Premium
Assessment of the mapping capabilities of ERS‐1 SAR data for flood mapping: a case study in Germany
Author(s) -
OBERSTADLER R.,
HÖNSCH H.,
HUTH D.
Publication year - 1997
Publication title -
hydrological processes
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.222
H-Index - 161
eISSN - 1099-1085
pISSN - 0885-6087
DOI - 10.1002/(sici)1099-1085(199708)11:10<1415::aid-hyp532>3.0.co;2-2
Subject(s) - flood myth , computer science , remote sensing , environmental science , hydrology (agriculture) , geology , geography , archaeology , geotechnical engineering
GAF examined, under contract to DARA (German Space Agency), the applicability of ERS‐1 SAR data for flood mapping under operational conditions. The flood event investigated was the flooding in the Rhine valley in winter 1993–1994. In order to carry out an examination close to the end‐user needs, the specific user requests concerning information about flood events were identified. The mapping accuracy in view of the flood extent and the flood level, the production of flood maps as well as the demonstration of the runoff turned out to be the most interesting points. The specific user information needs were considered in the project objectives to define the applicability as well as the deficits of ERS‐1 data concerning an operational use for flood mapping. After a detailed analysis of the time aspects of the traditional mapping method and a satellite data analysis, a visual interpretation as well as an automatic classification were applied, including various filter steps to derive the flood boundary. As a result, the visual interpretation proved to be the more accurate method. Crucial domains for both the visual interpretation and the automatic classification turned out to be settlements, forests and bushes as well as regions with layover and foreshortening effects. The comparison between the flood level derived from satellite data and the flood level registered by the water authority boards brought a height difference which ranged between 0·5 and 2·0 m. The relatively coarse resolution and problems with correct interpretation of the flood line proved to be the reason for this difference. In general the results are convenient, but in relation to field measurements of the water level they are too inaccurate. A cost and benefit analysis as well as a proposal for an operational GIS system using ERS‐1 SAR data are still under investigation. © 1997 John Wiley & Sons, Ltd.