He has also served on numerous National Research Council studies and the National Geographic Society's Committee on Research and Exploration. Awards include the USGS John Wesley Powell Award (2005) and the Cartography and Geographic Information Society's Distinguished Career Award (2014).

Presentation(s):

Send Email for Keith Clarke

Nina S.N Lam

Professor
Louisiana State University

Nina Lam is Professor and E.L. Abraham Distinguished Professor in the Department of Environmental Sciences at Louisiana State University. She was Chair of the Department (2007-2010), Program Director of U.S. National Science Foundation Geography and Spatial Sciences Program (1999-2001), and President of University Consortium on Geographic Information Science (UCGIS, 2004). Professor Lam’s research interests are in GIS, remote sensing, spatial analysis, environmental health, and community resilience. She has published on topics including spatial interpolation, fractals, cancer mortality, scale and uncertainties, AIDS in America, business recovery in New Orleans after Katrina, community resilience assessment, coastal resilience modeling using a coupled natural-human system approach. Lam has received top awards from within LSU (Distinguished Faculty, Rainmaker, Distinguished Research Master, and Outstanding Faculty Research Award), and outside LSU including Association of American Geographers’ (AAG) (Outstanding Contributions in Remote Sensing) and University Consortium on Geographic Information Science (UCGIS Fellow, The 2016 Inaugural Carolyn Merry Mentoring Award). Lam has coedited two books and authored and co-authored over 95 refereed articles. She has served as the Principal Investigor or Co-Principal Investigator of over 40 external grants. Lam has advised 5 post-doctoral associates, 17 PhDs, and 30 M.S. students.

Presentation(s):

Send Email for Nina Lam


Assets

7110.3 - Patterns of terrain feature extraction from variable terrain using multiple methods



Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Patterns of terrain feature extraction from variable terrain using multiple methods