Quantifying DEM Uncertainty and its Effect on Topographic Parameters

Abstract

Digital elevation models (DEMs) are representations of topography with inherent errors that constitute uncertainty. DEM data are often used in analyses without quantifying the effects of these errors. This paper describes a Monte Carlo methodology for evaluation of the effects of uncertainty on elevation and derived topographic parameters. Four methods for representing DEM uncertainty that utilize metadata and spatial characteristics of a DEM are presented. Seven statistics derived from simulation results were used to quantify the effect of DEM error. When uncertainty was quantified by the average relative absolute difference, elevation did not deviate. The range of deviation across the four methods for slope was 5 to 8 percent, 460 to 950 percent for derived catchment areas and 4 to 9 percent for the topographic index. This research demonstrates how application of this methodology can address DEM uncertainty, contributing to more responsible use of elevation and derived topographic parameters, and ultimately results obtained from their use.

Assessing uncertainties associated with digital elevation models for object based landslide delination

“…This methods has been employed by many researchers to evaluate error associated with GIS datasets (Wechsler and Charles, 2006). In order to make of Monte Carlo Simulation for DEMs uncertainty analysis, it is required to recognize DEMs as only one possible realization of the true elevation surface.…”

“…Accordingly in order to create this distribution a number (N) of random error fields are generated where each cell represents the possible error at a co-located elevation. Each random field is added to the DEMs generating a new realization of the elevation surface (Wechsler and Charles, 2006).…”

Digital elevation model uncertainty and hazard analysis using a geophysical flow model

“…Usually, these terrain representations are digital elevation models (DEMs). For this type of study, terrain elevation is rightly recognized as the most essential and fundamental of variables in geographical analysis (Mitasova et al 1996;Atkinson 2002;Wechsler & Kroll 2006;Stefanescu et al, in press). Dalbey et al (2008) introduced procedures for constructing hazard maps using ensembles of CFD model simulations (the TITAN2D code; Patra et al (2005)) of such flows constructed by establishing probability distributions of input uncertainties in flow initiation (location and volumes) and by sampling them.…”

“…statistic. However, many papers have reported on the limitations of a single value of accuracy, stressing that DEM error is spatially variable and highly correlated (Wechsler & Kroll 2006;Darnell et al 2008). Also the magnitude of the DEM error is closely related to the characteristics of the terrain surface.…”

“…In the literature, DEM error without spatial autocorrelation was considered to be a worst case scenario (Heuvelink et al 1989;Van Niel et al 2004;Oksanen 2006), but no analysis based on terrain morphology and the effect of different DEMs was done. Wechsler & Kroll (2006) developed four different methods for representing the spatial dependence of error through random fields to assess the effect on topographic parameters of the DEM uncertainty. The study showed that uncertainty in the DEM is manifested at higher elevations in locally steeper slopes, on both slope and elevation maps.…”