Uncertainty modelling and analysis of volume calculations based on a regular grid digital elevation model (DEM)

Chang Li, Qing Wang, Wenzhong Shi, Sisi Zhao

Research output: Journal article publicationJournal articleAcademic researchpeer-review

11 Citations (Scopus)

Abstract

The accuracy of earthwork calculations that compute terrain volume is critical to digital terrain analysis (DTA). The uncertainties in volume calculations (VCs) based on a DEM are primarily related to three factors: 1) model error (ME), which is caused by an adopted algorithm for a VC model, 2) discrete error (DE), which is usually caused by DEM resolution and terrain complexity, and 3) propagation error (PE), which is caused by the variables’ error. Based on these factors, the uncertainty modelling and analysis of VCs based on a regular grid DEM are investigated in this paper. Especially, how to quantify the uncertainty of VCs is proposed by a confidence interval based on truncation error (TE). In the experiments, the trapezoidal double rule (TDR) and Simpson's double rule (SDR) were used to calculate volume, where the TE is the major ME, and six simulated regular grid DEMs with different terrain complexity and resolution (i.e. DE) were generated by a Gauss synthetic surface to easily obtain the theoretical true value and eliminate the interference of data errors. For PE, Monte-Carlo simulation techniques and spatial autocorrelation were used to represent DEM uncertainty. This study can enrich uncertainty modelling and analysis-related theories of geographic information science.

Original languageEnglish
Pages (from-to)117-129
Number of pages13
JournalComputers and Geosciences
Volume114
DOIs
Publication statusPublished - May 2018

Keywords

  • DEM
  • DTA
  • Propagation
  • Terrain complexity
  • Truncation error
  • Uncertainty modelling
  • Volume calculation

ASJC Scopus subject areas

  • Information Systems
  • Computers in Earth Sciences

Fingerprint

Dive into the research topics of 'Uncertainty modelling and analysis of volume calculations based on a regular grid digital elevation model (DEM)'. Together they form a unique fingerprint.

Cite this