Flood Risk Assessment

1

Flood Risk Assessment using Remote Sensing and
Hydrodynamic Model in Xiang Jiang River, Hunan,
China
2
Study Area

• The Changsha channel segment of Xiang Jiang River
  is selected as study area in Hunan Province of
  China.

3
Objectives

• To prepare flood hazard map using Hydro-dynamic
  model (HEC-RAS) and satellite images.
• To develop a methodology and model for flood risk
  assessment using GIS.

4
Methodology
DEM
Hydrologic data
JERS-1
LANDSAT/ASTER/SPOT
Socio-Economic data
Hydro-dynamic model (HEC-RAS)
Population
Land use
Flood area
Flood depth
Population density map
Economic Value map
Flood frequency
Assessment method
Hazard factors data
Vulnerability factors data
Flood risk assessment model
Flood risk assessment
Flood risk analysis
Flood risk validation
Flood risk map
5
Data

• Satellite images

JERS/SAR (1998)
LANDSAT-ETM (1999)
JERS/VNIR (1997)

• Hydrologic data
• Topography maps
• Land use maps
• Socio-economic data

6
Flood Area Extraction from Satellite Images - 1
Histogram of JERS-SAR
JERS-SAR Image
Water extent extraction from JERS-SAR image
during flood
7
Flood Area Extraction from Satellite Images - 2
Water extent extraction from LANDSAT/ETM image
before flood
Histogram of band 7
Landsat-band743
Landsat-band7
8
Flood Area Extraction from Satellite Images - 3
Water extent before flood
Water extent during flood
Flood area extracted from 2 satellite images
(blue is for normal water and red for flood
extent)
9
Flood Modeling-Results from HEC-RAS
Cross section plot
Profile plot from 1997 to 2005
X-Y-Z Perspective Plot
Cross section output
10
Flood plain visualization-inundation area and
flood depth
1997
2000
1998
1999
2001
2005
2002
2003
11
Flood Experience Calculation
Flood experience for every pixel was calculated
by overlaying inundation area maps of eight years.
12
Land use Classification
Land use classification system at 2 levels, using
Object-based classification method and manual
editing
Land use map
13
Weighted Population Density Map
14
Hazard and Vulnerability
Classification criteria
15
Results
16
Flood Risk Map
17
Conclusions

• Risk to the affected population was assessed and
  it was found that out of 328,500 people of the
  study area about 65,000 were sited in high risk
  area.
• Combination of Remote Sensing, GIS and
  Hydrodynamic Model for flood risk assessment
  could reproduce the physical processes better by
  compensating the weaknesses of one by the
  strengths of others.
• Model builder of ArcGIS is an effective tool for
  Building flood risk assessment model.