Share Email Print

Proceedings Paper

An assessment of the Height Above Nearest Drainage terrain descriptor for the thematic enhancement of automatic SAR-based flood monitoring services
Author(s): Candace Chow; André Twele; Sandro Martinis
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Flood extent maps derived from Synthetic Aperture Radar (SAR) data can communicate spatially-explicit information in a timely and cost-effective manner to support disaster management. Automated processing chains for SAR-based flood mapping have the potential to substantially reduce the critical time delay between the delivery of post-event satellite data and the subsequent provision of satellite derived crisis information to emergency management authorities. However, the accuracy of SAR-based flood mapping can vary drastically due to the prevalent land cover and topography of a given scene. While expert-based image interpretation with the consideration of contextual information can effectively isolate flood surface features, a fully-automated feature differentiation algorithm mainly based on the grey levels of a given pixel is comparatively more limited for features with similar SAR-backscattering characteristics. The inclusion of ancillary data in the automatic classification procedure can effectively reduce instances of misclassification. In this work, a near-global ‘Height Above Nearest Drainage’ (HAND) index [10] was calculated with digital elevation data and drainage directions from the HydroSHEDS mapping project [2]. The index can be used to separate flood-prone regions from areas with a low probability of flood occurrence. Based on the HAND-index, an exclusion mask was computed to reduce water look-alikes with respect to the hydrologictopographic setting. The applicability of this near-global ancillary data set for the thematic improvement of Sentinel-1 and TerraSAR-X based services for flood and surface water monitoring has been validated both qualitatively and quantitatively. Application of a HAND-based exclusion mask resulted in improvements to the classification accuracy of SAR scenes with high amounts of water look-alikes and considerable elevation differences.

Paper Details

Date Published: 25 October 2016
PDF: 11 pages
Proc. SPIE 9998, Remote Sensing for Agriculture, Ecosystems, and Hydrology XVIII, 999808 (25 October 2016); doi: 10.1117/12.2240766
Show Author Affiliations
Candace Chow, Univ. Bern (Switzerland)
André Twele, German Aerospace Ctr. (Germany)
Sandro Martinis, German Aerospace Ctr. (Germany)

Published in SPIE Proceedings Vol. 9998:
Remote Sensing for Agriculture, Ecosystems, and Hydrology XVIII
Christopher M. U. Neale; Antonino Maltese, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?