We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.AP

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Applications

Title: Quality Assessment of the 20m SPOT DEM in Nigeria

Abstract: The 20m SPOT DEM (Digital Elevation Model) was acquired by the Office of the Surveyor-General of the Federation (OSGoF) in Nigeria for use in topographic mapping. A localized assessment of the DEM is needed to validate its stated accuracies as it is well known that the vertical accuracy of global DEMs varies in different landscape contexts. Hence, this study assessed the quality of the DEM in variable land cover types by comparing its heights against 780 Ground Control Points (GCPs) in Lagos State and the Federal Capital Territory (FCT) of Nigeria. The pattern of distribution of the height differences was analysed using spatial autocorrelation analysis, including other accuracy metrics employed. In the general accuracy assessment, the DEM yielded root mean square errors of 2.33m in Lagos and 3.69m in the FCT. It was shown that heights over the bare lands were the most accurately represented on the DEM while heights over built-up areas are the least accurate at both locations. Despite this, the SPOT DEM accuracies in the varying land cover types surpassed its stated global accuracy. While the spatial distribution of low and high height differences in Lagos State showed clustering, the converse was the case in the FCT. Following this assessment, it is recommended that the country should extend the application scope of the DEM in order to exploit its utility to the maximum.
Subjects: Applications (stat.AP); Other Computer Science (cs.OH)
Journal reference: Applied Geomatics volume 14, (2022)
DOI: 10.1007/s12518-021-00404-0
Cite as: arXiv:2205.02946 [stat.AP]
  (or arXiv:2205.02946v1 [stat.AP] for this version)

Submission history

From: Chukwuma Okolie [view email]
[v1] Thu, 5 May 2022 21:57:48 GMT (1559kb)

Link back to: arXiv, form interface, contact.