Wang Yanyan, Lu Jikang, Xiang Liyun and Chen Hao
(China Institute of Water Resources and Hydropower
Research (IWHR), Beijing, 100038)
Abstract: Based on GIS, an Urban Flood Damage Evaluation System (UFDES) for Shanghai City has been developed, which includes spatial geographic analysis, database, numerical simulation and modern statistical techniques. Distributions of population and assets, as well as calculation results, can be represented. Incorporated with urban flood numerical simulation model, an integrated spatial Model for Urban Flood Damage Evaluation (UFDEM) is established. Legible spatial queries and visual thematic maps for flood characteristics and asset losses are accomplished. As the result, some important reference could be suggested for decision-making on urban flood disaster reduction.
Keywords: database, urban flood damage evaluation, spatial analysis, GIS
As a basic research in the field of urban flood-control and disaster-relief, urban flood damage evaluation plays an important role in flood protection decision-making, flood risk management, urban planning, laws and regulations drafting and so on[1]. With the rapid urbanization and much concern on the flood, especially for the sustained flood disaster events recently, people have attached importance to urban flood mitigation. The study of urban flood damage evaluation is more vigorous than before, and the rationality has been improved due to the application of advanced technology and methods such as spatial geographic analysis, database techniques, numerical simulation and so on[2,3]. UFDES for Shanghai City based on GIS is a representative example[4].
Located in the east China-Yangtze delta and the downstream of Taihu basin, Shanghai City is adjacent to the sea with two sides and to the river with one side(see Fig1). Influenced by the tide and monsoon climate and with the upstream flood passing, such catastrophic weather as tropic storms and cloudburst which cause much damage to Shanghai, often happen in this area, and endanger the people’s livelihood and restrain the industry development and business prosperity. Trusted by Shanghai Flood Risk Information Center, we developed UFDES of Shanghai City upon MapInfo platform. Based on GIS, the UFDES imports particular spatial retrieving technique to the previous system in analyzing flood damage, with a lot of statistic data being spatially located and the output containing much spatial information. As a result, it can not only evaluate the effects and the damage of flood, but also shows the losses distribution. Contrasting to the traditional statistic estimation which is deficient in reflecting the spatial attributes of data, this mode can be more efficiently competed for flood-control and flood-mitigation and serves flood protection decision-making well[5].
The urban assets, with their uneven distributions on space, are various and complicated. Although city economic statistical yearbook can provide much valuable information, its spatial scale is somewhat larger, and can not fully meet the need for spatial analysis of assets. Thus, based on statistic of city economy, a kind of method, through spatial analysis of blocks and selecting a few characteristic indexes, is set up with more generality and higher accuracy[6]. The basic work includes: ① Digitizing paper maps in vector format and separating different characters by layer/level and color or attribute code; ② Dividing different block units from the block map layer, giving them the index numbers, calculating their areas, and putting all of the information into the databases; ③ Evaluating these blocks by some economic attributions (e.g. Prices of lands, building density etc), and performing the spatial retrieving solution for the assets on blocks.
It is considered that the density of population
corresponds to the density of urban buildings, thus we can calculate out the
population on a block by acquiring the proportion
of the block's building area to the entire region's building area.
(1)
Where, Pi, Ps - the population on the ith block and the total population in the entire area, respectively; Bi, γi - the area of the ith block and the building density of this block, respectively; n - the total of block units in the entire region calculated.
By combining the distribution of population with the statistic of per capita living space and family property, the spatial distribution of house and family properties can be attained accordingly.
Two patterns are adopted for distribution of factorial / commercial corporation, one of which is called “point pattern” for representing the distributions of larger corporations, the other is called “region pattern” for a good many smaller ones.
For those larger corporations, their geographic information is directly positioned on maps connecting with their general attributions. Thus, spatial database is created for describing their features. For smaller corporations, the region pattern means that their assets will be converted to block units. In this sense, this pattern is somewhat similar to that of population or family property, but there is a big difference on how to select characteristic indexes. In order to represent the spatial distribution of these smaller corporations’ assets more accurately, non-inhabitant land percentage and the land price of a block will be taken as major factors to characterize the assets conditions, particularly for those commercial corporations. Thus, the assets of smaller corporations in a block can be gotten through the following equation:
(2)
Where, XQi - the asset of smaller corporations in the ith block unit; XQs- the total assets of smaller corporations in the entire region calculated; Mi' - the area of non-inhabitant land in the ith block unit; and Mi' = Bi·ωi- Mi·α1- Mi·α2/k; In which Bi is the area of the ith block unit; ωi is the building density of the ith block; Mi is the area of dwelling in the ith block; α1,α2 is the percentage of low/high dwelling; k is the average floor number of high dwelling; ΣMi' - the total area of non-inhabitant land in the entire region; Ci - the adjustment index for the ith block according its land price; Ci =Lvi / Lv, Lvi and Lv represent the land price of the ith block unit and average land price, respectively.
The features of the damage evaluation model are shown as following (see Fig2): ① A powerful database with spatial information supplies a strong support to damage evaluation and economy prediction, and provides the vivid spatial exhibitions for different information; ② Dynamically linking with a flood numerical simulation model, it makes real-time evaluation with high accuracy as well as prediction of losses; ③ Using modern mathematical methods such as statistic, relative analysis and systematic analysis etc., to constitute evaluation relationships, the inner mechanism of asset loss evaluation under different flood conditions is more accurately presented; ④ With a special graph system based on GIS, spatial operations for information query and analysis become much easier; ⑤ All the sub-systems are integrated into a comprehensive application system, which offers a great convenience to decision-making of disaster mitigation.
Fig.2 The framework of
UFDEM
The system aims at the following purposes: ① Using GIS analysis to perform spatial distribution and diffusion of the information, and so provide the supports for economy and damage evaluation; ② Setting up a dynamic link between damage evaluation model and flood numerical simulation model to exchange the data by geographic conversion and overlay of the information; ③ Executing the information spatial query, statistical analysis and thematic mapping, and displaying the calculation results distinctly, i.e. The distribution of population and assets, flood situations under different risk frequencies, and the distributions of asset losses etc.
Designing the system complies with the principles as follows: ① According to the framework of UFDEM, the system should be an independent and united system from database analyzing to model operating and result outputting; ② The design of model linkage must be proper to insure the whole system running smoothly and rapidly; ③ The system has friendly user-computer interface for easy operation; ④ The system should be expendable and possess high security and good error traps. The structure of the system is shown in Fig.3.
Fig.3 The structure of
Shanghai UFDES
Flood Database
Information is the basis for damage evaluation.
Urban flood database collects such information as city economy, loss cases and
flood control engineering etc. In the meantime, it also includes spatial
database of different objects, such as blocks, green lands, rivers, roads,
bridges, corporations, schools, hospitals, shopping malls and so on, all of
which can be read and picked up from digital base maps. The database should
offer supports to economy prediction, damage evaluation and the output of
thematic maps. The function of the database system includes: ① Data update and maintenance, such as data's input, edit,
modification, index and browse; ② Information query, reports output and statistic analysis; ③ Graph operation for the results of statistic analysis; ④ Files management
and on-line help.
Running of UFDEM
As the key point of evaluation model, relationships between losses and flood characters, which is affected by many uncertain factors, is partly determined by practical experiences. So the parameter setting window is designed for users to modify the parameter required for establishing the relationship, which helps to enhance the model accuracy.
The calculation results of simulation model with spatial location, such as water depth/level, velocity, duration and arriving time, are directly applied by the evaluation model. On the other hand, considering that flood information may be attained by survey or statistics, real-time mapping function is designed for users to delineate inundated region as a new layer over base map and add the corresponding flood features, which is similar to the simulation results. To whichever mode, UFDEM, by spatially overlapping between flood information and socioeconomic database, can rapidly evaluate the flood damage. The flow chart of the running process of the model is shown in Fig.4.
Fig.4 The flow chart of
the running process of UFDEM
Spatial query and thematic mapping
By applying the powerful ability of GIS in managing large volumes of spatial data, legible spatial query function is developed for user to inquire the concerning information more easily and more expediently. It mainly involves the following five aspects: ① Query for socioeconomic information, flood control engineering, historical flood information, and other basic information such as topographic characteristics, rivers and so on; ② Query for flood situations, such as the maximum submerged range, water depth/level, velocity, duration and arriving time etc; ③ Query for population and assets inundated by flood; ④ Query for the calculation result of asset losses and loss rate. Users can make queries to get the above information at any concerned areas to which pattern the user would prefer.
To learn the effect of the flood completely and clearly, users not only concern the numerical value of the calculation results, but also concern the spatial distribution of flood characters, population and asset losses affected by flood. Thematic maps, which help users to find the tendency and mode hardly detected from datasheet, is considered an important means to reflect the spatial distribution of information. In this system, thematic maps consist of the following types: ① Maps of population and assets; ② Maps of flood characteristic; ③ Maps of inundated population and assets; ④ Maps of asset losses. The thematic map of submerged block area is shown in Fig.5. With transition color from deep red to light yellow to fill different blocks, the region covered with the deepest red represents the block with the maximum inundated area and the region covered with the lightest yellow represents the block with the minimum inundated area.
Fig.5 The thematic map of submerged block area
As an important system analyzing and processing information, Flood GIS should provide strong supports to damage evaluation as well as flood numerical simulation. In order to meet these needs, We choose MapInfo as the development platform, using MapBasic and VisualBasic as the main language to develop the system, which make the system operate easily and quickly with high accuracy and good error traps.
Spatial analysis technologies, especially GIS,
have recently been widely used in flood risk study and their effects are
remarkable and satisfactory. Based on GIS and implementing the functions of
spatial retrieving, spatial analysis and spatial query, UFDES of Shanghai City
developed by IWHR makes the management of data more efficient, the application
of UFDEM more practical, calculation results more accurate and the information
displaying more vivid. Consequently, this system supplies an automatic and
scientific means to help decision-making on Shanghai flood disaster reduction
and serves flood risk management preferably.
Acknowledgement
The development of UFDES was supported by Shanghai Flood Risk Information Center and Beijing Light & Sound Electric Technology Company. We also express thanks to our colleagues of Research Center on Disaster and Environment of IWHR for their helps.
References
[1] Chen Xiuwan, Flood Damge Evaluation System-Research on Remote Sensing and GIS Application, China Water Power Press, 1999 (in Chinese).
[2] Wang Yanyan and Lu Jikang: Application of Flood Damage Evaluation Technology, 3nd National Urban Water Resources Conference, Chengdu, Sepetember, 2000 (in Chinese).
[3] Chen Shupeng and Zhong Ershun, Perspectives on GIS Development in China, 2nd Annual GIS Infrastructure Planning and Management Conference, Beijing, May, 1998 (in English).
[4] Wang Yanyan and Lu Jikang et al: Research on Shanghai Flood damage evaluation, Report of China Institute of Water Resources and Hydropower Research, 2000 (in Chinese).
[5] Wan Qing et al:Analysis and Evaluation of Flood Disaster System, Science Process, 1999 (in Chinese).
[6] Chen Hao and Xiang Liyun et al:
Formative and Developing Mode and Damage Evaluation of Urban Flood Disaster,
Research Report of China Institute of Water Resources and Hydropower Research,
1997 (in Chinese).