Review of the Literature Introduction Our military and other government components such as the Department of Defense develop technology to keep United States citizens safe by improving on existing methods that identify and stop combatants. One method used to identify a suspect is by recovery DNA. When an improvised explosive device (IED) is detonated bomb fragments are recovered and examined in order to attempt to determine the identity of the bomb maker. Iris, facial, and vascular recognition are other biometrics tools used to identify our adversaries. Linking forensic functions with biometric capabilities is a relatively new form of technology and is discussed in the literature presented. Techniques According to a study by Chirchi, Waghmare, and Chirchi (2011), choosing the proper biometric tool to fit the specific situation requires knowledge of technological developments. One such development is the iris scan. Found to be a reliable form of authentication the military has evolved this form of biometric identification into a portable tool on the battlefield. The biometric automated toolset (BAT) is the primary system used by the U.S Central Command to store biometric data such as iris scans, (D’Agostino, 2008). The iris scan is a unique form of identification. In its genetic properties no two eyes are the same and furthermore the characteristic that is dependent on genetics is the pigmentation of the iris, (Chirchi, Waghmare, and Chirchi, 2011). Although not less reliable but a less developed form of biometric identification is facial recognition. It utilizes automated methods to verify the identity of a person based on physiological characteristics. Tolba, El-Baz, and ElHarby (2011) describe facial recognition as a way to detect facial patterns even in a crowded scene using classification algorithms. A computer algorithm “normalizes” the biometric signature so that it is in the same format as the signatures on the system’s database (Tolba, 2011). Facial recognition is seen as a convenient biometric tool due to being both machine-readable and human readable. The ubiquity of surveillance cameras means that, in a sense, a face can leave a trace and therefore be useful forensically, as are DNA and fingerprints, (DOD, 2007). Methods A significant tool in biometric identification is the use of DNA analysis, particular with recovering fingerprints. Esslinger, Siegel, Spillane, and Stallworth, (2004) research involved using short tandem repeat (SRT) analysis to detect human DNA from exploded pipe bomb devices. The effect on the DNA left on the components correlated with the material the pipe was made of (pvc vs. steel), the fragmentation pattern, and low vs. high explosives. One issue I noticed and it was briefly mentioned in the article, was with the reliability of the material the pipes were made. Steel is known to conduct heat better than PVC. The theory was since steel generates more heat during an explosion the chance for degradation of the DNA would increase. However since steel is more durable than PVC the percentage of larger fragments should increase. The more fragments, the more DNA could be collected. The data from the experiment showed the steel and PVC pipes had a similar success rate for DNA recover. Foran, Gehring, and Stallworth (2009) research included the recovery and analysis of mitochondrial DNA (mtDNA) from exploded pipe bombs. The importance and difference from STR analysis is that mtDNA analysis allows DNA that has been extracted from hair, fingernails, and bone to be examined when nuclear DNA cannot be recovered. Another significant difference is mtDNA sampling can be obtained from not only the subject but also related family members. The article discussed the materials and methods used in the test as well as the resulting bomb fragmentation and the correlation with the quality and quantity of DNA recovered. The results of the study showed the value of mtDNA analysis in identifying the manufactures of various detonated IEDs. Recovering fingerprints and other forms of DNA from various surface areas is not always textbook. Elements such as temperature, humidity, moisture, and material of surface area all affect the quality and ability to recover DNA. Shalhoub et al, (2008) researched a fast curing silicone-casting material (Isomark) as an effective method to obtain a reliable DNA profile from the casts of the fingerprints. Participants were asked to handle six different surfaces of various textures. This study was significant because various items are often used in IEDs that serve as projectiles. The Army field manual FM 3-34.119 (2005) describes various casings used such as pipes, soda cans, metal containers, all which turn into projectiles when detonated. Once recovered contents inside such as marbles, nails, rocks, and glass can all be examined for DNA. Through their research Shalhoub et al, (2008) concluded it was possible to recover DNA from Isomark casts made on all substrates tested. However, no link was noted between quality of finger marks obtained and the amount of DNA extracted from them, Shalhoub (2008). Summary Although the research discovered additional technology questions the research summaries concluded favorable results for recovering DNA from bomb components leading to identifying the bomb maker. Biometrics tools such as iris scanning, facial recognition, and fingerprinting are valuable components to identifying our adversaries and using that intelligence to mitigate against future attacks. References Chirchi, V., Waghmar, L.M., & Chirchi, E.R. (2011). Iris biometric recognition for person identification in security systems. International Journal of Computer Applications, 24(9). Retrieved August 25, 2011 from www.ijcaonline.org/volume24/number9/pxc3874002.pdf - India D’Agostino, D. 2008. Defense management: DoD can establish more guidance for biometrics. Retrieved October 2, 2011 from http://books.google.com/books?id=6hEMWW1M6osC&lpg=PA1&ots=wuEqgZCwb J&dq=biometric%20automated%20toolset&lr&pg=PP1#v=onepage&q=biometric %20automated%20toolset&f=false Department of Defense. 2007. Report of the defense science board task force on defense biometrics. Retrieved October 2, 2011 from www.fas.org/irp/agency/dod/dsb/biometrics.pdf Department of Defense. (2009). Biometrics task force annual report FY09. Retrieved September 4, 2011 from www.biometrics.dod.mil/Files/Documents/AnnualReports/fy09.pdf Esslinger, K., Siegel, J., Spillane, H., & Stallworth, S. (2004). Using STR analysis to detect human DNA from exploded pipe bomb devices. Journal of Forensic Science, 49(3). Retrieved September 7, 2011 from www.hartnell.edu/faculty/jhughey/Files/strpipebombanalysis.pdf Federal Bureau of Investigation (FBI). n.d. Terrorist explosive device analytical center (TEDAC). Retrieved September 15, 2011 from http://www.fbi.gov/aboutus/lab/tedac Foran, D., Gehring, M., & Stallworth, S. (2009). The recovery and analysis of mitochondrial DNA from exploded pipe bombs. Journal of Forensic Science (54)1. Retrieved September 7, 2011 from http://forbio.msu.edu/Recovery%20of%20mtDNA%20from%20exploded%20pipe %20bombs.pdf Makarski, R., Marrero, J. (2002). A surveillance society and the conflict state: leveraging ubiquitous surveillance and biometrics technology to improve homeland security. Retrieved September 4, 2011 from https://docs.google.com/viewer?a=v&pid=gmail&attid=0.1.1&thid=13237369b564 1eb4&mt=application/pdf&url=https://mail.google.com/mail/?ui%3D2%26ik%3D e063aef897%26view%3Datt%26th%3D13237369b5641eb4%26attid%3D0.1.1%2 6disp%3Dsafe%26zw&sig=AHIEtbRZ-Doe_xeF9h01W26wPdCmqr6Wng National Science and Technology Council (NSTC). 2008. Biometrics in government in post 9-11. Retrieved September 4, 2011 from www.biometrics.gov/.../Biometrics%20in%20Government%20Post%. Shalhoub, R., Quinones, I., Ames, C., Multaney, B., Curtis, S., Seeboruth, H., . . .Daniel, B. (2008). The recovery of latent fingermarks and DNA using a silicone-based casting material. Forensic Science International 178. p 190-203. Retrieved September 23, 2011 from http://www.forensic.sc.su.ac.th/seminar/seminari53/ref/52312342.pdf Tolba, A.S., El-Baz, A.H., & El-Harby, A.A. (2011). Face recognition: A literature review. International Journal of Signal Processing 2(2). Retrieved September 29, 2011 from www.scholar.google.co.uk/scholar?as_q=face+recognitionA%3A+A+literature+surv ey. United States Army. n.d. Chapter 15. Unexploded ordnance and improvised explosive devices. FM. 3-21.75 Chapter 15. Retrieved September 18, 2011 from https://rdl.train.army.mil/soldierPortal/atia/adlsc/view/public/24572-1/FM/321.75/chap15.htm Environ Earth Sci (2016) 75:12 DOI 10.1007/s12665-015-4830-8 ORIGINAL ARTICLE Flash flood susceptibility assessment in Jeddah city (Kingdom of Saudi Arabia) using bivariate and multivariate statistical models Ahmed M. Youssef1,2 • Biswajeet Pradhan3 • Saleh A. Sefry1 Received: 20 January 2015 / Accepted: 14 July 2015 / Published online: 18 December 2015  Springer-Verlag Berlin Heidelberg 2015 Abstract The city of Jeddah (Saudi Arabia) has experienced two catastrophic flash flood events in 2009 and 2011. These flood events had catastrophic effect on human lives and livelihoods around the wadi Muraikh, wadi Qus, wadi Methweb, and wadi Ghulail in which 113 people were dead and with 10,000 houses and 17,000 vehicles were damaged. Thus, a comprehensive flood management is required. The flood management requires information on different aspects such as the hydrological, geotechnical, environmental, social, and economic aspects of flooding. Flood susceptibility mapping for any area helps the decision makers to understand the flood trends and can aid in appropriate planning and flood prevention. In this study, two models were used for the generation of flood susceptibility maps for the Jeddah region. The first model includes bivariate probability analysis (frequency ratio), and the second model uses the multivariate analysis. For the multivariate model, the acquired weights of the FR model were entered into the logistic regression model to evaluate the correlation between flood occurrence and each related factor. This integration will overcome some of the weakness of the logistic regression, and the performance the & Biswajeet Pradhan biswajeet24@gmail.com; biswajeet@lycos.com Ahmed M. Youssef amyoussef70@yahoo.com 1 Geological Hazards Department, Applied Geology Sector, Saudi Geological Survey, P.O. Box 54141, Jeddah 21514, KSA 2 Geology Department, Faculty of Science, Sohag University, Sohag, Egypt 3 Department of Civil Engineering, Geospatial Information Science Research Center (GISRC), Faculty of Engineering, University Putra Malaysia (UPM), 43400 Serdang, Malaysia LR will be enhanced. A flood inventory map was prepared with a total of 127 flood locations. These flood locations were extracted from different sources including field investigation and high-resolution satellite image (IKONOS 1 m). These flood locations were randomly split into two groups, one dataset representing 70 % was used for training the models, and the remaining 30 % was used for models validation. Various independent flood-related factors such as slope, elevation, curvature, geology, landuse, soil drain, and distance from streams were included. The impact of each independent flood-related factors on flooding was evaluated by analyzing each independent factor with the historical flood inventory data. The training and validation datasets were used to evaluate the flood susceptibility maps using the success and the prediction rate methods. The results of the accuracy assessment showed a success rate of 90.4 and 91.6 % and a prediction rate of 89.6 and 91.3 % for FR and ensemble FR and LR models, respectively. In addition, a comparison has been made between real flood events in 2009 and the resultant susceptibility maps. Hence, it is concluded that the FR and ensemble Fr and LR models can provide an acceptable accuracy in the prediction of flood susceptibility in the Saudi Arabia. Our findings indicated that these flood susceptibility maps can assist planners, decision makers, and other agencies to deal with the flood management and planning in the area. Keywords Flash floods  Remote sensing  GIS  FR  Ensemble  Susceptibility  Jeddah  Saudi Arabia Introduction Flash floods are generated when precipitation saturates the drainage capacity of the basin slopes, causing impoundment of the drainage network, and resulting in 123 12 Page 2 of 16 exceptionally high discharge at the basin outlets (Youssef et al. 2009a). The rapid urban growth coupled with climate change in recent years has led to many environmental problems and increased risks of natural disasters (Kjeldsen 2010), including flooding and associated losses of human lives and property (Zwenzner and Voigt 2009). The occurrence of the flood disaster is expected to be raised due to unplanned expansion of urbanization (Tehrany et al. 2015). Flooding is one of the most costly disasters in terms of both property damage and human casualties (Alexander 1993). Most of the floods have an impact on people as the fear from the consequences exceeds the actual impacts (Green and Penning-Rowsell 1989). Hassan (2000) mentioned that frequent flash floods have seriously affected the highway and human activities along the coastal plains of the Red Sea. Numerous other studies on the flood hazards have been reported in different areas such as Youssef et al. (2005); Youssef and Hegab (2005). Flash floods are a major threat to human life and infrastructures such as urban areas, roads, and railways. The damage that can occur due to such disasters leads to huge economic cost, and consequently, floods can bring pathogens into urban environments and cause microbial development and diseases (Taylor et al. 2011; Dawod et al. 2012). Floods lead to human injury or death, and prevention of such events is preferable to compensation of damages. Regmi et al. (2013) indicated that in natural hazard-related research, huge databases are often needed. Youssef et al. (2009a) indicated that different natural factors such as hydrological and meteorological characteristics, soil types, geological structures, geomorphology, and vegetation are the most influential contributors to flooding. Human activities such as increasing the impervious materials (roads and buildings) as well cutting trees can accelerate flooding. Billa et al. (2006) indicated that flood control and prevention measures are urgently needed which will help in decreasing and minimize the tremendous and irreversible potential damages to agriculture, transportation, bridges, and urban infrastructure. Early warnings and emergency responses to floods are required (Feng and Wang 2011) so that governments and agencies can prevent as much damage as possible. However, measuring the benefits of flood reduction is difficult because they are not tangible and require a long time to be shown (Yi et al. 2010). In contrast, damage can be calculated both qualitatively and quantitatively (De Moel and Aerts 2011). Liu and De Smedt (2005) indicated that different new insights in the hydrological research can determine and mitigate flooding using geographic information system (GIS), digital soil-type maps, topography, and landuse/land cover data. Over the last two decades, remotely sensed data (active and passive sensors) have been used effectively for monitoring and analyzing different types of phenomena 123 Environ Earth Sci (2016) 75:12 and hazards (Mason et al. 2010; Elbialy et al. 2013; Pradhan et al. 2014; Youssef et al. 2009b, 2013, 2014a, b, c, 2015; Youssef 2015). Typically, studies of hazards require multitemporal datasets in order to identify spatial changes and the process of hazard occurrence (Martinez and Le Toan 2007). Bubeck et al. (2012) indicated that mapping of the flood-prone areas is an essential step in flood risk management. In addition to that, GIS represents a useful tool to investigate the flooding events. There are different types of approaches to study the flood hazard assessment. Many authors used GIS techniques in flood mapping (Chau et al. 2005; Mukerji et al. 2009). Other recently developed methods that were used to identify areas at risk of flooding (flood susceptibility) including qualitative and quantitative techniques such as multicriteria evaluation (Matori 2012), artificial neural networks (ANNs) (Campolo et al. 2003; Kia et al. 2012), frequency ratio (Lee et al. 2012), analytical hierarchy process (AHP) (Rozos et al. 2011), decision tree (DT) (Tehrany et al. 2013), logistic regression (Pradhan 2010a), adaptive neurofuzzy interface system (ANFIS) (Mukerji et al. 2009). Kourgialas and Karatzas (2011) indicated that the flood susceptibility map will manage any future flood problems. Tehrany et al. (2013, 2014a, b) reviewed the advantages and disadvantages of these statistical models. If large numbers of variables are used, the modeling process is time-consuming. Lee et al. (2012) applied individual bivariate probability models to map flood-susceptible areas in Busan, South Korea. A drawback of this method is that it considered the relationship between flood occurrence and each independent separately, while not considering the relationships among all the independent layers themselves. Pradhan (2010a) utilized multivariate logistic regression to examine the relations between a dependent variable and several independent variables to produce a susceptibility zonation map of Kelantan, Malaysia. He noted that logistic regression had several advantages: The variables can be either continuous or discrete, and they do not necessarily have to have normal distributions. Although the results of that study showed the efficiency of logistic regression, the impacts of classes of each variable were not considered. Accordingly, bivariate probability and logistic regression are both popular methods of statistical analysis for susceptibility mapping. Mostly, they are used individually, as either can produce a model of susceptibility. Tehrany et al. (2013) combined these methods for flood susceptibility analysis in tropical region of Kelantan, Malaysia. Prediction of flooding can be highly effective in preventing properties damage and life loss. Through scientific methods, flood-susceptible areas can be detected. The use of bivariate statistics (frequency ratio) and multivariate statistics (ensemble FR and LR) models in flood susceptibility mapping has not been explored in flood mapping in Environ Earth Sci (2016) 75:12 Saudi Arabia. The main objective of the current study is to apply flood susceptibility assessment using two statistical approaches of bivariate probability (FR) and logistic regression models (ensemble FR and LR) and to examine their relative efficiency and reliability for flood susceptibility analysis in an arid region (wadis Muraikh, Qus, Asheer, Methweb, and Ghulail in the Jeddah area, Saudi Arabia). This study also aims to determine optimized conditioning factors in flood susceptibility mapping through GIS analysis. Study area Location The study area includes five wadis, covering 219 km2 between latitudes 21240 0600 and 21330 4700 N and longitudes 39140 3500 and 39280 2400 E, named wadi Muraikh, wadi Qus, wadi Asheer, wadi Methweb, and wad Ghulail (Fig. 1). The study area receives flash flood water from the foothills through natural drainage channels of these wadis which are located in the eastern part. In the years of 2009 and 2011, these wadis were flooded causing the most damages. The morphometric characteristics of these five wadis in the study area are shown in Table 1. The altitudes of the study range between 30 m and 275 m above mean sea level. According to the meteorological data of the Jeddah area, the average annual precipitation is 52.5 mm/ year, the maximum rainfall is 284 mm in 1996, and the minimum rainfall is zero mm. Most of the rainfalls are infrequent and occur as intense thunderstorms from November to April. Geomorphology and geological setting Geomorphologically, the study area is situated toward the east of the Red Sea coastal plain which is characterized by the presence of hills and pediments. It includes different landforms such as small and mid-size low-laying rounded hills (elevation ranges from few tens of meters in the west to a few hundred meters in the east side) and flattened foothills (covered by alluvium deposits). The general sloping trend is toward the Jeddah metropolitan area which is located to the west of the study area (Fig. 2). The area is dissected by many streams (wadis) that transport the runoff water toward the Jeddah urban areas (coastal plain areas). Most of the flattened foothills include urban areas and roads, thus increasing the impervious cover and accordingly the runoff increased dramatically. Geologically, two units were detected, including the Neoproterozoic basement complex (oldest) and the Holocene sediments (youngest) (Qari 2009). Most of the hills Page 3 of 16 12 and pediments are characterized by the Neoproterozoic basement rocks which are consisted of volcanic rocks (andesite and dacite) and intrusive rocks (diorite and granite), whereas the Holocene sediments cover the flattened areas between the hills (wadis), including alluvium and aeolian deposits (Moore and Al-Rehaili 1989). Causes and consequences of flash flood events in the study area Flash floods can be caused according to various factors. Rainstorm events, geomorphological conditions of the area, and anthropogenic activities were considered to be the main causative factors causing these flood events along the study area. In the years of 2009 and 2011, two rainstorm events were happened in Jeddah area with a rainfall of 70 mm and 111 m in 3 h for each event, respectively. These events caused catastrophic floods. Most of the damages were occurred along the areas that located after the mouths of the narrow parts of the wadis. In addition to that, man-made earth dykes that were established by private peoples to collect waters for irrigation and agricultural activities were considered to have a big influence in increasing the damage force of the flash floods along the study area. These earth dykes collected a huge amount of waters at the rainstorm events of 2009 and 2011, accordingly the water pressure increased on these dykes, and finally, they broke down causing a big problem. Other anthropogenic factor includes uncontrolled urban distributions along the flat areas between the hilly part causing increment of runoff water and floods. According to the governmental records, many buildings, cars, lightweight trucks, highways, and roads were damaged and 113 persons were died due the flood events of 2009 and 2011. No warning system was established in the area before these flood events, and for that reason, large number of losses and damages were occurred. Urban areas impacted by flash floods Mapping the urban changes has fundamental impacts on the flash flood hazards in any area. In the current work, the diachronic changes of the urban areas were determined. The expansion of the Jeddah city toward the east was due to the increase in the population in the old Jeddah city. However, the new expansions lead to establishment of new residences and infrastructure works in the areas that are prone to flash floods. To create the urban changes in the study area, different steps were applied which includes: (1) interpretation of satellite images TM 1990, ETM? 2005, and IKONOS 2009. The urban areas were digitized from 123 12 Page 4 of 16 Environ Earth Sci (2016) 75:12 Fig. 1 a Location of the study area in relation to Saudi Arabia map, b location of the study area in relation to Jeddah city each map after stacking the bands for each image using ERDAS Imagine software. (2) All images were transferred to the ArcGIS environment [742 (RGB)] in order to map the boundaries of urban pattern of the study area (Fig. 3a– c). Overlaying techniques were used to identify the changes of the urban extent of the town for the years 1990, 2005, and 2009 (Fig. 3d). 123 Data, methodology, inventory map, and conditioning factors Data used In the current study, different datasets were used as shown in Table 2. The spatial database was constructed by Environ Earth Sci (2016) 75:12 Page 5 of 16 12 Table 1 Morphometric characteristics of the drainage basins of the study area Basin no. Basin area (km2) Basin length (m) Basin slope (m/m) Basin perimeter (m) 1 40.8 10,070.4 0.0501 40,183.7 2 68.8 20,846.2 0.0372 3 11.6 6510.6 0.0471 4 57 15,744.9 0.0463 5 40.6 12,268.1 0.0520 Mean elevation (m) Main stream distance (m) MSD slope (m/m) 88.5 10,838.5 0.0073 75,758.0 142.3 23,803.3 0.0069 22,132.3 90.4 6087.3 0.0084 54,687.5 123.8 17,159.5 0.0066 47,208.1 106.8 14,328.4 0.0059 Fig. 2 Geomorphological features of the study area incorporating relevant remote-sensing datasets. A geographic information system (GIS) was used to generate, compile, and host the datasets for data interpretation and analysis. First, datasets using different types of satellite imageries including thematic mapper (TM) for 1990 (30-m spatial resolution) enhanced thematic mapper plus (ETM?) for 2005 (30-m spatial resolution) and IKONOS images for 2008 and 2009 (1-m spatial resolution). These datasets were used for mapping of changes in urban areas and delineation of landuse and soil drain areas. Second, datasets were used in the current study that includes the geological map which was extracted from the Makkah quadrangle sheet GM-107c of 1985 (scale 1: 250,000). The geological map of the study area is used to determine rock units. Third, datasets include a digital elevation model (DEM) which was used for extracting slope, elevation, curvature, and streams of the study area. Fourth, datasets used in the current work are related to several field investigations. The multiple field investigation has been carried out during and after the flood events of 2009 and 2011 to help in collecting real data to test and validate the flood susceptibility models. A unified projection (UTMZone 37, WGS84 datum) was used for all the GIS and remote-sensing-derived datasets. Methodology Bivariate statistical analysis (frequency ratio) It is common to assume that flood occurrence is determined by flood-related factors, and that future flood areas will occur under the same conditions as past floods (Tehrany et al. 2013, 2014a, b). Using this assumption, the relationship between flood occurring in an area and the flood- 123 12 Page 6 of 16 Environ Earth Sci (2016) 75:12 Fig. 3 Urban areas expansion of the study area; a up to 1990, b up to 2005, c up to 2009, d the overall changes up to 2009 of the study area Table 2 Data sources and datasets used in the current study Dataset no. Classification Data type Scale and resolution Type of extracted data 1 Satellite imageries TM 1990 30 m Mapping of urban areas ETM 2005 30 m Geomorphological units IKONOS 2008 and 2009 1m Mapping flooding areas Makkah quadrangle GM-107C 1985 1:250,000 ? 2 Geological data Geomorphological units Geological units 3 Digital elevation model Grid 5m Elevation, slope, curvature, streams 4 Field investigation Information on the damaged areas by floods of 2009 and 2011 Field trips Flooding areas 2009, 2011 related factors can be distinguished from the relationship between flood not occurring in an area and the flood-related factors. Lee and Pradhan (2007), Yilmaz (2009), Pradhan 123 (2010b), and Pradhan and Lee (2010a) indicated that the ratio of the probability of the presence to the absence of any occurrence, such as a flood or landslide, can be used Environ Earth Sci (2016) 75:12 Page 7 of 16 for flood or landslide susceptibility assessment. To build a probability model, it is assumed that the hazard occurrence can be determined based on its related factors and that future problems will happen under the same conditions as past events (Akgun et al. 2012). The bivariate probability for each independent variable was measured by the variable’s relationship with flood occurrence. The greater the bivariate probability, the stronger the relationship between flood occurrence and the flood-related factors (Lee and Talib 2005; Lee and Pradhan 2007; Pradhan and Lee 2010b). Bivariate probability was performed (frequency ratio) which is a quantitative relationship between flood occurrences and different related parameters. The frequency ratio is defined in Eq. (1). By using the frequency ratio model, the spatial relationships between flood occurrence and related factors contributing to flood occurrence were derived. Wij ¼ FLij FNij ð1Þ where Wij is the frequency ratio of class i of parameter j; FLij is the frequency of observed flood in class i of parameter j; and FNij is the frequency of non-observed flood in class i of parameter j. Therefore, the greater the ratio above unity, the stronger the relationship between flood occurrence and the given factor’s class attribute, and the lower the ratio below unity, the lesser the relationship between flood occurrence and the given factor’s class attribute (Tehrany et al. 2014a, b). The calculated frequency ratios are given in Table 3. To calculate the flood susceptibility index (FSI), each factor’s frequency ratio values were summed (Lee and Min 2001; Lee and Pradhan 2006; Youssef et al. 2014a). In the current study, the FSI was determined by sum of each factor’s frequency ratio as expressed in Eq. (2). FSI ¼ n X Wij ð2Þ j¼1 where FSI, flood susceptibility index; Wij , weight of class i in parameter j; and n, number of parameters. Multivariate statistical analysis (logistic regression model) Logistic regression (LR) is a popular multivariate statistical analysis method that allows for examination of the multivariate regression relationship between a dependent variable and several independent variables (Pradhan and Lee 2010b). Logistic regression was created to measure the probability of a disaster in an area using a specific formula generated using the independent variables (Eq. 3). One of the requirements of this method is dependent data consisting of values of 0 and 1, which indicates the absence 12 and existence of disasters, respectively. Here, the dependent and reclassified independent variables derived by bivariate probability were converted from raster to ASCII format, which is required in SPSS. In the current study, Dai and Lee (2002) approach in which there is equal proportion of floods and non-flood areas was applied. For the ensemble of FR and LR, the weight values of the FR were normalized in the range of 0 and 1. Flood-related factors were classified based on these normalized values. Finally, they entered into the LR model to get the ensemble FR and LR model. The coefficients are measured and listed in Table 3. The higher the logistic coefficient, the greater the expected impact on flooding occurrence. Using the derived logistic coefficients, the probability (p) of flood occurrence was calculated as in Eq. (3). p¼ 1 1 þ ez ð3Þ where p is the probability of an event occurring. In this situation, the value p is the estimated probability of flood occurrence. The probability varies from 0 to 1 on an Sshaped curve, and z is the linear combination. It follows that logistic regression involves fitting Eq. (4). z ¼ b0 þ b1 x 1 þ b 2 x 2 þ    þ b n x n ð4Þ where b0 is the intercept of the model, bi (i = 0, 1, 2, …, n) represents the coefficients of the logistic regression model, and xi (i = 0, 1, 2, …, n) denotes the independent variables (Lee and Sambath 2006). Flood inventory map Flooded areas were identified, and an inventory map containing 127 flooded areas was generated over the whole study area through field investigation and the examination of high-resolution satellite image with one meter pixel size before and after floods. During the change analysis, IKONOS image of 2008 was used as reference when no floods had happened, and IKONOS image of 2009 was taken after the flood scenario of November 2009. Based on the literature review, 70 % of the flooded locations were used randomly to prepare the model which called training dataset in order to assess the spatial distribution of flooding (Pradhan 2010a). The rest of the locations 30 % (validation data) were used for validating the model. Figure 4 shows an inventory map of flooded locations of the study areas in the Jeddah region. Flood conditioning parameters Conditioning parameters are needed as independent variables to produce of flood susceptibility mapping (Liu and De Smedt 2005). These parameters can contribute to the 123 12 Page 8 of 16 Environ Earth Sci (2016) 75:12 Table 3 Results of the frequency ratio analysis (FR) for all parameter classes and factors weight for ensemble FR and LR model Factor Slope () Elevation (m) Curvature Geology Landuse Soil drain Distance from streams (m) Class 0 FR model Pixels number in domain Pixels number in domain %a Flood area number Flood area number %b 3,257,400 37.49 75 84.27 2.25 0–4.46 535,673 6.17 4 4.49 0.73 4.46–6.05 750,463 8.64 2 2.25 0.26 6.05–7.88 697,476 8.03 0 0.00 0.00 7.88–11.30 580,625 6.68 1 1.12 0.17 11.30–15.77 610,209 7.02 2 2.25 0.32 0.00 15.77–20.77 653,287 7.52 0 0.00 20.77o–26.02 543,975 6.26 3 3.37 0.54 26.02–33.91 536,243 6.17 0 0.00 0.00 33.91–67.03 522,277 6.01 2 2.25 0.37 36–63 m 886,973 10.21 18 20.22 1.98 63–78 m 936,627 10.78 14 15.73 1.46 78–92 m 865,299 9.96 10 11.24 1.13 92–106 m 106–117 m 881,886 933,583 10.15 10.75 9 9 10.11 10.11 1.00 0.94 117–126 m 857,983 9.88 9 10.11 1.02 126–137 m 831,802 9.57 11 12.36 1.29 137–152 m 843,050 9.70 8 8.99 0.93 152–176 m 842,626 9.70 1 1.12 0.12 176–275 m 807,799 9.30 0 0.00 0.00 Concave 2,219,662 25.55 3 3.37 0.13 Flat 4,480,185 51.57 80 89.89 1.74 Convex 1,987,781 22.88 6 6.74 0.29 Volcanic 1,269,601 14.61 6 6.74 0.46 Intrusive 3,480,890 40.07 10 11.24 0.28 Sedimentary 4,011,747 46.18 73 82.02 1.78 Barren areas (rock) 3,021,839 34.78 1 1.12 0.03 Urban in valleys 2,145,564 24.70 32 35.96 1.46 Barren areas (soil) 1,901,885 21.89 50 56.18 2.57 Urban on hilly areas 1,617,571 18.62 6 6.74 0.36 Poorly drain 5,132,633 59.08 16 17.98 0.30 Well drain 3,630,193 41.79 73 82.02 1.96 98,960 1.14 15 16.85 14.80 10 m 10–50 m 330,251 3.80 32 35.96 9.46 50–100 m 395,012 4.55 16 17.98 3.95 7,863,405 90.51 26 29.21 0.32 [100 m occurrence of flooding in a specific area (Tehrany et al. 2014a, b). There are many relevant independent parameters that can be used and analyzed in modeling. However, the model will be comprehensive and will require lots of information, which can be challenging to acquire. Sanyal and Lu (2004) mentioned that it is important to determine 123 Frequency ratio(b/a) LR model Factor weight 0.06 0.025 0.0001 0.019 0.03 0.021 0.01 which variables can be eliminated without harming the model’s precision. Recently, Campolo et al. (2003) aimed to produce models that use the minimum number of independent factors, which give high accurate model results. The independent data should be measurable, and there should be a conjunction between the independent and Environ Earth Sci (2016) 75:12 Page 9 of 16 12 Fig. 4 Flood inventory data used to test and validate the flood susceptibility models dependent data. Ayalew and Yamagishi (2005) indicated that the independent data should be collected from the whole study area, while it should not represent uniform spatial information. Furthermore, its effect should not lead to two types of outcome at the end of the process. These conditioning parameters can be in nominal, ordinal, interval, or ratio scale format (Park et al. 2013). Many factors may be influential in flood occurrence for a specific area, while the same factors may not be effective for other environments. No exact agreement exists on which parameters should be used in flood susceptibility assessments (Tehrany et al. 2014a, b). Numerous researchers used some of these factors according to their importance and vital role in flood studies. Skilodimou et al. (2003) mentioned that the anthropogenic factors including urban areas, road network, and landuses should be taken into consideration in flood susceptibility assessment which are related to flood events. Different factors were used in the current work, which are selected according to different literature (Pradhan 2010a; Kia et al. 2012; Lee et al. 2012; Tehrany et al. 2014a, b). The data were collected and compiled into spatial databases using ArcGIS 10.2 software. The independent parameters consisted of slope, elevation, curvature, geology, landuse, soil drain, and distance from streams. Each variable was resized to be presented by a 5 9 5 m grid, and the grid of the study area covers about 219 km2. The quantile method was used to classify each independent variable. In the quantile classification method, each class contains the same number of features. This method is used by several authors due to its efficiency in classification (Papadopoulou-Vrynioti et al. 2013; Tehrany et al. 2014a, b; Youssef et al. 2014a, b, c; Umar et al. 2014). In the current study, different layers/factors were classified into various classes based on the above-mentioned literature review. 1. 2. 3. 4. 5. 6. In the slope (0–67.03) and elevation (63–275 m) maps, ten categories were constructed for each analysis (Fig. 5a, b). In the curvature map, three classes were considered: convex, flat, and concave (Fig. 5c). In the geology map, three geological classes were used: intrusive, volcanic, and sedimentary (Fig. 5d). The map of the four landuse groups [Barren areas (Rock), urban in valley, urban on hill areas, and Barren areas (soil)] as shown in Fig. 5e. Figure 5f presents the soil drain map (poorly drain soils and well and very well-drained soils). In the case of the distance from streams map, four buffer categories (10, 50, 100, and 100 m) were compiled using the buffer tool in ArcGIS 10.2 software. The approach of drainage system delineation for the catchments was obtained from the digital elevation model (DEM) (5-m resolution) derived from the 2-m DEM developed by King Abdulaziz City for Science and Technology and data created from the 123 12 Page 10 of 16 Fig. 5 List of all the conditioning data layers. a Slope, b elevation, c curvature, d geological units, e landuse, f soil drain, and g distance from streams 123 Environ Earth Sci (2016) 75:12 Environ Earth Sci (2016) 75:12 light detection and ranging (LiDAR). All the network drainages of wadi Muraikh, wadi Qus, wadi Asheer, wadi Methweb, and wadi Ghulail were extracted using the watershed modeling systems (WMS 8.1) (Fig. 5g). Results and discussion Application of frequency ratio (FR) The spatial relationship and the frequency ratio values between flood locations and flood conditioning factors are shown in Table 3. By applying the frequency ratio model, it was found that the lower slopes (close to zero degree) have higher FR value of 2.25, followed by class 0–4.46 (0.73), whereas other slope classes have a FR lower than 0.6. Lower slopes that close to zero are expected to have higher weight values of FR. Elevation class of 36–63 m has a highest frequency ratio value (1.98), whereas class of 176–275 m has lowest frequency ratio (0.00). Results showed that the FR values for elevation classes have a general trend that increased for the lower elevation. For the curvature factor, flat areas have higher FR with values of 1.74, whereas the convex and concave classes have the lowest FR value of 0.29 and 0.13, respectively. As for the geological factors, the FR value was 1.78 in sedimentary rocks and 0.46 in the volcanic rocks, followed by a value of 0.28 for intrusive rocks. The reason for increasing the frequency ratio in sedimentary rocks is related to the presence of these sediments in the wadis (alluvial deposits). For the landuse factor, Barren areas (soil) have higher FR with values of 2.57, followed by urban in valleys with a FR ratio of 1.46, then urban on hilly areas with a FR value of 0.36, followed by Barren areas (rocks) with a FR value of 0.03. For the soil drain factor, the well-drained soil class has higher FR with value of 1.96 and the poor-drained soil class has a FR value of 0.32. The higher the FR for the well drain soil indicated that these well drain soil along the wadis. In terms of distance from streams, distances between 0 and 10 m have highest FR with values of 14.8, whereas distance from stream class greater than 100 m has a FR of 0.32. Rating layers for the different flood-related factors are constructed based on the frequency ratio values. The resulting FSI map is depicted in Fig. 6a. The FSI values vary from 1.06 to 27.08. The flood susceptibility value represents the relative susceptibility to flood occurrence. So the greater the frequency ratio value, the higher the susceptibility to flood occurrence and the lower the value, the lower the susceptibility to flood occurrence. The FSI map was classified according to quantile classifier method into five different flood susceptibility zones, as shown in Fig. 6a. Page 11 of 16 12 Application of logistic regression (LR) model In this study, the input data for the logistic regression model for predicting flood susceptibility are prepared. Logistic regression was performed using independent factors that were reclassified and weighted based on the bivariate probability (frequency ratio). Using the logistic regression model, the spatial relationship between flood occurrence and flood-related factor floods was assessed. The correlations between flood and each factor were calculated. In this analysis, the ‘‘continuous data’’ such as slope, elevation, curvature, geology, landuse, soil drain, and distance from streams were treated as ‘‘ordinal’’ in SPSS. The logistic regression coefficient for each thematic layer was computed and is shown in Table 3 and Eq. (5), where each of the continuous thematic layers has only single coefficient values. Zn ¼ ð0:51 þ 0:06  ‘‘Slope’’ þ 0:025  ‘‘Elevation’’ þ 0:0001  ‘‘Curvature’’ þ 0:019  ‘‘Geology’’ þ 0:03  ‘‘Landuse’’ þ 0:012  ‘‘Soil drain’’ þ 0:01  ‘‘Distance from streams’’ ð5Þ Hosmer–Lemeshow test showed that the goodness of fit of the equation can be accepted because the significance of Chi-square is larger than 0.05. The value of Cox and Snell (R2) and Nagelkerke (R2) showed that the independent variables can explain the dependent variables. In the current study, the FSI was determined according to Eq. (5). The resulting FSI map is depicted in Fig. 6b. The FSI values range from 0.5257 to 0.995. The flood susceptibility value represents the relative susceptibility to flood occurrence. So the greater the FSI, the higher the susceptibility to flooding occurrence and the lower the value, the lower the susceptibility to flooding occurrence. Model validation The prediction accuracy and quality of the developed model are examined here. The model was validated by comparing the acquired flood probability map with existing flood data (Lee et al. 2007; Tien Bui et al. 2012; Pourghasemi et al. 2012). Specifically, the results were quantitatively examined using the receiver operating characteristic (ROC) curve, in which the basis of the assessment is the true- and false-positive rates (Chauhan et al. 2010). To examine the reliability and efficiency of the flood probability map, both the success rate and prediction rate curve were calculated. The training flooded datasets were used to generate the flood model, but could not be used to assess the prediction capability of the model. In 123 12 Page 12 of 16 Environ Earth Sci (2016) 75:12 Fig. 6 Flood susceptibility maps produced from a FR and b ensemble FR and LR models general, the prediction rate shows how well the model could predict flooding in a given area (Tien Bui et al. 2012). This can show the predictive capability of a model (Maier and Dandy 2000; Pourghasemi et al. 2012), in that 123 the area under the prediction rate curves (AUC) can measure the prediction accuracy qualitatively (Pradhan and Lee 2010c). The success rate result was obtained using the training dataset, which used 70 % of the inventory flood Environ Earth Sci (2016) 75:12 locations (89 flood locations). The success rate curves are presented in Fig. 7a. The model produced a value of 0.904 which represents 90.4 % and 0.916 which represents 91.6 % for the area under the curve (AUC), success accuracy for the LR and FR. The prediction accuracy was calculated using the validating datasets for the 30 % of the flooded areas (38 flood locations) that were not used in the training process. The AUC values varied from 0.5 to 1.0. The value of 1.0 represented the highest accuracy, showing that the model was completely capable of predicting disaster occurrence without any bias (Pradhan et al. 2010). Thus, AUC values close to 1.0 are considered to show that a model is precise and reliable (Tien Bui et al. 2012). The prediction curve is shown in Fig. 7b. The model produced the AUC value of 0.896 which represents 89.6 % and 0.913 which represents 91.3 % for the area under the curve (AUC), prediction accuracy for the LR and FR. Moreover, another way was provided in order to test the performance of the used methods according to the field data that represent the flood events happened in 2009. The comparison shows good matches between the areas were flooded in the 2009, along wadi Muraikh, wadi Qus, wadi Asheer, wadi Methweb, and wadi Ghulail of the study area, and the products of the model. Finally, it can be seen that the high susceptible zones are mostly located along the wadis and the western parts of the catchments. A large part of the Fig. 7 Success rate (a) and prediction rate (b) curves for models derived from the FR and ensemble FR and LR Page 13 of 16 12 catchments were classified as very low, low, and moderate susceptible zones using the applied models. Conclusion Flood susceptibility mapping is one of the first steps in overall flood management program and has been adopted by many countries. Many authors used various methods for this purpose. Prediction of flooding can be highly useful in preventing the overall damage and loss due to a flood. Through scientific analysis of floods, flood-susceptible areas can be detected and thus flood damages can be decreased. The aim of this work was to investigate the application of FR and ensemble FR and LR models (bivariate and multivariate statistical analysis) in detection of flood-prone regions in wadis Muraikh, Qus, Methweb, and Ghulail in the Jeddah area, Saudi Arabia. Consequently, two flood susceptibility maps were constructed. The frequency ratio and ensemble FR and LR models were applied using independent factors that were reclassified and weighted based on the bivariate probability, in order to produce a more precise flood susceptibility map of the study area. The flood inventory map, which represented 127 locations that were flooded in 2009 and 2011, was used for the statistical analysis. Among all the flood events, 89 cases were used to build the model. The remaining 38 flooded locations were used to validate the developed model. Seven independent variables were used as independent data. These consisted of slope, elevation, curvature, geology, landuse, soil drain, and distance from streams. Initially, all the independent variables were resized to a 5 9 5 m grid. The grid of the study area covers about 219 km2. As a next step, each variable was classified using the quantile method, and the bivariate probability was used to evaluate the relationship of each class with flood occurrence. To assign the weights that were derived using bivariate probability, each independent variable was reclassified and normalized to be used in logistic regression multivariate statistical analysis. The logistic regression coefficients were calculated in SPSS V.22 software, and the probability index was then calculated. Finally, a flood susceptibility map of the study area was constructed by classifying the probability index into five classes: very low, low, medium, high, and very high susceptibility. AUC method was used to examine the efficiency of the proposed method. In order to evaluate and compare these models, the receiver operating characteristics (ROC) were used. The success rate and the prediction rate were measured with values of 91.6 and 91.3 %, and 90.4 and 89.3 % for both bivariate probability (FR) and multivariate probability (ensemble FR and LR), respectively. Results showed that the FR model produced better fits to the training and 123 12 Page 14 of 16 validating data compared to the ensemble FR and LR model. However, the ensemble FR and LR model was enhanced the using of LR model. Result revealed that the slope is the most significant factor in flood generation followed by landuse, elevation, soil drain, geology, distance from streams, whereas the curvature has the lowest significance. The current study indicated that FR and ensemble FR and LR models are more applicable in flood susceptibility mapping in Saudi Arabia. Their results are acceptable, and the process of the analysis is understandable. Having precise and reliable flood susceptibility maps can decrease the costs and damages if uncontrolled urban areas extend to cover the flood-prone areas. They can be used as an efficient method to produce flood susceptibility map in GIS platform. The current generated maps can be used as the basis of further research in hazard mapping, hydrological studies, and disaster management. Moreover, it can aid the planners, decision makers, and governments engaged in flood management and planning in the study area and to take proper actions in order to control and mitigate this flood phenomenon in the study area. Furthermore, the produced maps can be used to prevent excessive urbanization in susceptible flood-prone areas to reduce the potential damage caused by floods. Acknowledgments Authors would like to thank two anonymous reviewers for their valuable comments and editorial comments of Prof. James W. LaMoreaux. References Akgun A, Sezer EA, Nefeslioglu HA, Gokceoglu C, Pradhan B (2012) An easy-to-use MATLAB program (MamLand) for the assessment of landslide susceptibility using a Mamdani fuzzy algorithm. Comput Geosci 38(1):23–34. doi:10.1016/j.cageo.2011.04.012 Alexander D (1993) Natural disasters. UCL Press, London, p 631 Ayalew L, Yamagishi H (2005) The application of GIS-based logistic regression for landslide susceptibility mapping in the KakudaYahiko Mountains, Central Japan. Geomorphology 65:15–31. doi:10.1016/j.geomorph.2004.06.010 Billa L, Shattri M, Mahmud AR, Ghazali AH (2006) Comprehensive planning and the role of SDSS in flood disaster management in Malaysia. Disaster Prev Manage 15:233–240. doi:10.1108/ 09653560610659775 Bubeck P, Botzen W, Aerts J (2012) A review of risk perceptions and other factors that influence flood mitigation behavior. Risk Anal 32:1481–1495. doi:10.1111/j.1539-6924.2011.01783.x Campolo M, Soldati A, Andreussi P (2003) Artificial neural network approach to flood forecasting in the River Arno. Hydrol Sci J 48:381–398 Chau K, Wu C, Li Y (2005) Comparison of several flood forecasting models in Yangtze River. J Hydrol Eng 10:485–491. doi:10. 1061/ASCE1084-0699(2005)10:6(485) Chauhan S, Sharma M, Arora MK (2010) Landslide susceptibility zonation of the Chamoli region, Garhwal Himalayas, using logistic regression model. Landslides 7:411–423. doi:10.1007/ s10346-010-0202-3 123 Environ Earth Sci (2016) 75:12 Dai FC, Lee CF (2002) Landslide characteristics and slope instability modeling using GIS, Lantau Island, Hong Kong. Geomorphology 42:213–228 Dawod GM, Mirza MN, Al-Ghamdi KA (2012) GIS-based estimation of flood hazard impacts on road network in Makkah city, Saudi Arabia. Environ Earth Sci 67:2205–2215. doi:10.1007/s12665012-1660-9 De Moel H, Aerts J (2011) Effect of uncertainty in landuse, damage models and inundation depth on flood damage estimates. Nat Hazards 58:407–425. doi:10.1007/s11069-010-9675-6 Elbialy S, Mahmoud A, Pradhan B, Buchroithner M (2013) Application of spaceborne synthetic aperture radar data for extraction of soil moisture and its use in hydrological modelling at Gottleuba Catchment, Saxony, Germany. J Flood Risk Manag. doi:10.1111/jfr3.12037 Feng CC, Wang YC (2011) GIScience research challenges for emergency management in Southeast Asia. Nat Hazards 59:597–616. doi:10.1007/s11069-011-9778-8 Green CH, Penning-Rowsell EC (1989) Flooding and the quantification of intangibles. J Inst Water Env Man 3(1):27–30 Hassan OA (2000) Salient geoenvironmental parameters of Ras Malaab—Abu Zenima Area, Gulf of Suez, Egypt, with an emphasis on flash flood potential and mitigative measures; Egypt. J Remote Sensing Space Sci 3:37–58 Kia MB, Pirasteh S, Pradhan B, Mahmud AR, Sulaiman WNA, Moradi A (2012) An artificial neural network model for flood simulation using GIS: Johor River Basin, Malaysia. Environ Earth Sci 67:251–264. doi:10.1007/s12665-011-1504-z Kjeldsen TR (2010) Modelling the impact of urbanization on flood frequency relationships in the UK. Hydrol Res 41:391–405. doi:10.2166/nh.2010.056 Kourgialas NN, Karatzas GP (2011) Flood management and a GIS modelling method to assess flood-hazard areas—a case study. Hydrol Sci J 56:212–225. doi:10.1080/02626667.2011.555836 Lee S, Min K (2001) Statistical analysis of landslide susceptibility at Yongin, Korea. Environ Geol 40:1095–1113. doi:10.1007/ s002540100310 Lee S, Pradhan B (2006) Probabilistic landslide hazards and risk mapping on Penang Island, Malaysia. J Earth Syst Sci 115(6):661–672 Lee S, Pradhan B (2007) Landslide hazard mapping at Selangor, Malaysia using frequency ratio and logistic regression models. Landslides 4:33–41. doi:10.1007/s10346-006-0047-y Lee S, Sambath T (2006) Landslide susceptibility mapping in the Damrei Romel area, Cambodia using frequency ratio and logistic regression models. Environ Geol 50:847–855. doi:10.1007/ s00254-006-0256-7 Lee S, Talib JA (2005) Probabilistic landslide susceptibility and factor effect analysis. Environ Geol 47:982–990. doi:10.1007/ s00254-005-1228-z Lee S, Ryu JH, Kim IS (2007) Landslide susceptibility analysis and its verification using likelihood ratio, logistic regression, and artificial neural network models: case study of Youngin, Korea. Landslides 4:327–338. doi:10.1007/s10346-007-0088-x Lee MJ, Kang JE, Jeon S (2012) Application of frequency ratio model and validation for predictive flooded area susceptibility mapping using GIS. In: Geoscience and remote sensing symposium (IGARSS), 2012 IEEE International, Munich, pp 895–898 Liu Y, De Smedt F (2005) Flood modeling for complex terrain using GIS and remote sensed information. Water Resour Manag 19:605–624. doi:10.1007/s11269-005-6808-x Maier HR, Dandy GC (2000) Neural networks for the prediction and forecasting of water resources variables: a review of modelling issues and applications. Environ Model Softw 15:101–124. doi:10.1016/S1364-8152(99)00007-9 Environ Earth Sci (2016) 75:12 Martinez JM, Le Toan T (2007) Mapping of flood dynamics and spatial distribution of vegetation in the Amazon floodplain using multitemporal SAR data. Remote Sens Environ 108:209–223. doi:10.1016/j.rse.2006.11.012 Mason DC, Speck R, Devereux B, Schumann GP, Neal JC, Bates PD (2010) Flood detection in urban areas using TerraSAR-X. IEEE Trans Geosci Remote 48:882–894. doi:10.1109/TGRS.2009. 2029236 Matori A (2012) Detecting flood susceptible areas using GIS-based analytic hierarchy process. In: International conference on future environment and energy, Singapore Moore TA, Al-Rehaili MH (1989) Geologic Map of the Makkah Quadrangle, sheet 21D, Kingdom of Saudi Arabia, Saudi Arabian Directorate General of Mineral Resources Geoscience Map GM-107C, scale 1:250,000 Mukerji A, Chatterjee C, Raghuwanshi NS (2009) Flood forecasting using ANN, neuro-fuzzy, and neuro-GA models. J Hydrol Eng 14:647–652. doi:10.1061/(ASCE)HE.1943-5584.0000040 Papadopoulou-Vrynioti K, Bathrellos GD, Skilodimou HD, Kaviris G, Makropoulos K (2013) Karst collapse susceptibility mapping considering peak ground acceleration in a rapidly growing urban area. Eng Geol 158:77–88. doi:10.1016/j.enggeo.2013.02.009 Park S, Choi C, Kim B, Kim J (2013) Landslide susceptibility mapping using frequency ratio, analytic hierarchy process, logistic regression, and artificial neural network methods at the Inje area, Korea. Environ Earth Sci 68:1443–1464. doi:10.1007/ s12665-012-1842-5 Pourghasemi HR, Pradhan B, Gokceoglu C (2012) Application of fuzzy logic and analytical hierarchy process (AHP) to landslide susceptibility mapping at Haraz watershed, Iran. Nat Hazards 63:965–996. doi:10.1007/s11069-012-0217-2 Pradhan B (2010a) Flood susceptible mapping and risk area delineation using logistic regression, GIS and remote sensing. J Spat Hydrol 9(2):1–18 Pradhan B (2010b) Landslide susceptibility mapping of a catchment area using frequency ratio, fuzzy logic and multivariate logistic regression approaches. J Indian Soc Remote 38:301–320. doi:10. 1007/s12524-010-0020-z Pradhan B, Lee S (2010a) Delineation of landslide hazard areas on Penang Island, Malaysia, by using frequency ratio, logistic regression, and artificial neural network models. Environ Earth Sci 60:1037–1054. doi:10.1007/s12665-009-0245-8 Pradhan B, Lee S (2010b) Landslide susceptibility assessment and factor effect analysis: backpropagation artificial neural networks and their comparison with frequency ratio and bivariate logistic regression modelling. Environ Model Softw 25:747–759. doi:10. 1016/j.envsoft.2009.10.016 Pradhan B, Lee S (2010c) Regional landslide susceptibility analysis using back-propagation neural networks model at Cameron Highland, Malaysia. Landslides 7(1):13–30. doi:10.1007/ s10346-009-0183-2 Pradhan B, Youssef AM, Varathrajoo R (2010) Approaches for delineating landslide hazard areas using different training sites in an advanced neural network model. Geo-spat Inf Sci 13:93–102. doi:10.1007/s11806-010-0236-2 Pradhan B, Hagemann U, Tehrany MS, Prechtel N (2014) An easy to use ArcMap based texture analysis program for extraction of flooded areas from TerraSAR-X satellite image. Comput Geosci 63:34–43. doi:10.1016/j.cageo.2013.10.011 Qari MHT (2009) Geomorphology of Jeddah governorate, with emphasis on drainage systems. JKAU Earth Sci 20(1):93–116 Regmi AD, Yoshida K, Dhital MR, Pradhan B (2013) Weathering and mineralogical variation in gneissic rocks and their effect in Sangrumba Landslide, East Nepal. Environ Earth Sci 71(6):2711–2727. doi:10.1007/s12665-013-2649-8 Page 15 of 16 12 Rozos D, Bathrellos GD, Skillodimou HD (2011) Comparison of the implementation of rock engineering system and analytic hierarchy process methods, upon landslide susceptibility mapping, using GIS: a case study from the Eastern Achaia County of Peloponnesus, Greece. Environ Earth Sci 63:49–63. doi:10.1007/ s12665-010-0687-z Sanyal J, Lu X (2004) Application of remote sensing in flood management with special reference to monsoon Asia: a review. Nat Hazards 33:283–301. doi:10.1023/B:NHAZ.0000037035. 65105.95 Skilodimou H, Livaditis G, Bathrellos G, Verikiou-Papaspiridakou E (2003) Investigating the flooding events of the urban regions of Glyfada and Voula, Attica, Greece: a contribution to Urban Geomorphology. Geogr Ann A 85:197–204 Taylor J, Davies M, Clifton D, Ridley I, Biddulph P (2011) Flood management: prediction of microbial contamination in largescale floods in urban environments. Environ Int 37:1019–1029. doi:10.1016/j.envint.2011.03.015 Tehrany MS, Pradhan B, Jebur MN (2013) Spatial prediction of flood susceptible areas using rule based decision tree (DT) and a novel ensemble bivariate and multivariate statistical models in GIS. J Hydrol 504:69–79. doi:10.1016/j.jhydrol.2014.03.008 Tehrany MS, Pradhan B, Jebur MN (2014a) Flood susceptibility mapping using a novel ensemble weights-of-evidence and support vector machine models in GIS. J Hydrol 512:332–343. doi:10.1016/j.jhydrol.2014.03.008 Tehrany MS, Lee MJ, Pradhan B, Jebur MN, Lee S (2014b) Flood susceptibility mapping using integrated bivariate and multivariate statistical models. Environ Earth Sci. doi:10.1007/s12665014-3289-3 Tehrany MS, Pradhan B, Mansor S, Ahmad N (2015) Flood susceptibility assessment using GIS-based support vector machine model with different kernel types. Catena 125:91–101. doi:10.1016/j.catena.2014.10.017 Tien Bui D, Pradhan B, Lofman O, Revhaug I, Dick OB (2012) Landslide susceptibility mapping at Hoa Binh province (Vietnam) using an adaptive neuro-fuzzy inference system and GIS. Comput Geosci 45:199–211. doi:10.1016/j.cageo.2011.10.031 Umar Z, Pradhan B, Ahmad A, Jebur MN, Tehrany MS (2014) Earthquake induced landslide susceptibility mapping using an integrated ensemble frequency ratio and logistic regression models in West Sumatera Province, Indonesia. Catena 118:124–135. doi:10.1016/j.catena.2014.02.005 Yi CS, Lee JH, Shim MP (2010) GIS-based distributed technique for assessing economic loss from flood damage: pre-feasibility study for the Anyang Stream Basin in Korea. Nat Hazards 55:251–272. doi:10.1007/s11069-010-9524-7 Yilmaz I (2009) Landslide susceptibility mapping using frequency ratio, logistic regression, artificial neural networks and their comparison: a case study from Kat landslides (Tokat—Turkey). Comput Geosci 35:1125–1138. doi:10.1016/j.cageo.2008.08.007 Youssef AM (2015) Landslide susceptibility delineation in the ArRayth Area, Jizan, Kingdom of Saudi Arabia, by using analytical hierarchy process, frequency ratio, and logistic regression models. Environ Earth Sci. doi:10.1007/s12665-014-4008-9 Youssef AM, Hegab MA (2005) Using geographic information systems and statistics for developing a database management system of the flood hazard for Ras Gharib area, Eastern Desert, Egypt. In: The fourth international conference on the geology of Africa, Assiut, Egypt, November, vol 2, pp 1–15 Youssef AM, Abdel Moneim AA, Abu El-Maged SA (2005) Flood hazard assessment and its associated problems using geographic information systems, Sohag Governorate, Egypt. In: The fourth international conference on the geology of Africa, Assiut, Egypt, November, vol 1, pp 1–17 123 12 Page 16 of 16 Youssef AM, Pradhan B, Gaber AFD, Buchroithner MF (2009a) Geomorphological hazards analysis along the Egyptian Red Sea Coast between Safaga and Quseir. Nat Hazards Earth Syst Sci 9:751–766. doi:10.1007/s12665-014-3661-3 Youssef AM, Maerz NH, Hassan AM (2009b) Remote sensing applications to geological problems in Egypt: case study, slope instability investigation, Sharm El-Sheikh/Ras-Nasrani Area, Southern Sinai. Landslides 6(4):353–360. doi:10.1007/s10346009-0158-3 Youssef AM, Pradhan B, Maerz NH (2013) Debris flow impact assessment caused by 14 April 2012 rainfall along the Al-Hada Highway, Kingdom of Saudi Arabia using high-resolution satellite imagery. Arab J Geosci. doi:10.1007/s12517-0130935-0 Youssef AM, Al-kathery M, Pradhan B (2014a) Landslide susceptibility mapping at Al-Hasher Area Jizan (Saudi Arabia) using GIS-based frequency ratio and index of entropy models. Geosci J. doi:10.1007/s12303-014-0032-8 123 Environ Earth Sci (2016) 75:12 Youssef AM, Pradhan B, Jebur MN, El-Harbi HM (2014b) Landslide susceptibility mapping using ensemble bivariate and multivariate statistical models in Fayfa area Saudi Arabia. Environ Earth Sci. doi:10.1007/s12665-014-3661-3 Youssef AM, Al-kathery M, Pradhan B, Elsahly T (2014c) Debris flow impact assessment along the Al-Raith Road, Kingdom of Saudi Arabia, using remote sensing data and field investigations. Geomat Nat Hazards Risk. doi:10.1080/19475705.2014.933130 Youssef AM, Sefry SA, Pradhan B, Abu Alfadail E (2015) Analysis on causes of flash flood in Jeddah city (Kingdom of Saudi Arabia) of 2009 and 2011 using multi-sensor remote sensing data and GIS. Geomat Nat Hazards Risk. doi:10.1080/19475705. 2015.1012750 Zwenzner H, Voigt S (2009) Improved estimation of flood parameters by combining space based SAR data with very high resolution digital elevation data. Hydrol Earth Syst Sci 13:567–576. doi:10. 5194/hess-13-567-2009 Emergency Management in Saudi Arabia: Past, Present and Future Yassar A. Alamri1 “He who is secure in his house, healthy in his body and has his food for the day, has owned the world” - Prophet Mohammed Introduction The management of potentially hazardous situations such as religious mass gatherings has been the duty of the people of Makkah (now part of Saudi Arabia) for many centuries. Inhabitants of Makkah used to evacuate their houses to accommodate the incoming pilgrims, and servants of the Holy Mosque used to distribute cold water to quench pilgrims’ thirst. This concept of serving mass gatherings formed the nucleus of the first emergency management plans in the Kingdom of Saudi Arabia. Today, Saudi Arabia covers most of the Arabian Peninsula and has faced many other risks in addition to those arising from religious mass gatherings. In order to improve on the existing emergency management policies and plans, it is of crucial importance to examine the current emergency management system. It is also pivotal to reflect back on previous disasters and learn lessons from them to avoid committing the same mistakes again. It is saddening to discover that most emergency policies implemented are either out-of-date, not fully documented or not easily obtainable. This chapter will look at current hazards and vulnerabilities in Saudi Arabia. It will also provide a list of major disasters in Saudi history, and describe the current emergency management policies in the country. Finally, lessons learned from these disasters and areas of improvement will be critically discussed. The Kingdom of Saudi Arabia The Kingdom of Saudi Arabia is located in western Asia. It takes up most of the Arabian Peninsula, with a surface area of 2,149,690 km2 and a population of 27.137 million (Central Department of Statistics and Information, 2010, United Nations: Statistics Division, 2008). Of this population, 30% are 14 years or younger and only 4.75% are 60 years or older. International migrant stock, such as guest workers, represent 27.8% of the total population (Ministry of Economy and Planning, 2010-2014). Saudi Arabia’s population living in rural areas makes up 18.6% of the total population. The geography is varied, from coastal regions in the eastern and western parts, to mountainous regions in the south-west, and finally to the Rub’ al Khali desert running along the country’s southern boarders where almost no life exists. The country is divided into 13 provinces which are further divided into governorates; each of these has a capital that is headed by a governor. Figure 1 shows a simplified map of Saudi Arabia and its major cities. 1 Medical student and PhD candidate (MBChB/PhD) at Christchurch School of Medicine, University of Otago and Van der Veer Research Institute, P.O. Box 4345, Christchurch 8140, New Zealand. E-mail: yasao@hotmail.com 1 Figure 1. Map of Kingdom of Saudi Arabia (source: United States Central Intelligence Agency) Hazards in Saudi Arabia A hazard can be thought of as a potential risk endangering human life or health, property or the environment. However, if this risk does lead to an incident, it is referred to as an emergency situation or, if the damage is overwhelming, a disaster. Such events are often the result of human factors, environmental hazards or natural causes. Although considerable overlap occurs between these factors, there is usually one factor that contributes significantly more than the others. This section will review hazards in Saudi Arabia classified according to the main contributory factor. 1. Human-related risks: • Terrorist attacks: Up until recent years, terrorist attacks have very rarely, if at all, been heard of in Saudi Arabia. Citizens and foreigners have co-habitated for decades, even before the foundation of the current Saudi Arabia. This was especially the case in areas known for trade, such as Jiddah on the Red Sea, where merchants from Syria, Lebanon, Jordan, Egypt, Yemen, Oman, and India regularly mingled and traded with local merchants. With the rapid modernization that occurred to the country, more and more citizenforeigner interactions were formed. This increased presence and power of foreigners in the Kingdom is viewed by some extremists as posing a “threat.” Lacking adequate knowledge of Islamic laws, they took out-of-context quotes from Holy Scriptures to 2 justify taking their souls, along with many others’ of their fellow citizens and foreigners. This has resulted in the unfortunate occurrence of several terrorist attacks in Saudi Arabia in the past few years (discussed later on in the chapter). Added to the human and structural losses, these bombings resulted in transient internal instability in the country, albeit brief, as well as interrupting public and international relations leading to an unprecedented shift in regional and international political dynamics. • Motor Vehicle Crashes (MVCs): MVCs are the leading cause of mortality and morbidity in Saudi Arabia. There have been almost 500,000 MVCs in 2008 alone, resulting in over 6,000 deaths (Ministry of Interior, 2008). This means that there are 1,350 MVCs, 101 people injured and 18 people killed everyday! This, in part, has been attributed to the social and economic development in the country, leading to a considerable increase in the numbers of drivers and vehicles. In turn, this has overwhelmed traffic services in urban and rural areas. Supporting this theory is the notable increase in MVCs and deaths seen during the special seasons on the Islamic calendar (discussed next). For example, the province of Makkah has witnessed more MVC-related deaths (26.02%) in 2008 than the rest of the 13 provinces of Saudi Arabia. The vast majority of MVCs result from driver-related offences, as opposed to road- or vehicle-related causes (Ministry of Interior, 2008). Driver-related offences can be divided into the following categories: road-code offences, vehicle misuse, driving misjudgments and other offences. Of these categories, road-code offences have been the most common, with overspeeding and running red lights having accounted for more than 50% of all MVAs in 2008. Of note, more than one-third of all MVA-related deaths are seen in the 18-29 years age group (which is most expected to undertake such driving stunts). Targeting such risk factors, therefore, has the potential of dramatically improving morbidity and mortality resulting from MVCs in Saudi Arabia. MVCs are on the rise internationally, but they are particularly problematic in Saudi Arabia. In a review of MVCs from all Gulf countries, Saudi Arabia had the highest incidence of accidents including pedestrians (Al-Tukhi, 1990). Not only has this been claiming the lives of many people in Saudi Arabia, but it has also been exhausting national resources that could be better utilized. • Ramadan and Hajj seasons: Ramadan and Hajj are two special seasons on the Islamic calendar, for which a massive influx of people from all over the world come to Saudi Arabia. Ramadan is the ninth month on the Islamic calendar, while Hajj occurs on the 12th month. Given that the Islamic calendar employs a lunar cycle, these events do not equate to a particular time on the Gregorian solar calendar (which is usually 11-12 days longer). This also means that these events cycle between seasons (i.e. summer, fall, winter and spring) every few years. During the fasting month of Ramadan, it is an Islamic belief that good deeds are exponentially greater. As a result, many Muslims from around the world make an effort to visit the Holy Mosques in Makkah and Medina to perform prayers and other rituals. This leads to a cumulative number of visitors of about 2 million people over a period of only 30 days. With this number of visitors, simple practicalities, such as when to perform physical prayers, can result in profound adverse effects that can exhaust available resources. For example, an observational study from Al-Noor Specialist Hospital in 3 Makkah has shown that most emergency department admissions were during the evening shift (4pm-12am). This was attributed to the fact that most patients were fasting and had been exposed to the high temperatures of summer while performing prayers during the day (Dhaffar et al., 2005). Hajj refers to the major pilgrimage to the Holy Mosque in Makkah, carried out over 5 days on the 12th month of the Islamic calendar. It is obligatory for each adult Muslim, physically and financially capable, to perform Hajj at least once in their lifetime. During Hajj season, there is an almost sudden increase in Makkah’s population from 200,000 permanent inhabitants to well over 3 million people. This increase puts major stress on Makkah’s modest supplies of food and water as well as its health services. In addition, the limited space in Makkah has raised concerns about pilgrim overcrowding and trampling, increased MVCs, spread of infectious diseases and other public health implications. 2. Technological hazards: Technological hazards refer to the partial malfunction or total breakdown of equipment leading to the early cessation of an operation short of its intended goal. Technological hazard is increasingly becoming a recognized separate category of hazards. Depending upon the type of operation ceased, technological hazards can result in power outages, environmental damage or health risks for the human workforce. Since Saudi Arabia is one of the leading oil-producing countries, this paragraph will focus on the risks posed by technological hazards in the oil industry. Technological hazards in the oil industry can occur at any stage of oil processing: from extraction to refinement to exportation. Some of the incidents that can occur include damage to oil wells, leaking pipelines, accidental ballast water discharge from loading terminals and accidental oil spillages. All Saudi factories involved in oil-related operations are very active in the protection and maintenance of the equipment in accordance with SASO (Saudi Standards, Metrology and Quality Organization) standards. Unfortunately, however, incidents still occur in spite of all precautionary measures, and there have been about 36 recorded oil spills in the Arabian Gulf alone as of 2005 (Al-Suwian, 2001). Several field studies from King Fahad University of Petroleum and Minerals have not shown any significant pollution of the Arabian Gulf by heavy metals or hydrocarbons (Al-Suwian, 2001). However, the Arabian Gulf is especially likely to become more polluted since it is enclosed and receives only a slow rate of water exchange with the open sea. It also has a high salinity and a rapid rate of water evaporation leading to an even higher salinity. All of this poses a great threat to living marine species, the ecological structure of the Gulf, as well as people working in the area. 3. Natural disasters: Saudi Arabia has recently become known for media-attracting incidents such as terrorist attacks and major MVCs. However, less attention has been given to natural disasters, even though their incidence has been on the rise. Floods are the most frequently encountered natural disaster in Saudi Arabia. They have been the cause of 7 of the 10 most damaging natural disasters in the history of the country between 1900 and 2010 (refer to Table 1). The reason behind floods being a major threat in Saudi Arabia is multi-faceted. Rains have been relatively scarce in the area, and this has lead to the under-development 4 of a proper drainage system in the country. Compounding this problem is the geography of some of the most populated cities in Saudi Arabia. Cities, such as Jiddah and Makkah, are on low ground and are surrounded by mountains. When rains fall on these mountains, water runs in valleys towards these cities. With poor drainage systems, this continuous flow of water could easily lead to a flash flood. Disaster Date No. Killed Flood 24/11/2009 163 Epidemic 11/09/2000 76 03/2000 Epidemic 57 Epidemic 9/02/2001 35 Flood 28/04/2005 34 Flood 24/12/1985 32 Flood 22/01/2005 29 Flood 4/04/1964 20 Flood 8/04/2002 19 12 Flood 11/11/2003 Table 1. Top 10 natural disasters in Saudi Arabia for the period 1964 to 2010, sorted by the number of people killed (source: International Disaster Database) Vulnerability in Saudi Arabia Vulnerability in any country can be gauged by how it prepares for and reacts to emergency situations and hazards. This section will examine vulnerabilities in Saudi Arabia in terms of emergency preparedness and reaction to emergencies once they occur. Emergency preparedness vulnerabilities Saudi Arabia has certain vulnerabilities that can hinder the country’s ability to be better prepared for hazards discussed previously. One of these is the short time available to prepare for high risk seasons, namely, Hajj and Ramadan season. This line-up of mass gathering seasons leaves no time for proper emergency preparedness projects. Usually, preparations of these seasons start at least a month before Ramadan. As people start to leave after the Ramadan season, more and more people arrive in Makkah in preparation for Hajj. This takes up the period leading to the actual Hajj season. After Hajj, at least two to four weeks are spent on cleaning the Holy Mosque and fixing any damage caused by the season itself rather than initiating new emergency preparedness projects. The scale and timing of these mass gathering seasons leave no choice for emergency planners but to operate on full capacity, and surge capacity of human and physical resources is almost null. Any extra resources are only used for increasing the operating capacity to handle more visitors rather than to increase the surge capacity. 5 Furthermore, with all the crowding during these mass gathering seasons, emergency preparedness activities take longer to establish and are more expensive to run because of the logistics and practicalities of establishing a preparedness program in a very crowded city (i.e. Makkah). Basically, the nature and timing of high risk seasons in Makkah make the population of Makkah and its visitors more vulnerable to disasters and its impact. Another, although less significant, factor to exacerbate the vulnerability of Saudi Arabia to the impact of potential disasters is the recent trend of reluctance from international experts, including emergency planners, to work in the country especially after the recent terrorist attacks (Maben et al., 2010). This has affected the progress of a wide range of collaborative developmental projects including emergency preparedness projects, for which more expertise and skill than available in the country is required. Emergency preparedness is based on experience-sharing, and international expertise is central to any readiness activities and without such expertise the vulnerability to the effect of disasters is multiplied. Vulnerabilities in reaction to emergencies A country’s reaction to emergencies once they strike determines the extent of the damage. Multiple factors could improve or hinder the reaction to emergencies. Saudi Arabic has several factors that could hinder recovery efforts and increase the vulnerability to disasters impact. These are usually social and demographic factors, such as the high rate of illiteracy and language barriers among vulnerable populations. Illiteracy and lack of proper education can negatively affect people’s attitudes towards emergency preparedness. In 2007, illiteracy rates were 23.6% in females and 8.6% of males over 15 years (Ministry of Economy and Planning, 2010-2014). Not being able to read safety brochures or use the internet and other media resources for public announcements can have adverse consequences and place the population on higher risk of being a victim of disasters. For example, during the rainfall that resulted in the flood in Jiddah in 2009 (discussed in the next section), many people ignored warnings about using motor vehicles for unnecessary trips simply because illiteracy means less attention to such messages. Some people under-estimated the risk and decided to take a trip in their cars to “enjoy” the rain, and these were the cars that were swept away by the flood and clogged main streets. Moreover, some people have the attitude that “what God wills to happen, will happen”; however, this contradicts Islamic beliefs. Islamic teachings state that every person has to do their best in taking precautions, as well as believing in God and relying on Him. In short, lower education level and illiteracy leads to less effective risk-communication and under-appreciation of the power of disasters. Many communities in Saudi Arabia have a higher vulnerability to the impact of disaster because people do not appreciate risks and ignore official messages. Another problem is the language barrier among immigrant workers in Saudi Arabia. Immigrant workers made up 53.1% of the workforce in Saudi Arabia in 2008 (Ministry of Economy and Planning, 2010-2014). In spite of this large number, most precautionary warnings issued by officials during disasters are still publicized in Arabic! There has been a call for occupational emergency personnel who can speak languages most commonly used by foreign workers (e.g. Urdu and Filipino); attempts to date have been unsuccessful. The media is still largely in Arabic and less of other languages. This 6 miscommunication leads to increased vulnerability of minority groups in Saudi Arabia who are labor workers living in high risk areas. In summary, the vulnerability to disasters and their impact is compounded in Saudi Arabia by multiple factors, such as the nature of the mass gatherings, the high illiteracy rate and miscommunication of risk to minority groups. These factors all tend to slow down preparedness activities and make recovery after disasters even slower. History of disasters in Saudi Arabia Almost all major disasters in Saudi Arabia can be attributed to one or more of the hazards and vulnerabilities mentioned in the previous sections. Unfortunately, there is no official publicly-available database that keeps a record of disasters in the country. Most official information available comes from newspapers local to the region where the disaster occurred. The International Disaster Database (IDD) of the WHO provides the best record of disasters in Saudi Arabia (International Disaster Database, 2010). For this section, data recorded in the IDD have been compared to information published in the relevant medical literature as well as in local newspapers around the time of any given disaster to check for accuracy (2000, Aguilera et al., 2002, Almulla, 2008, Lerner et al., 2007, Thompson et al., 2004). Table 2 shows a chronological list of major disasters in the past 50 years in Saudi Arabia. The following is a description of the most significant disasters in the history of Saudi Arabia: 1964 rains: this is the earliest recorded account of a natural disaster in Saudi Arabia. Heavy rains poured continuously on parts of the country leading to a flood that killed 20 people and left about 1,000 people either injured or homeless. No further details are recorded. Fire incident in Hajj season 1975: during Hajj season in 1975, a fire broke out in one of the pilgrim’s tents near Makkah and quickly spread to other tents. The fire was caused by an explosion of a gas cylinder, and led to the death of 200 pilgrims. Seizure of the Holy Mosque in Makkah: on 20 November 1979 the Holy Mosque in Makkah was occupied by a group of armed Muslim extremists. The attackers had planned to seize the Mosque by filling coffins with weapons and smuggled them into the Mosque. On the morning of the day of seizure, they chained the gates of the Mosque, killed the two guards on-duty at the time, and held present worshippers hostages. They called on the people to revoke the current Saudi Monarchy and obey their leader, Abdullah Hamid Al-Qahtani. After more than two weeks of cross-fire with the Saudi Army, and with the help of Pakistani and French forces, the siege of the Mosque was ended. At least 250 people were killed and 600 injured, including worshippers, troops and insurgents. The surviving insurgents were captured by Saudi authorities and later executed. Ras al-Khafji thunderstorm: in October 1982, a severe thunderstorm hit Ras al-Khafji city on the east coast of Saudi Arabia. Hail stones were reported to be as big as tennis balls. This was followed by four hours of heavy rains. The net damage included 11 fatalities. 7 Type of disaster Heavy rains Fire during Hajj Militant occupation of Holy Mosque in Makkah Floods in north-western Saudi Iranian riots during Hajj Stampede inside pedestrian tunnel during Hajj Date April 1964 December 1975 November 1979 No. affected 1,000 NDA 600 No. killed 20 200 250 December 1985 5,000 At least 32 July 1987 July 1990 649 NDA 402 1,426 Fire during Hajj April 1997 343 Rift Valley Fever outbreak Jizan floods September 2000 April 2004 More than 1,500 500 87 NDA 430 5 Destroyed 2,680 km2 of hoses, lands and roads $900,000 Jiddah floods Effect estimates NDA NDA Help from Pakistani and French forces $450,000 NDA Compounded by failure of ventilation system inside the tunnel November 2009 More than 163 10,000 Table 2. Top 10 disasters causing major damages in Saudi Arabia between 1960 and 2010 (NDA = No Data Available) 1985 flood: on 24 December 1985, heavy rains poured on north-western regions of Saudi Arabia, leading to what has been described as the worst flood in the area in 50 years. Estimates of damage were not recorded, except that there were at least 32 people killed from the flood. Iranian riots in Hajj 1987: in July 1987, the Civil Defense forces and Saudi Police had to open fire against Iranian demonstrators after arguments escalated to fights between the two parties. This incident claimed the lives of 402 people, and wounded 649. This led to political tension between the two countries, and Iranian pilgrims were held from entering Saudi Arabia for Hajj seasons 1988 and 1989. Stampede in Hajj season 1990: as pilgrims were moving between the sacred sites on the second day of Hajj season in 1990, a massive stampede occurred in a tunnel south of Makkah. The stampede occurred after what is thought to be a failure in the ventilation system inside the tunnel. This led to the suffocation and death of 1,426 pilgrims, most of whom were from south-east Asia. Stampede in Hajj season 1994: During one of the rituals of Hajj, a stampede occurred as pilgrims leaving the site crossed roads with those coming in. This led to a massive disorder culminating in the death of 270 pilgrims, most of whom were trampled. 8 Khobar tower attack: on 25 June 1996, a terrorist truck bomb (estimated to carry 20,000 pounds of TNT equivalents) exploded in Dhahran, eastern Saudi Arabia. The attack was aimed against troops of US Air Forces, US Army and coalition forces who billeted in Khobar towers military compounds. The attack resulted in the death of 19 people and the injury of 555 people. 1996 Charkhi Dadri mid-air collision: even though this tragic event occurred outside the country, it deserves to be mentioned since it is considered the deadliest mid-air collision in history. On 12 November 1996, Saudi flight 763 was en route to Saudi Arabia from India when it collided with Kazakhstan Airlines flight 1907. All 349 people onboard both flights were killed. Yanbu flood: heavy rains poured on western Saudi Arabia in January 1997, mainly affecting Yanbu and peripheries of Jiddah. The rain lasted for 24 hours, killing 10 people and causing damage to an area of over 130,000 km2 of land. Asir flood: Asir is a province in the Southwest of Saudi Arabia. On Monday 25 March 1997, heavy rains poured on the region, leading to floods that resulted in 16 fatalities and damaged an area of just below 100,000 km2 of land. Fire incident in Hajj season 1997: in April 1997, a gas stove exploded in one of the pilgrim’s tents, leading to a massive fire that quickly spread to other nearby tents. It claimed the lives of 343 pilgrims, and more than 1,500 were wounded. This stimulated authorities to design the currently used fire-proof tents, as well as banning gas-operated material. Meningitis outbreaks in Hajj and Ramadan: outbreaks of N. meningitides serogroup W135 have been reported from as early as 1987. In the Ramadan of 1992, an epidemic occurred, but all cases have been confined to residents of Saudi Arabia. However, in Hajj season 2000, another outbreak of the same infection occurred, only to include pilgrims from various countries this time. This had led to the spread of the infection to countries from which those infected pilgrims came. The reported cumulative number of deaths is 57, but is likely to be considerably higher. The 2000 Rift Valley Fever outbreak: beginning in early September 2000, it had been noticed that goats and sheep were being found dead in some areas of the far south of Saudi Arabia. Soon after, reports of hemorrhagic fevers from the same region started to increase, which had subsequently been identified as Rift Valley Fever. The Saudi Ministry of Health declared an epidemic (i.e. the first of its kind in Saudi Arabia), and advised citizens to wear mosquito repellants. Areas where dead animals were found were quarantined; live stock in endemic areas were checked and exterminated if found ill. At least 87 people died and more than 500 people were afflicted by this infection. Makkah 2002 flood: heavy rains started falling on Makkah area on 8 April 2002 and lasted for a whole week. This led to flooding of water in some areas, claiming the lives of 19 people; hundreds of Makkah residents were rescued by the GDCD that week. 9 Makkah 2003 flood: not quite recovered from previous year’s rain, Makkah experienced yet another heavy shower described as the worst rains in Makkah in 25 years. Water levels were reported to have reached 6 meters. Twelve people were killed; however, estimates of physical damage are not available. Riyadh 2003 bombings: on 12 May 2003, attacks on three different housing compounds were conducted by a group of nine radical terrorists. These sites are thought to have been chosen because they contained a large number of Westerners and non-Muslims. Seven vehicles, packed with explosives, gained entry into the compounds after attackers killed the guards. The attackers then detonated their bombs and the vehicles, resulting in a significant damage to buildings and vehicles and leaving large craters. Thirty-four people were killed and 194 were injured. Jizan 2004 floods: less than four months apart, two floods hit the Jizan region, leading to what has been described as Jizan’s worst floods in 45 years. The floods left over 400 people homeless, killed 13 people and devastated many roads and farms. Medina 2005 flood: very heavy showers fell on Medina region in January 2005. This resulted in a flood that caused the Yatamah dam to fail, killing 29 people. Seventeen people were injured, 50 were left homeless and 43 had to be evacuated. Riyadh 2005 flood: heavy rains poured on the Riyadh region of Saudi Arabia, as well as on other areas in neighboring countries (i.e. Oman and the United Arab Emirates). The resultant flood claimed the lives of seven people; 700 people had to be evacuated via GDCD helicopters and another 700 were left homeless. Hostel collapse in Makkah: in Hajj season 2006, a hostel near the Holy Mosque collapsed after a fire had spread in lower floors of the building. Most pilgrims were out in the Mosque as it was time for the noon prayer. The collapse killed 76 people, most of whom were people passing by the building, and another 64 were injured. Jiddah 2009 flood: at around 6:30 a.m. on Wednesday 25 November 2009, rain started falling heavily in Jiddah, and continued for around 12 hours. The amount of water in this relatively brief downpour (around 90 mm3) doubled the average annual rainfall in Jiddah. With a sound infrastructure and a proper drainage system lacking, this rain turned into the worst disaster that Jiddah has experienced in 27 years or so. The downpour resulted in the formation of water tides coming from the hills on the east of the city, heading west towards the Red Sea and cutting their way through the city. Several residential houses collapsed, forcing many inhabitants to upper floors and roofs. Labs and databases at King Abdulaziz University and King Abdulaziz Hospital were destroyed, wasting valuable resources, specimens and medical records. Major roads of the city were blocked by meters-high of water waves or by cars that have been washed out. As a result, thousands of pilgrims had to wait in buses for hours before getting to Makkah for the first day of Hajj. Furthermore, King Abdullah 10 Bridge on the South of Jiddah had partially collapsed, adding to the chaos and fright to the situation. Power and telecommunication services were not spared either. As early as 11 a.m., floods had already resulted in a temporary power outage on the whole western region of Saudi Arabia (i.e. Makkah, Medina and Jiddah). Many people were not even able to call for help as communication with emergency services (e.g. civil defense forces, police or emergency medical services) failed due to the overwhelmed network and power outage. Overall, 161 people lost their lives as a result of the floods, either drowning or from car crashes. This disaster had an estimated cost of around US$900 million to reconstruct Jiddah and help its victims. 11 Photos of Jiddah 2009 flood (source: personal communication) Riyadh 2010 flood: on 3 May 2010, Riyadh city experienced a brief 45-minute water shower, accompanied by light hail and winds gusting up to 24 km/hour. As brief as the downpour was, however, it resulted in floods and car crashes across the city. Local newspapers reported that at least two people were killed, and that the floods caused around 275 car crashes. Even though King Khalid International Airport was not affected, many people missed their scheduled flights due to poor road conditions. A survey committee, appointed by the Governor of Riyadh, has started assessing the extent of and the reasons behind the damage that resulted from the rain. Development of emergency management plans in Saudi Arabia The development of emergency management plans in Saudi Arabia started more than 80 years ago, and has been progressing slowly since then. The first nucleus of an emergency management body was a fire brigade that was formed in Makkah in 1927 (Ministry of Interior, 2001). Its purpose was to serve pilgrims that came to Makkah every year. It was the first of its kind in Saudi Arabia, and it was managed by the Makkah 12 Provincial Council. In 1948, the Makkah Fire Brigade joined the later-established Center of General Security to form the General Security and Fire Services. Over the following 32 years, the General Security and Fire Services grew to include 5 fire brigades in Makkah alone. Meanwhile, fire brigades formed in a number of other cities including: Medina, Jiddah, Riyadh, Qasim, and Dammam. In 1965, a Royal Decree by King Faisal dissolved the General Security and Fire Services, and instead formed the current General Directorate of Civil Defense (GDCD). This was following recommendations by the International Association of Fire Fighters. The scope of the GDCD was wider than previous emergency management bodies because it was made the official body of civilian defense during peace and in times of instability. During this time, the GDCD received generous funds to expand its human and material resources. In addition, the GDCG centers started operating in more and more urban and rural areas in the Kingdom with the help of the evolving telecommunication networks. Later on, in 1987, King Fahad ordered a reform of the GDCD’s structure, goals and responsibilities. As a result, staff from the GDCD administration paid several visits to neighboring and other friend countries to investigate civil defense advancements and useful experiences in these countries. After extensive meetings, the current Civil Defense Law was decreed, which included 36 sections. The following is a translation of two sections relevant to emergency management: Section one defines “civil defense” as protocols and operations required to protect civilians as well as public and private properties from the dangers of fires, natural disasters, wars and other accidents. It also encompasses rescuing those afflicted by such catastrophes, ensuring transportation safety, and protecting national resources in times of peace and emergency. Article four of section two defines the role of the GDCD in emergencies and wars as: 1. Organizing and operating the national alarm system in cases of emergencies or attacks by a foreign army. 2. Managing electrical power, and organizing evacuation and shelter plans. 3. Extinguishing fires and rescuing civilians and providing basic life-support measures in damaged areas. 4. Marking areas afflicted by nuclear damage, and directing civilians away from them. 5. Corresponding with other governmental bodies (e.g. Ministry of Transportation) to ensure safe transportation of civilians. 6. Removing debris from damaged areas, and rehabilitating them for safe use as soon as possible. Current structure of the General Directorate of Civil Defense The current structure of the GDCD is divided into three levels: Board of GDCD, Executive Committee, and volunteers (Ministry of Interior, 2001). The following is a description of the members of each level, and the most important roles for which each level is responsible. Figure 2 shows a schema of the current structure of the GDCD. 1. Board of GDCD: 13 This is made up of the Minister of Interior as Chairman, Assistant Minister of Interior as Deputy-Chairman and a number of members who represent divisions of the GDCD or sectors that work closely with the GDCD, such as fire services, police and emergency medical services. Those members are appointed by a Royal Decree often after the recommendation of the Chairman or his deputy. The Board of GDCD is responsible for: 1. Establishing general GDCD policies and planning future projects. 2. Establishing safety and fitness standards that must be met in all projects to ensure civilian safety and protect public and private properties. 3. Establishing guidelines for training programs for GDCD personnel. 4. Establishing policies for the recruitment of GDCD volunteers and defining their roles and rights. 5. Forming divisions of the GDCD, defining their responsibilities and appointing a manager to each division. 6. Reviewing the suggested budget annually before seeking approval from the Ministry of Finance. 2. Executive Committee: This committee consists of members appointed by the ...
Purchase answer to see full attachment