Back to Search Start Over

Evaluation of the Horizontal Resolution of SRTM Elevation Data

Authors :
Josef Kellndorf
Wayne S. Walker
Leland Pierce
Oton Barros
Source :
Photogrammetric Engineering & Remote Sensing. 72:1235-1244
Publication Year :
2006
Publisher :
American Society for Photogrammetry and Remote Sensing, 2006.

Abstract

The SRTM dataset is available at the UsGs seamless website with one arc-second pixel spacing for the U.S. Recently, the value for horizontal resolution has been questioned. One paper (Smith and Sandwell, 2003) suggests that 60 meters may be more accurate, implying that the resolution is twice the provided spacing. For users of this data, the horizontal resolution is very important for their analyses. Hence, this paper addresses this important question by using two different approaches: coherence spectra and step-response. The coherence spectra approach uses statistical techniques to compare the SRTM dataset against a more accurate one, while the step response approach uses the observed step response in many areas of the dataset to estimate the width of the averaging function used to produce the SRTM data. Results from this study show that the resolution is between 1 and 1.6 pixels, depending on the local variability of the elevation data; with higher resolution near sharp edges and corners, and lower resolution in smoother areas.

Details

ISSN :
00991112
Volume :
72
Database :
OpenAIRE
Journal :
Photogrammetric Engineering & Remote Sensing
Accession number :
edsair.doi...........283826ac691b9d0710b0310b19ca1ab4