Quantitative dynamic flood monitoring with NOAA AVHRR
The analysis of the spatial extent and temporal pattern of flood inundation from remotely sensed imagery is of critical importance to flood mitigation. With a high frequency of global coverage, NOAA/AVHRR has the advantage of detecting flood dynamics during devastating floods. In this paper, we describe a systematic approach to flood monitoring using AVHRR data. Four critical issues for successful flood monitoring with AVHRR were identified: correct identification of water bodies, effective reduction of cloud contamination, accurate area estimation of flood extent, and dynamic monitoring of flood processes. In accordance with the spectral characteristics of water and land in AVHRR channels, a simple but effective water identification method was developed with the ability to reduce cloud influences. The area of flooded regions was calculated with the consideration of areal distortion due to map projection, and mixed pixels at water/land boundaries. Flood dynamics were analysed from flood distributions in both space and time. The maximum spatial extent of floods, generated by compiling the time series of flood maps, was informative about flood damages. We report a successful use of this approach to monitor the Huaihe river flood, a centennial devastating disaster occurred in the Huaihe river basin of China in the summer of 1991.
No Reference information available - sign in for access.
No Citation information available - sign in for access.
No Supplementary Data.
No Article Media