Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

5.1 Introduction

Although there is a sharp difference between the manifestation of hazards and disasters, disaster research currently receives more attention than it did in the past. While hazard studies examine the beneficial and adverse effects of extreme events on society, disaster research focuses more on causes, physical risks, and emergency responses (White et al. 2001). Therefore, similar to other key terminologies of hazard literature, the definition of hazard varies. Nevertheless, identifying potential hazard areas is an integral part of the flood risk assessment process. Considerable efforts have been made worldwide to assess flood hazards using geospatial techniques (see Chap. 2). The primary intent of such efforts is to delineate hazardous areas, which can then assist in developing appropriate measures to lessen the negative effects of floods on people and the economy. A coarse- to moderate-resolution flood hazard map may have little relevance in developed countries; however, in developing countries, where precise information is seriously lacking, such hazard maps are extremely useful for saving the lives and property of millions of people, particularly marginal groups.

In this chapter, flood hazard assessment is presented by the integrated use of GIS and RS. To facilitate flood hazard analysis, flood frequency and floodwater depth maps were constructed from satellite and GIS data. Sect. 5.2 describes the evaluation of flood hazards using flood-affected frequency. DEM and HWL data of major rivers in the study area were used to derive multi-temporal flood depth maps (see Chap. 4), which were used to identify hazard areas for the flood depth category (see Sect. 5.3). Finally, in Sect. 5.4, the modeling of flood hazards was conducted by considering a combination of flood depth and flood-affected frequency.

5.2 Analysis of Flood Frequency

Several hydrological parameters, such as the depth of flooding, rate of water-level rises, flood frequency, water velocity, physical exposure of land, and sediment loads, are influencing flood hazards of a particular site. As all parameters were not available, two hydraulic components (e.g., flood frequency and floodwater depth) were considered in this study for the evaluation and determination of potential flood hazard areas.

The concept of flood-affected frequency (Islam and Sado 2000) is used to assess flood hazards. The classified images obtained from the threshold algorithm (see Chap. 4) were used for this purpose. Nine images were classified into water and non-water areas and were lumped together to generate a flood-affected frequency map. Initially, the classified images of 1988, 1998, and 2000 were combined to construct a single flood frequency map. Likewise, the flood extent maps of 2001, 2003, and 2004 were combined, and the process continues for the rest of the years. This procedure resulted in a three-flood extent map, which provided an opportunity to obtain a common boundary essential to developing a flood-affected frequency map (see Fig. 5.1).

Fig. 5.1
figure 00051

Concept of flood-affected frequency (Adapted from Islam and Sado 2000)

Full size image

The inundation area that appeared in all of the images considered the highly affected areas and therefore a potentially high-damage zone. The common inundated areas that appeared in two and one of the three maps were deemed medium- and low-damage areas, respectively. An inundated area that was not present in any of the images was classified as a non-flooded area (see Fig. 5.2a). The final flood-affected frequency map was reclassified into four categories according to the severity of inundation, corresponding to flood rankings of class 1, class 2, class 3, and class 4 as non-flood, low, medium, and highly hazardous areas, respectively (see Table 5.1).

Fig. 5.2
figure 00052

(a) Flood-affected frequency and (b) floodwater depth

Full size image
Table 5.1 Classification of flood-affected frequency
Full size table

The flood-affected frequency map revealed that at least 23% of the study area was in a high-hazard zone that is subject to regular inundation. More than 42% of the area was in a non-flooded category that was not inundated by river waters, 13.3% was in a medium-flooded zone, and 20.5% was in a low-flooded category (see Table 5.1).

5.3 Analysis of Floodwater Depth

As described in Chap. 4, a series of flood depth maps were developed using a DEM with the peak water-level data of major rivers. These flood depths were reclassified into four categories: no flooding, shallow, medium, and deep. A rule-based approach was used in the model-builder utility of ArcGIS (v. 10) to derive a floodwater depth map for the study area. If a deep flood is observed in a single image and a medium flood is observed in at least one of the other images, then it was regarded as deep; from the remaining pixels, if it appeared medium in two images, then it was considered medium. If it appeared shallow in two images, then it was considered shallow. After the classification of the deep, medium, and shallow depth categories, the remaining inundated areas were considered non-inundated or non-flooded. Thus, four floodwater depth categories were obtained from the model, corresponding to depth rankings of class 1, class 2, class 3, and class 4 as non-flooded areas, shallow, medium, and deep (see Fig. 5.2b), respectively.

The total area and percentage occupied by each depth category is shown in Table 5.2. It revealed that 38% of the study area is within the category of no flooding. Shallow, medium, and deep flood depth categories represent 15, 1.5, and 45%, respectively. It should be noted that during the calculation of the depth map, ongoing land development by means of earth/sand filling was not considered. As land development by sand filling is a popular practice that became widespread recently to raise land levels from surrounding river levels, the phenomenon may affect the depth of flooding in the study area. Therefore, the depth maps developed here only represent the condition of land during the creation of the elevation model of the study area by the concerned authority. To validate this study’s result, the final flood depth map was superimposed on the flood-affected frequency map, and the difference was found to be only 2% in terms of flood-affected areas. This may suggest the efficacy of these two hydrologic parameters for flood hazard assessment.

Table 5.2 Area occupied by each category of floodwater depth
Full size table

5.4 Development of the Flood Hazard Map

Even though geospatial techniques, particularly RS, has shown great potential for hazard assessments (McKean et al. 1991), the big challenge is the lack of generally accepted methods for producing hazard maps (Rhoads 1986). Moreover, many well-developed methods in developed countries may have little relevance to developing countries such as Bangladesh due to a lack of data and/or restricted access to data sources. Therefore, the availability of information and the context of the area being studied are important considerations when evaluating flood hazards. In this study, a simple 2-D multiplication-ranking matrix was applied by combining flood frequency and flood depth maps (Fig. 5.3).

Fig. 5.3
figure 00053

Calculation of flood hazard

Full size image

The flood-affected frequency map and floodwater depth maps were combined to derive the final flood hazard map. A 2-D multiplication-ranking matrix (see Table 5.3) suggested by Ochi et al. (1991) was used for flood hazard ranking. The guiding principle of this technique was that flood hazards for a particular cell increase in a nonlinear manner, which depends on both flood occurrence and depth. If a cell represents non-flooded in both maps, then it was considered a non-hazard zone. Low-flooded areas in the flood frequency map and shallow depth in the floodwater depth map were ranked as low-hazard areas. Similarly, medium-flooded areas and medium depth were categorized as a medium-hazard zone. Pixels representing high-flooded areas and deep flooding were considered a high-hazard zone. This operation on two thematic layers resulted in a new map with 16 unique categories. Based on hazard intensity, the derived flood hazard map was reclassified into four hazard categories corresponding to class 1, class 2, class 3, and class 4 as no-hazard, low-hazard, medium-hazard, and high hazard zones, respectively (see Fig. 5.4).

Table 5.3 Ranking matrix for two thematic hydraulic parameters
Full size table
Fig. 5.4
figure 00054

Flood hazard map derived from flood-affected frequency and floodwater depth simultaneously

Full size image

The final flood hazard map was intersected with population and housing data to estimate the exposure of these two elements to flood hazards. Table 5.4 shows the distribution of the population in different flood hazard zones. The analysis revealed that 23.7% of the population was exposed to low–high flood hazards, while 76.3% were located in the no-hazard zone. Among the total population exposed, 24% were elderly, 25% were children, and 24% were females (see Table 5.4). Note that the data used to estimate population exposure are based on the 2001 population census; hence, the actual populations in the different flood hazard zones may be greater than this study.

Table 5.4 Exposure of the population to flood hazards
Full size table

Exposure of housing in different hazard zones is presented in Table 5.5. This shows that more than 25% of total houses in the study area are exposed to flood hazards, of which 6% is exposed to a high-hazard zone. The estimation of various housing types reveals that 17% of katcha houses are exposed to medium- to high-hazard categories, while 18% is in the low-hazard zone, revealing the vulnerability of human settlements to floods in Dhaka. Likewise, more than 7% of semi-pucca houses are exposed to medium–high flood hazards. Pucca houses, which are usually homes to wealthy people, are also exposed to flood hazards, as shown in Table 5.5. As housing data are based on the year 2004–2005, the actual exposure of human settlements to floods may be higher given that Dhaka is expanding rapidly.

Table 5.5 Percentage of human settlements in different flood hazard zones
Full size table

5.5 Summary

This chapter presented the evaluation of flood hazards in the study area by considering flood frequency and flood depth data obtained from GIS and RS images. The study used a simple and cost-effective method of utilizing the GIS to create the flood hazard map from the available dataset.

Using the flood extent maps of 1988–2009, a flood-affected frequency map was constructed, which showed that at least 23% of the area was in a high-flooded zone. In contrast, the flood depth map revealed that 45% of the area was in a deeply flooded zone. Using a 2-D multiplication-ranking matrix, a flood hazard map was derived by considering both flood depth and flood-affected frequency simultaneously.

The exposure analysis revealed that 23% of the total population in the study area was exposed to low- to high-hazard zones. Among the populations exposed, the elderly population constituted 24%, children represented 25%, and females comprised of 24%. This clearly indicates that floods remain a significant threat to the people of Dhaka. In addition, 25% of the total housing units comprising katcha, semi-pucca, and pucca were exposed to flood hazards, of which 6% were in the high-hazard zone. Of the housing types, 18% of katcha houses were exposed to the low-hazard zone, 7% were exposed to the medium-hazard zone, and 9% were exposed to the high-hazard zone. As these houses are home to low-income people, the potential effects of floods remain a significant threat for this group. Similarly, 4% of semi-pucca houses and 3% of pucca houses were exposed to the high-hazard zone, revealing that people in the study area are either not aware of flood hazards or, if they are, they have few alternatives to settling in the high-hazard zone.

As observed in Dhaka, increasing population pressure is forcing many people to move to the vacant land by filling in surface water bodies and floodplains. Consequently, the flood hazard is increasing. To ameliorate flood-induced damage, the derived flood hazard map is invaluable. Urban planners can use this information to make environmentally sound land-use decisions. An important advantage is that the developed hazard map can quickly be updated if modifications occur subsequent to the original study (e.g., changes in land use).