Flooding is one of the most frequent and damage causing natural disaster in Danang city well as other parts of Vietnam. Mitigation of flood hazard requires understanding about hydrological, meteorological as well as topographical condition of study area. This study aims to build flood inundation map for an alluvial lowland area using ALOS PALSAR data integrated with high resolution DEM. This case study area is located in the flood prone region in Danang city, Vietnam that used to experienced several flood events in past. Study area comprises of 98 km2 of lowland area located in the south of Danang City with elevation ranging from 0m to 10m. The data used for DEM generation is in-situ elevation points surveyed by Department of Natural Resource and Environment, Danang city in 2009. For flood inundation mapping, two ALOS PALSAR images in September 15th 2007 and October 31st 2007 that close to the major flood event in 2007 were collected with respect to the time before and during flood. In addition, Landsat TM imagery in 2007 was used to classify the land-cover of study area that related to flood hazard potential. Except for the DEM generation, all other processing was carried out using the GRASS GIS framework. In order to generate high resolution DEM, an interpolation algorithm utilizing equality-inequality constraint implemented in BS-Horizon FORTRAN program (Nonogaki et al., 2012) was used. This method is based on a bi-cubic spline algorithm and 5m resolution DEM was generated. The output DEM has elevation from 0-10.6m, the RMSE is only 0.16m. Compared to other interpolation methods such as Inverse Distance Weighting (IDW) or Regularized spline with tension (RST), Horizon-DEM provides a better representation of surface and the vertical error is also minimal. In order to develop the flood inundation map, the difference in Normalized Backscattering Coefficient between two PALSAR images was calculated. Subsequently, a threshold of 7dB was used to separate water and non-water objects. Topographic data was also used to remove all the noise flood pixels in the higher elevation area by applying a threshold of 5m elevation from DEM. Using this method, flood inundated area was separated from non-water areas and permanent water such as river channels or lakes. Finally, landform classification map was built based on land-cover derived from Landsat data, flood map and elevation information from DEM. Based on this landform map, a flood risk map was generated based on the probability of submergence of each landform unit. Landforms units derived rule-based classification of land cover map, 5m resolution DEM data and flood inundation map not only facilitates the understanding of the nature of flood but also in flood risk zoning. The methodology developed in this study would be useful in low relief areas in Vietnam and other parts of the world.
Copyright © 2015. All Rights Reserved.