Flood Hazard Assessment for the Construction of Flood Hazard Map and Land Development Priority Map Using NOAA/AVHRR Data and GIS - A Case Study in Bangladesh
Hazard Map
Flood hazard maps were constructed by considering the interactive effect of flood-affected frequency and flood depth on the land cover categories, physiographic and geological divisions. Flood Hazard Map using Flood-affected Frequency as Hydraulic Factor The flood-affected frequency map (Fig. 3) was constructed by using the images of September 18, 1988; October 31, 1995; and 18 September 1998, consisted of four classes --non-hazardous, low, medium and high damaged areas. Before considering the interactive effect of land cover categories, physiographic divisions and geologic divisions on the flood-affected frequency, each hazard map consisted of three ranks (HR 1-3), which were only developed by land cover categories or physiographic divisions or geologic divisions. Hazard ranks were considered from 1 to 27 after combining the hazard rank of land cover categories (HR 1- 3), physiographic divisions (HR 1-3) and geologic divisions (HR 1-3) simultaneously, using the ranking matrix of three-dimensional multiplication mode (Fig. 5). A model was considered for the assessment of the flood hazard. The schematic concept of the model is shown in Fig 6.  Fig. 6. Schematic concept of model Flood Hazard Map using Flood Depth as Hydraulic Factor Three flood depth maps were constructed by using three images of September 18, 1988; October 31, 1995; and September 18, 1998, respectively, which consisted of four classes --non-flooded area, shallow, medium and deep flood. Therefore, initially three different hazard maps were developed for three different flood depth maps by considering the interactive effect of land cover categories, physiographic divisions and geologic divisions on the flood depth, using the ranking matrix of three-dimensional multiplication mode (Fig. 6). Each hazard map consisted of hazard ranks ranging from 1 to 27. Table 3(a) and (b) show the comparison among the hazard maps of 1988 to those of 1995 and 1998, when only flood depths were considered independently. The rows in these tables present the area percentage occupied by the hazard ranks of the hazard map for the 1988 flood, while columns represent the area percentage occupied by the hazard ranks of the hazard maps developed by 1995 and 1998 flood, respectively. The tables show that flood hazard map developed for the flood depth of 1988 exhibits the deviation of the marginal distribution toward higher ranks among these three hazard maps. Therefore, the hazard map of 1988 event for the flood depth was selected as a hazard map for flood depth among the three hazard maps, because the histogram of the hazard map of 1988 shows many frequencies over higher hazard ranks among the three hazard maps. Design and development purposes for flood countermeasures considering higher hazard ranks provide a higher factor of safety. 
Considering the Interactive Effect of Flood-affected Frequency and Flood Depth Two flood hazard maps were developed by flood-affected frequency and flood depth, respectively, considering the flood hazard rank ranging from 1 to 27. Table 4 shows comparison between the area occupied by the same hazard rank, when the flood hazard maps were developed by flood depth and floodaffected frequency independently. The columns of the table show the area percentage occupied by the hazard rank of the hazard map developed by using flood depth, while rows of the table present the area percentage occupied by the hazard rank of the hazard map developed by using flood-affected frequency. In these two hazard maps, 56.31% areas exhibited the same hazard ranks and 43.69% were different, as shown in Table 4. Therefore, the authors need to integrate the flood hazard maps developed by considering flood-affected frequency and floodwater depth. Comparing between the hazard maps of flood-affected frequency and flood depth, the higher rank for a pixel was assigned for that pixel for the new developed hazard map. As a result, the newly developed hazard map represents the higher rank between the two ranks of two hazard maps for each pixel. Finally, more pixels were occupied by higher ranks of this hazard map. The flood hazard map developed by flood-affected frequency shows the deviation of the marginal distribution toward higher ranks between these two hazard maps. Consequently, it is to be said that the flood-affected frequency comparatively dominates higher hazard ranks for flood hazard map and land development priority map. Watercourses were not included to the developed flood hazard map. Therefore, the drainage map was overlaid onto the hazard map, and final hazard map with watercourses is shown in Fig. 7. Ten hazard ranks, 1, 2, 3, 4, 6, 8, 9, 12, 18, 27, ranging from 1 to 27, are shown, because the ranking matrix of threedimensional multiplication mode in Fig. 6 was used. Flood plain area of the Meghna River including Sylhet trough and floodplain area of the Brahmaputra River fall into highest flood hazard areas; these have a hazard rank of 27, and these areas also capture high flood-affected frequency (Fig 3) as well as deep or medium flood depth (Fig. 4). Flood plain area of the Ganges River, Faridpur trough and lower southwest part (tidal area) except for mangrove areas fall into higher hazard areas; hazard rank 12 and 18. Southeast lower part (eastern hill), northwest upper part and west part of lower flood plain of Ganges River show low hazard rank; hazard rank 1, 2, 3 and 4. 
LAND DEVELOPMENT PRIORITY MAP FOR FLOOD COUNTERMEASURE The major cities of Bangladesh are extremely highly populated. The planning of river works for flood countermeasure should be undertaken by considering the economical effects of the infrastructure and the importance of the concerned areas. Therefore, using flood hazard map and population density map, a land development priority map was developed. Digital population data was prepared using the population map of Bangladesh. Urban and industrial areas show highly dense population, while agricultural low lands and agricultural flat plains show low-density population. According to the population density, the digital population data was categorized into five zones. Areas that show a population density 1 to 500 per square kilometer were considered as zone 1, similarly, areas that show a population density 501 to 1000, 1001 to2500, 2501 to 4000 and over 4000 were considered as zone 2, zone 3, zone 4 and zone 5, respectively. Hazard ranks of flood hazard map were categorized into five groups. Hazard ranks 1, 2 and 3 are grouped as 1. Similarly, 4 and 6 are group 2; 8 and 9 are group 3; 12 and 18 are group 4, and 27 is group 5.  Fig. 7. Flood hazard map This new hazard-grouped map was incorporated with digital population categories map and finally, using ranking matrix of two-dimensional multiplication mode, a land development priority map was developed. Fig. 8 shows the land development priority map for flood countermeasures. The land development priority score (PS) range from 1 to 25. Higher score indicate that higher priority must be given for the development for flood countermeasure. Therefore, the highest score 25 shows the first priority (priority rank, PR=1) for the development, and score 1 shows the last priority for the land development. Therefore, land development priority rank (HR) ranges from 1 to 14 for the PS 25 to 1. A priority rank 1 indicates first priority for the development and then 2, 3, 4 and so on. This development priority map shows the priority score on the basis of pixel. Comparing between the hazard map (Fig. 7) and the development priority map (Fig. 8), it is understood that some high hazardous areas do not show the high score for the development. The northeast part of the Meghna River and southwest lower parts of Bangladesh show high hazard ranks, whereas the development map shows the low score for the development of the same areas. Some parts of Dhaka and Narayanganj districts show the higher development score due to the high-density population with high hazard area. Study has shown that Dhaka, the capital city of Bangladesh, was highly affected during the 1988 flood (Sado & Islam 1997).  Fig. 8. Land development priority map The hazard ranks for the administrative districts were estimated by using the mean value of the pixels belonging to the particular administrative district using the following equation 
where ni =number of pixels occupied by ith rank of hazard map for each administrative district; and hi= value of ith rank. Similarly, the development score rank for the administrative districts were estimated by using the mean value of the pixels belonging to the particular administrative district using the following equation 
where ni =number of pixels occupied by ith rank of development priority map for each administrative district; pi =value of ith rank. A comparison between the flood hazard ranks and development priority score is shown in Table 5. Comparing the hazard ranks and development priority score, it is found that some higher hazardous districts, Jamalpur, Netrokona and Sunamganj, do not show the higher score for development, because those areas are comparatively low-density populated areas. On the other hand, some lower hazard districts, Munshiganj, Dhaka and Chandpur, show a higher development score due to the highdense population, compared to the above-mentioned districts. 
CONCLUSIONS Flood hazard assessment can be performed using NOAA AVHRR data with physiographic, geological, elevation, administrative district and drainage network data. Flood-affected frequency and flood depth are essential components for the evaluation of flood hazard. In this study, categories of flood affected frequency and flood depth were estimated using NOAA satellite data. Flood hazard rank assessment was undertaken on the basis of land cover classification, physiographic divisions, geological divisions and administrative districts. The summarized results can be concluded as follows: Flood hazard assessments were undertaken and a new flood hazard map for Bangladesh was developed by using the flood events of 1988, 1995 and 1998; considering the interactive effect of flood-affected frequency and flood depth, those were estimated from NOAA AVHRR images of September 18, 1988; October 31, 1995; and September 18, 1998. The floods hazard maps based on each pixel and each administrative district represent the magnitude of flood damage for each pixel and each administrative district, respectively. This type of map helps the responsible authorities to better comprehend the inundation characteristics of the floodplains. The land development priority map for flood countermeasure was based on each pixel and each administrative district. Although flood hazard rank for some urban areas are comparatively less than the hazard for some rural areas, development should be undertaken for those urban areas (higher dense populated area) on a reasonable priority basis. The results described in this study should provide helpful information about flood risk management and should be useful in assigning priority for the development of very high risk areas for flood control planning, and the construction and development of flood countermeasures. In addition, this study may have considerable management implications for emergency preparedness, including aid and relief operation in high risk areas in the future. Finally, these types of flood hazard and land development priority map in digital form can be used as a database to be shared among the various government and non-government agencies responsible for the construction and development of flood defence. REFERENCES
Page 1 of 3
| Next | |