 |
| Titel |
Flood mapping by combining the strengths of optical and Sentinel active radar remote sensing |
| VerfasserIn |
H. C. Winsemius, G. R. Brakenridge, R. Westerhoff, J. Huizinga, N. Villars, C. Bishop |
| Konferenz |
EGU General Assembly 2012
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 14 (2012) |
| Datensatznummer |
250070679
|
|
|
|
|
|
| Zusammenfassung |
| Flood mapping with remote sensing plays an important role in large scale disaster
management procedures. For this purpose, the Dartmouth Flood Observatory (DFO)
gained experience since 1993 with the production of flood maps from optical satellite
imagery and has currently established, together with NASA collaborators, a fully
automated, global, near real-time service. Another consortium is also presently
working on an automated, near real-time, global flood mapping procedure called
the ’Global Flood Observatory’ (GFO), which will make use of high resolution
Sentinel data. The procedure is currently tested on Envisat active radar (ASAR)
imagery. Both the DFO and GFO projects provide open data output of their data and
maps.
The optical and radar approaches to flood mapping each have advantages and suffer from
shortcomings. Optical remote sensing via the U.S. MODIS and VIIRS sensors is constrained
by cloud cover but can attain a high revisit frequency (>2 /day), whereas the Envisat ASAR is
not affected by cloud cover, but uses a lower revisit frequency (generally once/3 days,
depending on the location). In this contribution, we demonstrate the combination of both
approaches into one flood mapping result. This results in improved flood mapping in a
case study over the Chao Phraya basin (Bangkok surroundings) during the recent
October-November 2011 extreme flooding. The combined map shows that during overpass,
ASAR reveals flooded regions over cloud-obscured areas, which clearly follow elevated
features in the landscape such as roads, embankments and railways. Meanwhile, the high
frequency of delivery of the optical information ensures timely information. Also, the
quite different water classification methods used for the optical and ASAR data
sources show good agreement and have been successfully merged into one GIS data
product. This can also be automatically generated and disseminated on a global
basis. |
|
|
|
|
|
|