Reduce Exposure to Reduce Risk

3. SYSTEM ARCHITCTURE of data or web services. The geoportal hosts the This section presents the high level and low level provided data and makes it available for the user to architecture of the ONG. search, discover and consume the data and services. 3.1 High Level Architecture Figure 3 shows the high level user interaction. The provider publishes the data to the geoportal in form Geoportal Publish data and services Search, discover and consume data and services Provider User Web Service Data Figure 3: ONG High level user interaction geoportal, which consists of vector data and raster Figure 4 shows the high level system architecture, data provided in form of various consumable services where the user connects to the central catalogue over and downloadable format. using internet. The central catalogue is hosted on the Country 1 catalog HTTP Country 2 catalog Country 3 catalog User Country 4 catalog Central catalog Country 5 catalog Figure 4: ONG High level user architecture Each country has its own catalog which is embedded in the central catalogue; an additional security layer prevents the unauthorized access of the catalog Reduce Exposure to Reduce Risk 240

File Geodatabases and MXDs Raster Data (File System) Managed Server with Esri’s ArcGIS Server and Arc Viewer SSeeccuurriittyy LLaayyeerr IInntteerrffaaccee LLaayyeerr Data Analysis Data Access Services Services Navigation / Display/ Print / Query / WMS WFS FTP Identify Select / Draw / Measure Internet Country User Country User Country User Country User Figure 5: ONG Low level architecture 3.2 Low Level Architecture as a front screen / link on Asia-Pacific Gateway for As shown in figure 5 the country user is able to DRR and development. access the ONG through the user interface and the security role. The security roles can be customized REFERENCES for each user, as individual user or as a group. Berry, J. (2009, September). GIS and the Cloud- Internally the ONG consists of a managed server and Computing Conundrum. GeoWorld. Retrieved from geodatabases. The geodatabases are further divided http://www.geoplace.com/ME2/dirmod.asp?sid=DA7 into file geodatabases and raster data. The reason for 2DA013599412F85B2FD29498DD7E3&nm=a+test this segregation is to enable the country users to &type=MultiPublishing&mod=PublishingTitles&mid access the data more efficiently and securely. The =2F0B36C074B04B3DAACB3F3733414366&tier=4 geodatabases can’t be accessed independently, hence &id=7F932400DCF44F8E9ACF43A7536A77D5 adding an additional layer of security. Figure 5 also shows the services that will be provided Esri, A. (2007a). Geospatial Service-Oriented by the ONG. The services provided are divided into Architecture ( SOA ), (June). two sets Esri, A. (2007b). Developing and Deploying an a. Data Access Services Integrated Geoenabled SOA Business Solution : A b. Data Analysis Services Case Study, (June). Data access services consist of REST, WMS, WFS and FTP, this will enable the end user to access the Esri, A. (2009). ArcGIS ® Server in Practice Series : data by different means depending upon his Best Practices for Creating an ArcGIS Server Web requirement. Data analysis services are a basic set of services that will be provided to the user to perform Mapping Application, (August). basic analysis on the data. These services can be provided on internet as well on the client side. Esri, A. (2011). Estimating the Cost of a GIS in the Amazon TM Cloud, (January). 4. CONCLUSION Keeping in view the architecture and users of the Goodchild, M. F. (2007). Geographic information online national geoportal it is highly recommended sharing: the case of the Geospatial One-Stop portal. that the managed remote server setup should be Annals of the Association of American Geographers, adopted as the client server model for the deployment 97(2), 250–266. of the online national geoportal. The major advantage would that no additional human resource would have Kouyoumjian, V. (2010). The New Age of Cloud to be hired by the stakeholder to manage the licenses Computing and GIS. ESRI white paper. Retrieved for Windows Server, ArcGIS Server, Esri Geoportal from and ArcGIS Desktop. In addition to this periodic data http://downloads2.esri.com/campus/uploads/library/p backups will be marinated by the hosting service dfs/96771.pdf provider hence creating an automatic archive, cost comparison also shows that managed remote server Mell, P., Grance, T., & Grance, T. (2011). The NIST setup is also a cost effective solution when compared to other models. And the flexibility of the proposed Definition of Cloud Computing (Draft) architecture would allow it be seamlessly integrated Recommendations of the National Institute of Standards and Technology. NIST Special Publication. Retrieved from Reduce Exposure to Reduce Risk 241

http://scholar.google.com/scholar?hl=en&btnG=Sear ch&q=intitle:The+NIST+Definition+of+Cloud+Com puting+Recommendations+of+the+National+Institut e+of+Standards+and+Technology#1 Tait, M. G. (2005). Implementing geoportals: applications of distributed GIS. Computers, Environment and Urban Systems, 29(1), 33–47. doi:10.1016/j.compenvurbsys.2004.05.011 Reduce Exposure to Reduce Risk 242

AUTOMATED DISASTER MANAGEMENT SYSTEM FOR EARLY RECOVERY THROUGH GIS M. P. A. W. Gamage, K. P. D. H. De Silva, S. Janahan, S. T. Mackmillian, R. H. Daniel, S. D. T. N. Siyambalapitiya Sri Lanka Institute of Information Technology [email protected], [email protected], [email protected], [email protected], [email protected], [email protected] ABSTRACT : A superior disaster management solution would be one that is integrated with web-based technology that people are already familiar with, and this is the idea behind this particular disaster management program. One prime example of tried and tested web based solutions is Geographical Information Systems (GIS). Through integration with various Google Apps including Google Maps, this research aims to develop a system which allows users and administrators to exploit existing technology with greater efficiency and share larger volumes of information with each other. Apart from GIS, the system will also exploit basic features of smart phones to rapidly share key information, especially concerning navigational visual data. This facility used GPS and GPRS techniques to route tracking services and also neural network concepts for prediction of specific disasters. Neuroph jar files were used to feed the prediction system with past data for 10 years and error back propagation algorithm was used to train the system. Therefore, this system combines existing technological solutions to create a powerful and fully-accessible disaster management system that can be rapidly deployed and used by virtually among every class of user. KEYWORDS : Disaster Management, Geographical Information System (GIS), Global Positioning System (GPS), General Packet Radio Services ( GPRS ) 1. INTRODUCTION Disaster Management completely relies on data from The Disaster Management System (DMS) proposed various sources. Data has to be gathered, organized here, is a web system, which is completely properly and displayed accordingly to take any further implemented using JAVA and MySQL to provide actions. It‟s not a right time to search and gather data most viable option for solving pre and post disaster during disaster or after the disaster for disasters may management related problems in Sri Lanka. The also have affected the government organization which system will help to save lives and enhance relief provides data. services for the people who are affected by disaster. This research was conducted with the aim of gathering Disaster management can be divided in to five major and organizing related information through GIS factors according to their behaviors. (Johnson, 2000) technologies in order to efficiently used during and These phases are involved with different services and after a disaster. related to one another. The research was conducted The existing systems, which were developed for the focusing on these five factors. above purpose have been riddled with various flaws and limited functionalities. They are as follows Planning: -Contains the activities to analyze and document the disaster impacts. It measures the Inadequate monitoring mechanism: The key dangers, risks, mitigation, alertness, response, and problem with the existing systems is the inability to recovery needs. properly report and manage the resources, aids, Mitigation: -Contains the tasks which are undertaken information, communication etc. before the disaster to reduce the damages during the disaster. Lack of IT knowledge of users: This problem Preparedness: -Involves the activities needed to requires long-term educational and training programs increase when mitigation measurements have not for organizations involved in disaster relief and local prevented disasters. The interested parties develop governments in the areas effected. plans to save lives and minimize damages. Response:-Contains activities following a disaster. Slow progress regarding reporting and monitoring: These activities are designed to provide emergency Another concern is that despite access and knowledge assistance for victims such as shelter, food and of IT, users had to spend a relatively large amount of medicines. their time on the system to update or gather Recovery:-It includes the activities which are taken up information. to bring the situation back to normal. This can be divided into two phases, such as short term response 243 plans and long term response plans. Reduce Exposure to Reduce Risk

Prior to the disaster occurring; programs need to have Telephones play an important role in warning flexibility to adjust and scale up easily so that they can communities about an impending disaster. In some respond in a timely manner. Similarly, capacity aspects countries, mechanisms called „telephone trees‟ are and coordination efforts among government, non used to warn communities of impending danger (Sahu, government, private sector and other actors will 2010) greatly affect the effectiveness of the system. Information systems can play a major role in the One of the recent changes in technological trends is the effectiveness of natural disaster management rise in the popularity of smart phones. Unlike an strategies. Similarly, recent advances in poverty ordinary phone, a smart phone comes with a variety of mapping techniques have improved the estimate of navigational and operational features that give users vulnerable populations such as orphans or people with unprecedented amount of technological flexibility. disabilities, providing a more accurate input for These features could be easily exploited during a planning (Vakis, 2006) disaster and help provide better disaster management related information gathering. The integration of smart Considering above factors, this research aims to phones with GIS will enable accurate real-time develop a system which is more reliable, efficient with information to be shared from across different areas more functionality. And it was implemented to manage within a disaster zone. This is made possible to the all sorts of needs before a disaster occurs. And also it integration of a smart phone‟s Global Positioning is implemented to be usable by all levels of people in System (GPS) with Google‟s “Google Maps” service. an easy manner. When the GPS is activated, it will provide the user 2. METHODOLOGY with his latitude and longitude coordinates. This DMS consists with various types of users and will be information will then be automatically transferred to having different levels of privileges according to their Google maps, which will display the disaster point on needs. Users can be connected with the system through the map. Hence, it will provide users with an internet using PCs, smart phones and normal phones. interactive map that can be updated in regular Server will be connected with a database server to intervals. GPSs are generally considered to be very store and retrieve the information needed. accurate and in some instances, they can provide accurate positioning within 20 meters of the target. Server Database Apart from displaying a target position on a map, the Users integration of GIS and GPS will also help improve the data communication amongst different users and a Figure 1 System design central command/admin. Some valuable information that could be shared can include the number of This system is expected to be utilized as a disaster affected personnel, the status of infrastructure, the preparedness collaboration tool and aims to help to status of emergency relief supplies, etc. Users can recover from disasters at earliest possible. upload this information in real-time and improve the overall efficiency of a disaster relief program because 2.1 GIS Based Disaster Information Updating scarce resources can be transferred to areas that need All phases of disaster management depend on data them most. from a variety of sources. During an actual disaster it is critical to have the right data, at the right time, For the system to be utilized properly, the chief task of displayed logically, to respond and take appropriate the admin would be to ensure that the information is action. easily accessible across various platforms and to send GIS facilitates this process by allowing planners to out email or SMS alerts to users in the area. Looking view the appropriate combinations of spatial data out for key information and then sharing it across through computer generated maps. Values at risk can volunteer organizations will help improve the disaster be displayed quickly and efficiently through a GIS. management program and reduce the frequency of Utilizing existing databases linked to geographic errors. More importantly, even ordinary mobile phones features in GIS makes this possible. will be able to work with the GIS. Ordinary phones will be able to send the information manually to Reduce Exposure to Reduce Risk central database, which will display the information on Google Maps. This could be done via GPRS and SMS. 2.2 Flood Prediction Repetitive flood-related incidents arise on a yearly basis that is not as significant in extent, but they also endanger property and lives of citizens and so there is a need to react adequately. To protect the property and inhabitants effectively it would be useful to have a system available that would continuously track the 244

development of meteorological conditions, state of stage of the crisis, it is vital to get as much information water in stream channels and be capable, pursuant to as possible about critical infrastructure and aid the information and also based on the existing distribution areas. experience, of assessing on a nearly real-time basis the potential threat of floods. (Rapant, 2010) Emergency personnel often need detailed information concerning water supply, infrastructure details, The Flood prediction is an early warning system, hospital details, electrical distribution and etc. which is capable of learning from a natural language Utilizing a map based computerized system; all source and predict when necessary. Most of existing departments can share information through databases research findings have not focused on building agent on computer-generated maps in one location. Without with machine learning and also they were not able to this capability, emergency workers must gain access to implement fully because of major limitations. And no a number of department managers, their unique maps, machine learning systems are being used in a country and their unique data. Most emergencies do not allow like Sri Lanka to predict the natural disasters. That time to gather these resources. This results in indicates it still needs more expert human assistance to emergency responders having to guess, estimate, or operate in an environment like in Sri Lanka. And the make decisions without adequate information. This information needed to predict the flood is under costs time, money, and in some cases lives. GIS various organizations. This was one of the major provides a mechanism to centralize and visually hallucinations of the field of artificial intelligence display critical information during a disaster. DMS which is still a miracle. uses GIS to fulfill data requirement needs for planning and emergency operations and GIS become the In this research, Neuroph jar files were sued to do the backbone of disaster management. Disaster neural network functionalities to predict the possibility management activities are focused on three primary of flood using data for past 10 years. The flood objectives. These objectives are protecting life, prediction system will take the measurable factor that property, and the environment. causes the flood and learns some patterns how and when the flood is occurring and how much area being The DMS includes the following key functions: affected because of it. And then the neural network is to learn and recognize patterns based on past 1. Communication and navigation data: This will information and provide predictions according to the include contact numbers and interactive maps (Google pattern. Maps) to help users contact key institutions like hospitals and police stations to verify their operational Figure 2 : Test Data for Prediction status and other information. 2. Operational information: If these institutions are The network was trained using error back propagation functioning, the system will allow administrators to algorithm for result. The objective of training is to enter key details such as available personnel, capacity, reduce the global error. In the back propagation equipment, etc. This will help disaster management algorithm, an attempt is made to reduce this global officials to evaluate the effectiveness of these error by adjusting the weights and biases. This institutions and then divert the necessary resources involves minimization of the global error using depending on the requirements. steepest descent or gradient-descent approach. The 3. Navigational details for volunteers and affected network weights and biases are adjusted by moving a persons: information, depending on the condition of small step in the direction of a negative gradient of the roads and other transport networks, will be shared with error function during each iteration. The iterations are volunteers and those who have been affected by the reported until a specified convergence or number of disaster. This information can be elevated through iterations is achieved. real-time updates or satellite information. This will help people move to hospitals or report at police 2.3 Institute and infrastructure mapping stations in a quickly and efficient manner. The most crucial period of disaster management is 4.Direct communication: After the status and right immediately after the disaster strikes. At this operational capabilities of key institutions have been verified, the hospitals, police stations or other Reduce Exposure to Reduce Risk authorized persons will be able to send special updates regarding casualties or medical requirements. This will improve the frequency and information awareness available to organizers. 2.4 Route Tracking When a natural disaster such as a flood or an earthquake occurs, it causes the transportation network to suffer a large-scale destruction. Formerly several flood and earthquake experiences, the destroyed network system is possible to affect the disaster area rescue and the emergency evacuation work execution. 245

Therefore, when the natural disasters occur, how to reduce needless duplication or excess supply to one rapidly restore the disaster area network system‟s basic particular area and vice versa. capability is an important research issue. Specially, the disaster relief work should consider the \"gold 72 Volunteer registration: to prevent a misallocation of hours\" principle. (Wang, 2006) One of the main volunteers amongst different disaster zones, it is objectives of this research work is to provide a important to have a proper volunteer registration systematic framework in analyzing the network system in place that will help administrators deploy reconstruction plan after the network suffering the volunteers depending on the circumstances and natural disasters destruction. severity of different disaster zones. This is important because deploying volunteers also means deploying The route-tracking feature will be based on Google equipment and aid supply maps. The system will provide detailed information Aid distribution: To help with the proper distribution regarding the movement and location of key areas. of aid, the system maintains key details of disaster Once the details of a relief camp or some other zones and relief camps. The information will include institution have been entered into the system, it will be the number of affected personnel, the required displayed on a Google map. While traveling through a equipment, and any special requests. particular route, users will be able to update the status Inventory management: Another interrelated issue of the roads and other infrastructure. Moreover, if they with aid distribution is the management of aid supplies come across any difficulties with traveling, they can or inventories. An effective management of the post an update and alert administrators. Afterwards, inventory will help administrators to effectively this information can be used to plot out a different distribute resources amongst different disaster zones route or figure out a different method of transportation. and make request for resources depending on the fluctuations of demand and supply. 2.5 Missing People Tracking 3. CONCLUSION Another major tragedy of a natural disaster is no The reliability of a disaster management system is efficient way of finding missing people and identifying crucial and depends on the speed and accuracy of the unidentified dead bodies. In Sri Lanka, this fact information gathered at a disaster zone. To increase was demonstrated following the aftermath of the 2004 the accuracy of the prediction, the system should store tsunami, where thousands of victims were affected and large set of historical data. And to be an effective real many listed missing and as unidentified dead bodies. time system it should be able to connect to data sources such as Dept of Meteorology, Ministry of According to literature survey, the key problems are Disaster Management, Dept of Irrigation etc.. to the inability to properly monitor and report on the update information. In Sri Lankan context, a system status of missing people. with all three languages would encourage the users to get facilities through the system. This research aimed DMS provides commenting, multiple photo uploading, to use the GIS & GPS facilities, connect to a web notifications, sharing etc. It helps to spread the missing application and mobile application to implement person information quickly and efficiently within the aspects of a disaster recovery situation and there by public. And there are also functionalities available to provide minimum of loss during earthquake, land check whether more than one person is reporting for a slide, flood or tsunami. same missing person and then the system will connect the people who are reporting together. REFERENCES Johnson R., GIS Technology for Disasters and Emergency The system also has functionalities to keep track of the Management, New York: ESRI, 2000, unidentified bodies. It enables to predict if the person who reported as missing is in the registry of the Rapant P, Unucka J, Vondrák I, regional flood early warning unidentified bodies. System uses the information system, LVI ed., GeoScience Engineering, 2010. matching techniques to match the missing person and the unidentified bodies and output the percentage of Sahu R. S. A, guidebook on technologies for disaster matching to the responsible authority. For information preparedness and mitigation , LVI ed., Asian and Pacific matching, system uses the information such as age Centre for Transfer of Technology (APCTT). 2010. group, blood group, eye color, dress information (dress type, dress color) , tattoo marks, birth marks and etc. Vakis R., Complementing Natural Disasters Management : the Role of Social Protection, February 2006. 2.6 Volunteer registration, Aid management Wang C. A study on emergency evacuation a n d r e s c u e One fundamental feature of post-disaster management network reconstruction for natural disasters with multi class in recent history is the misallocation of resources. users travel behavior constraints, Journal of the Eastern Asia Most often, it is not the lack of resources that hamper Society for Transportation Studies, vol. 6, 2005. aid efforts but the inability to properly allocate those resources to people in need at the appropriate time. 246 This is true with both volunteers and aid material. Any effective disaster management program should help Reduce Exposure to Reduce Risk

CLOUD COMPUTING AS A NOVEL APPROACH TO NATURAL DISASTER MANAGEMENTS IN SRI LANKA J.M.D.R. Menike1 Ranjith Premalal De Silva2 1PGIA, University of Peradeniya,Sri Lanka, 2 Uva Wellassa University, Sri Lanka. ABSTRACT During the last decade Sri Lanka experienced several environmental disasters such as tsunami, flood, landslides, droughts, extreme wind events and lightning. The most devastating natural disaster was the tsunami which hit on eastern and southwestern parts of the island. It was estimated nearly 30000 to 50000 people were dead and about 100,000 of people have been affected. The major reason for this much of defeat is the lack of awareness about the disaster. In addition to that draw backs on emergency management, insufficient resources, poor communication have made the situation worst. To overcome this sort of issues the government initiated the disaster information management system with the aids from UNDP. This web site provides database about the natural disasters from 1974 and allows civilians to upload the data about the natural disasters. The government web site says there is a scarcity of systematic, homogenous compatible records on typology of disasters and small and medium scale disasters are hidden from the society which can create problems in decision making. Foresee requirement to get rid of these issues is computer based solution known as cloud computing. Cloud computing facilitates to share databases, stored data and geospatial data via cloud. That is very important in situation like Tsunami which is new to civilians and no previous records exists. Not only the row data, but the instructions from expertise, scientists, and experience from different countries can share via the cloud. In cloud computing end user can access to the cloud even via light weight computers with broad band connection, mobile phone apps or web browsers. During a natural disaster many local area networks, hardware devices, storage data get damage. Natural disasters can severely damage internet access which makes difficulties to connect with cloud. Using mobile phones apps and other portable computer devices with broad band connection help to get access to cloud. Not only is that, on the spot data uploading and updating is possible as all the databases were linked to the cloud. Clouds provide 24 hours 7 days service. They store data in high security. That provides high level of computing. Since the platform and infrastructure are major components of the cloud the software updating, upgrading and versioning are looked out by the cloud provider(Bhatt 2011). The major challenge is the secure and privacy of the data. Most of the cloud providers such as Google, Microsoft pay big attention to client’s data security and privacy. In future cloud computing will be the upcoming technology for natural disaster management. PREDICTING THE DISTRIBUTION OF PORT-AU- PRINCE’S IDP CAMPS James F. Bramante Tropical Marine Science Institute, National University of Singapore, E-mail: [email protected] ABSTRACT In January, 2010 an earthquake struck Haiti near its capital of Port-au-Prince, causing possibly the largest urban natural disaster in modern times. Within a week of the earthquake, dozens of informal camps were erected across Port-au-Prince by victims of the earthquake, termed internally displaced persons (IDPs). Most of these camps were formed spontaneously by those whose homes were destroyed during the earthquake, or by those who feared aftershocks would make their current dwellings unsafe. These IDP camps became the focus of humanitarian efforts and this in turn drew more IDPs to the camps, increasing their size and density. Ostensibly the first settlers of these camps chose the sites according to some decision-making model, even if as simple as the first large open space near one’s damaged home. This paper attempts to determine the extent to which the results of these decisions can be explained using geographic factors, such as topography, population density, and availability of open space. A logistic regression model was successfully created to predict the location of these informal IDP camps as a function of these variables. In addition, a multiple regression of a restricted set of Reduce Exposure to Reduce Risk 247

these variables explains at least 20% of the variance in IDP camp size across the study area. This paper includes a discussion of the statistical methods and pitfalls involved in geostatistical analysis of this form, with implications for emergency planners. KEY WORDS: Haiti, Internally Displaced Persons, IDP Camps, Geostatistics, Logistic Regression LAND USE / LAND COVER MAPPING OF NAGAPATTINAM BLOCK AND DISASTER MANAGEMENT USING GEOMATICS Tamil Selvan Vajravelu*, Ramakrishnan*, Ramalinagam*, Ganesan. G **, Ghouse Mohamed Shaik** *Inst of Remote sensing, ** Sri Venkateswara College of Engg and Technology Anna University Chennai ABSTRACT The earth has abundance of renewable and non-renewable natural resources. However, sub optimal, haphazard and unscrupulous exploitation of these resources leads to irreversible deterioration of the environment in the long run. Such an adverse scenario is evolving, rather rapidly in coastal areas, and livelihood is becoming unsustainable and the future generation is deprived of their rightful stake on these natural resources. Three out of the four metro cities of India are located on the sea coast. The coastal regions are densely populated due to access to deltaic agricultural land, rivers, marine resources etc. At the same the area is associated with natural calamities like cyclone, flood, Tsunami, coastal erosion etc. In due course of development, the human population unknowingly disturbs the production potential the natural resources. As a result environmental degradation is accelerated and supporting capacity of the nature is severely eroded. For sustainable development in coastal regions, there is an urgent need for proper management of existing resources. Hence strategies like land use planning, water resources conservation and augmentation etc are to be evolved at once. Site specific technologies are to be identified and promoted. The land cover / land use information for a particular year and their change over the years is necessary to assess the rate of exploitation of the resources and for identifying suitable sites for installing appropriate technologies for sustainable development. Nagapattinam is a costal block of Tamil Nadu state located in eastern coast of India. It forms a part of Nagapattinam District situated in Bay of Bengal. The block is frequently prone to natural calamity by storm and heavy rains. It is one of the coastal blocks devastated by Tsunami in 2004. It is categorized as saline block. It has 15.50 km of coastline and a maximum width of 11.50 Km from coastline. The geographical area is 186 Sq. Km. In the present study land use and land cover maps of the study area for the year 2003 and 2010 are prepared using IRS- 1C, LISS-III imagery of 2003 and Google Earth imageries of 2011. Collateral data such as Topo sheet, Panchayat maps are consulted for mapping. Extensive field visits are under taken for ground truth verification of the identified land use / land cover classes. The land use / land cover of study area is classified up to three levels. Significant changes in land use / land cover are observed. The geographical distribution of land use land cover changes are detected, quantified and tabulated. The data may be used for identifying suitable site specific technologies for sustainable development of the coastal block. Reduce Exposure to Reduce Risk 248

Poster Reduce Exposure to Reduce Risk 249

ESTIMATION OF SOIL LOSS USING GEO - INFORMATION TECHNIQUES Sreenivasulu Vemu, Udaya Bhaskar Pinnamaneni Department of Civil Engineering, JNT University, Kakinada, Andhra Pradesh, India-533003 E-mail: [email protected] ABSTRACT: Soil erosion is the greatest destroyer of land resources in Indravati catchment. It carries the highest amount of sediment compared to other catchment in India. This catchment spreading an area of 41,285 square km is drained by river Indravati, which is one of the northern tributaries of the river Godavari in its lower reach. In the present study, USLE is used to estimate sediment yield at the outlet of river Indravati catchment. Both magnitude and spatial distribution of potential soil erosion in the catchment is determined. From the model output predictions, it is found that average erosion rate predicted is 18.00 tons/ha/year and sediment yield at the out let of the catchment is 22.31 Million tons per year. The predicted sediment yield verified with the observed data. Generated soil loss map will be useful to soil conservationist and decision makers for watershed management. Overall 19.71 % of the area is undergoing high erosion rates which are a major contributor to the sediment yield (78.04 %) in the catchment. This area represents high-priority area for management in order to reduce soil losses, which are mostly found in upstream of the catchment. It is indicated that the areas of high soil erosion can be accounted for in terms of steep unstable terrain, and the occurrence of highly erodible soils and low vegetation cover. KEYWORDS: Soil Erosion, USLE, Sediment yield, GIS, Remote sensing. 1. INTRODUCTION models have been developed over the past 50 years, Land degradation due to water erosion and globally. All the erosion models are developed by deterioration of water quality by point source and taking the existing models into consideration. non-point source are some of the main problems in most of the watersheds in India. In addition to losses The Universal Soil Loss Equation (USLE), in its of soil, many other problems are created by soil original and modified forms, is the most widely used erosion: siltation of reservoirs, canals and rivers; model to estimate soil loss from watersheds (Rao et deposition of unfertile material on cultivated lands; al, 1994). That the various parameters of USLE can harmful effects on water-supply, fishing, power be derived from rainfall distribution, soil generation; and destruction of fertile agricultural characteristics, topographic parameters, vegetative land. According to Ministry of Irrigation in India, as cover and information on conservation support much as 175 Mha i.e. 53 % of geographical area is (erosion control) practice are often available in the subjected to serious environmental degradation. form of maps or can be mapped through collection of Nearly 60 % of the cultural area is suffering from the data from possible sources. Due to geographic nature effects of erosion, taking the toll of the land at the of these factors USLE can easily be modeled into rate of 5 to 7 Mha each year (Balakrishna, 1986). GIS (Jain 1994). The USLE model applications in the Study on global soil loss has indicated that soil loss grid environment with GIS would allow us to analyze rate in the U.S. is 16 t/ha/yr, in Europe it ranges soil erosion in much more detail since the process has between 10 – 20 t/ha/yr, while in Asia, Africa and a spatially distributed character (Ashish Pandey et al. South America, between 20 and 40 t/ha/yr (Pimentel 2007). The GIS and Remote Sensing (RS) provide et. al., 1993). spatial input data to the model, while the Universal Soil Loss Equation (USLE) can be used to predict the In assessing soil erosion, researchers always confront sediment yield from the watershed. with the problem of selecting the appropriate model to use in a given area (Meijerink and Lieshout 1996). In the present study, USLE is used to estimate It is always important to adopt a suitable model that potential soil erosion from Indravati catchment. Both can be applied to the critical conditions of an area magnitude and spatial distribution of potential soil (Chisci and Morgan 1988). Some models are area- erosion in the catchment is determined. An ArcGIS specific and may not perform well in other areas, package is used in developing digital data and since they are designed with a specific application in another GIS package Integrated Land and Water mind (Shrestha 2000). Therefore, selection of a Information Systems (ILWIS) is used for processing proper model suitable for an area should be the first remote sensing data. ILWIS is also used in spatial step in erosion modeling. Numerous soil erosion Reduce Exposure to Reduce Risk 250

data analysis to determine magnitude and spatial computes average annual soil loss (A) which is a distribution of potential soil erosion. product of five different factors that affect soil loss, 2. STUDY AREA and is given by: The Indravati is one of the northern tributaries of the Godavari in its lower reach. The Indravati catchment A = R K LS C P lies between latitudes 18º 27΄ N to 20º 41΄ N and longitudes 80º 05΄ E to 83º 07΄ E (Figure 1). The river Equation 1 Indravati rises at an altitude of about 914 m near Thuamal Rampur village in the Kalahandi district of Where, A = average annual soil loss in tons per Orissa on the western slopes of the Eastern Ghats and hectare, R = rainfall-runoff erosivity factor joins Godavari at an altitude of about 125 m. The (MJ/ha.mm/h), K = soil erodibility factor main river flows for a length of about 477 Km. The (t.ha.h/ha/MJ/mm), LS = topographic or slope Indravati basin with a catchment area of 41285 Km2 length/steepness factor, C = cover and cropping- constitutes 13.32 % of the total Godavari basin. The management factor, P = supporting practices (land basin has high hills, deep valleys and large plateaus. use) factor. All of the factors are dimensionless, with The mean annual rainfall in of this area is about 1288 the exception of R and K. The preparation of spatial mm, most of which occurs between May and data base for this model is explained below. September. Average potential evaporation rates are 6.5 mm per day, while average minimum and 3.1.1 Rainfall Erosivity Factor (R) maximum temperature are 13ºC and 39 ºC The erosivity factor R is often determined from respectively. There are no major irrigation projects rainfall intensity if such data are available. In existing in the study area. The major land covers in majority of cases rainfall intensity data are very rare, the catchment are forest (68 %), followed by consequently attempts have been made to determine agriculture (22 %). Agriculture is the main erosivity from daily rainfall data (Jain et al., 2001). occupation of the people in the area. In River Indravati catchment, no station has rainfall intensity data. Therefore R is determined using mean Figure 1: Location map of Indravati Catchment annual rainfall as recommended by Morgan and Davidson (1991). The expression is given below. 3. METHODOLOGY 3.1 Soil Erosion Model - USLE R = P * 0.5 Techniques for prediction of soil loss have evolved Equation 2 over the years. The most widely used equation for soil loss prediction of the catchment is the Universal Where, P = mean annual rainfall in mm and R = Soil Loss Equation (USLE). The USLE equation rainfall erosivity factor in MJ/ha.mm/h. A 20-year time series of monthly girded average precipitation dataset from the Climatic Research Unit -Average Climatology 2.0 (CRU-CL2.0) (http://www.cru.uea.ac.uk/cru/data/tmc. htm) from 1982 to 2002 is used in preparing R factor layer. Inverse distance method, which is very fast and efficient weighted average interpolation method in ILWIS, is used to show spatial distribution of mean R factor values in Indravati catchment. 3.1.2 Soil Erodibility Factor (K) The soil erodibility factor (K) represents both susceptibility of soil to erosion and the amount and rate of runoff, as measured under standard plot condition. In the study area no detailed soil map in the large scale is available. The soil map prepared by National Atlas and Thematic Mapping Organization, Department of Science and Technology, Government of India on 1: 2 Million scale is used to prepare K factor. 3.1.3 Slope length and steepness factor (LS) The topography affects the runoff characteristics and transport processes of sediment on a watershed scale. A 90 m resolution DEM from the Shuttle Radar Topography Mission (SRTM) is downloaded from ftp://e0mss21u.ecs.nasa.gov/srtm/ , and gaps of no data is filled with coarser Gtopo 30 DEM Reduce Exposure to Reduce Risk 251

(http://lpdaac.usgs.gov/gtopo30/hydro/index.asp).Thi and/or strip cropping to that with straight row s rectified 90 m resolution DEM is used to prepare farming up-and-down slope. As there is only a very the LS factor as discussed below. small area has conservation practices in the study area, P factor values are assumed as 1 for the basin Slope Length Factor (L): Mc Cool et al. (1987) 3.2 Sediment Yield Estimation presented the following relationship to compute the The ratio of sediment delivered at a given area in the slope length or L factor: stream system to the gross erosion is the sediment delivery ratio for that drainage area. Thus, the annual L = (λ/22.1)m sediment yield of a watershed is defined as follows: Equation 3 where L = slope length factor; λ = field slope length SY = (A) (SDR) (m); m = dimensionless exponent that depends on slope steepness, being 0.5 for slopes exceeding 5 Equation 6 percent, 0.4 for 4 percent slopes and 0.3 for slopes less than 3 percent. A grid size of 100 m is used as Where, A = total gross erosion computed from field slope length (λ). Similar assumption of field USLE, SDR = sediment delivery ratio. A general slope length is made by several researchers (Onyando equation for computing watershed delivery ratios is et al., 2005; Fistikoglu and Harmancioglu, 2002; Jain not yet available since they depend on several et al., 2001). properties of the watershed like infiltration, roughness, vegetation cover, hydrograph or runoff Slope Steepness Factor (S):For slope length longer drainage, etc. Since much of the above data are not than 4 m, the slope steepness factor is derived using available for the study area to derive SDR, some of the following equations (McCool et al., 1987): the simple models given by different researchers have been tried to estimate sediment yield at the S = 10.8 sin θ + 0.03 (for slope gradient < 9 %) outlet of the basin, but the one given below by Williams and Berndt‟s (1972) is finally chosen Equation 4a because it gives reasonable results despite using few S = 16.8 sin θ − 0.05 (for slope gradient ≥ 9 %) catchment characteristics. Equation 4b SDR = 0.627 SLP 0.403 Equation 7 where S = slope steepness factor and θ = slope angle Where, SLP = % slope of main stream channel. in degree. The slope steepness factor is 3.3 Spatial Distribution of Soil Loss dimensionless. After completing data input procedure and preparation of the appropriate maps as data layers, 3.1.4 Cover (C) and Conservation practices (P) they are analyzed in the GIS, to provide a estimate of the gross erosion map on 200 m X 200 m pixel size. factors Average soil loss is calculated as the product of each pixel value with pixel area then dividing with total The C factor is derived from NDVI distribution area of the basin. The USLE model is applied for the following two scenarios. obtained from Landsat images downloaded (via http://edcsns17.cr.usgs.gov/EarthExplorer/) on internet. NDVI is positively correlated with the amount of green biomass, so it can be used to give an indication for differences in green vegetation coverage. NDVI-values are scaled to approximate C 3.3.1 Estimation of Average Annual Soil Loss Average annual soil loss is estimated based on 20- values using the following formula, developed by year average rainfall erosivity factor and K, LS, C, P factors. European Soil Bureau: C = e – α (NDVI/(β-NDVI)) 3.3.2 Estimation of yearly soil loss Equation 5 Among the rainfall data, soil erodibility, slope- length, cover and support factors, rainfall data is where; α, β are the parameters that determine the available on yearly basis for the period of 1992 to shape of the NDVI-C curve. An α-value of 2 and a β- 2002. Therefore, R factor is computed for each of the value of 1 are assigned to give the reasonable results years. Using yearly R factor and keeping the remaining factors as constant, yearly soil loss (after Van der Knijff et al., 2000). (t/ha/yr) is calculated. P is the conservation practice factor, reflects the impact of support practices in the average annual erosion rate. It is the ratio of soil loss with contouring Reduce Exposure to Reduce Risk 252

4. RESULTS AND DISCUSSION 0 and 4 cover 78 % of the catchment area, and only The results obtained by analyzing the data are 10 % of the catchment has LS values greater than 11. presented and discussed in this section. The average Spatial distribution of C factor was derived for the annual R factor values vary from 550 to 670 MJ.mm year 1998 and the C value in the study area varies ha-1 h-1 with a mean value of 602 MJ.mm ha-1 h-1 and from 0.1 to 0.3. After completing data input a standard deviation is 25. The K value in the study procedure and preparation of the appropriate maps as area varies from 0.6 to 0.8. DEM of the study area data layers, they were multiplied in the GIS, to revealed that 42 % of area between altitude from 500 provide a estimate of the gross erosion map on 200 m m to 700 m. The combined spatial distribution of LS X 200 m pixel size. Gross erosion map was factor is derived using the DEM of the study area. LS reclassified (Figure 2) as per the guidelines suggested factor values in the study area vary from 0.2 to 587 by Singh et al. (1992) for Indian conditions. with a mean value of 4. Areas with LS value between Table 3: Computed and Observed Values (CWC, India) of Sediment Yield for Different Years Figure 2: Spatial distribution of soil loss Year Gross Predicted Observed in Indravati catchment erosion sediment sediment 92-93 ( t ha−1yr−1) yield at the yield at the 93-94 94-95 17.64 outlet outlet 95-96 18.76 ( t ha−1yr−1) ( t ha−1yr−1) 96-97 21.02 5.31 4.98 97-98 18.99 5.65 2.42 98-99 14.83 6.33 7.27 99-00 14.39 5.72 4.48 00-01 16.00 4.46 2.07 01-02 19.25 4.33 2.20 Average 17.60 4.82 1.61 20.2 5.79 3.61 17.87 5.29 1.76 6.08 4.15 5.37 3.46 Table 1: Soil loss and erosion risk classes From the model output predictions ( Table 1) it is found that on average, 74.11 Million tons of soil are Soil Erosion Area Area Soil Soil loss moved annually per year and average erosion rate loss risk (Km2) (%) loss (%) predicted is 18 tons/ha/year. Sediment delivery ratio (t ha-1 Classes (Millio (SDR) for the catchment is found to be 0.3 using the y-1) 22400.88 54.26 n tons) 7.81 empirical equation of Williams and Berndt. By Slight multiplying the gross erosion with SDR, sediment <5 5.79 yield at the out let of the basin is found to be 22.3 Million tons per year. The observed average annual 5 – 10 Moderate 6584.12 15.95 4.59 6.20 sediment yield at Pathagudem gauge site which is obtained from Central Water Commission (CWC), 10 – 20 High 4161.32 10.08 5.89 7.95 Government of India, at the out let of the basin support our results (Table 2). Almost half of the 20 – 40 Very 3433.12 8.32 9.81 13.24 Indravati catchment (54.26 %) falls under slight high 2638.72 6.39 14.84 20.03 erosion risk class where soil loss is lower than 5 t h-1 y-1 (Table 2). Areas covered by moderate, high, very 40 – 80 Severe high, severe and very severe erosion potential zones are 15.95, 10.08, 8.32, 6.39 and 5.01 percent > 80 Very 2066.76 5.01 33.17 44.77 respectively (Table 1). Overall 19.71 % of the area is Total severe 41284.84 100.00 74.11 100.00 undergoing high erosion rates which are a major contributor to the sediment yield (78.04 %) in the Table 2: Computed and observed values catchment. This area represents high-priority area for (CWC, India) of sediment yield management in order to reduce soil losses, which are mostly found in upstream of the catchment. Station Duration Observed Computed % Name & No. Million tons error About 32 and 42% of the erosion area occur in Pathagudem Annual altitude zones between 200 to 400 and 500 to 700 m, (AGG00B5) average 21.21 22.31 + respectively. About 32% of the total soil loss occurs (1992- 4.93 between 500 and 700 m, while altitude zones 2002) between 300 to 400, 400 to 500, 700 to 800, and 800 Reduce Exposure to Reduce Risk 253

to 900 m contribute almost an equal percent of total Fistikoglu, O., and Harmancioglu, N.B., 2002. soil loss (i.e., ~11%). Most of the slight to very Integration of GIS with USLE in assessment of severe risk erosion is also found in the zone between soil erosion. Kluwer Academic Publishers. 500 and 700 m. Therefore, it is essential to take Water Resources Management 16, pp. 447–467 conservation practices in this zone. Jain, S.K., 1994. Integration of GIS and remote Annual rainfall erosivity factor for different years sensing in soil erosion studies, Report No. (i.e., 1992–2002) are generated. Soil loss for each CS(AR)-186, National Institute of Hydrology, year is computed using respective annual rainfall Roorkee, India. erosivity layer. Sediment yield at the outlet of the basin is calculated by multiplying gross erosion with Jain, S.K., Kumar, S. andVarghese, J., 2001. SDR of the catchment, and the results are presented Estimation of soil erosion for a in Table 3. It is found that the model estimated Himalayanwatershed using GIS technique. highest sediment yield (6.33 t ha-1year-1) in the year Kluwer Academic Publishers. Water Resources 1994–1995 and lowest sediment yield (4.33 t ha- Management 15, pp.41–54 1year-1) in the year 1997–1998. Overall analysis indicates that USLE model has over predicted soil McCool, D.K., Foster, G.R., Mutchler, C.K., and loss, except in the year 1994–1995. Therefore, it is Meyer, L.D., 1987. Revised slope steepness observed that USLE can be used to predict long-term factor for the universal Soil Loss Equation. average soil loss rates in the basin than yearly Trans of ASAE 30(5), pp.1387–1396. prediction. Meijerink, A.M.J., and Lieshout, A.M.V., 1996. 5. CONCLUSION Comparison of approaches for erosion modelling A quantitative assessment of soil loss is made using using flow accumulation with GIS. HydroGIS Universal Soil Loss Equation for Indravati 235,pp. 437-444. catchment. All the thematic layers of R, K, LS and C are integrated to generate erosion risk map to find out Morgan, R. P. C., and Davidson, D. A., 1991. Soil spatial distribution of soil loss within the GIS Erosion and Conservation, Longman Group, environment. Since the USLE model does not take U.K. into account transportation and deposition, the actual sediment yield at the outlet is likely to be less than Onyando, J.O., Kisoyan, P., and Chemelil, M.C., the estimated. Generated soil loss map is also able to 2005. Estimation of Potential Soil Erosion for indicate high erosion risk area which is useful to soil River Perkerra Catchment in Kenya. Water conservationist and decision makers.. Resources Management 19, pp. 133–143 REFERENCES Pimentel David., 1993. World soil erosion and conservation (Edited), Cambridge University Ashish Pandey., Chowdary, V.M. and Mal, B.C., Press, U.K. 2007. Identification of critical erosion prone areas in the small agricultural watershed using Rao, V.V., Chakravarty, A.K., and Sharma. U., 1994. USLE, GIS and remote sensing, Water Resour Watershed prioritization based on sediment yield Manage , 21, pp. 729–746. modeling and IRS-1A LISS data, Asian-pacific Remote Sensing Journal, 6(2), pp. 59-65 Balakrishnan, P., 1986. Issues in water resources development and management & the role of Shrestha, D.P., 2000. Aspects of erosion and remote sensing, Technical report of ISRO, India, sedimentation in the Nepalese Himalaya: No.ISRO-NNRMS- TR. pp.67-87 Highland- Lowland relations. PhD thesis, Ghent University, Ghent Chisci, G., and Morgan, R.P.C., 1988. Modelling soil erosion by water: Why and how, In: Morgan Singh, G., Babu, R., narain, P., Bhusan, L.S., and RPC, Rickson RJ (eds) Erosion assessment and Abrol, I.P.,1992. Soil erosion rates in India. J modelling, Commission of the European Soil and water cons, 47(1). Pp. 97-99. communities report no. EUR 10860 EN, pp. 121–146 Williams, J.R., and Berndt, H.D., 1972. Sediment yield computed with universal equation. J Hydrol Div, ASCE 98(12), pp. 2087–2098. Van der Knijff, J.M., Jones, R.J.A., and Montanarella, L., 2000 Soil erosion risk assessment in Italy, European Soil Bureau, EUR 19044 EN. Reduce Exposure to Reduce Risk 254

GEOMETRIC EXTRUSION OF 2D FOOTPRINTS TO 3D BUILDING MODEL FOR DISASTER MANAGEMENT Saher Murad and Rizwan Bulbul National University of Sciences and Technology, Institute of Geographical Information Systems H-12 Sector, Islamabad, Pakistan, {saher09, bulbul}@igis.nust.edu.pk ABSTRACT: 3D city models encompass great importance in GIS as they give the realistic presentation of the city features on the Earth‟s surface. In the context of a 3D city model, the main object that can be modeled is building owing to its great importance vis-à-vis enormous needs of society mainly living space to provide shelter from changing weathers and several others such as ensuring privacy, storage of valuables and belongings and obviously a place for rest and comfort. Ever increasing high rate of population which is commensurate to city development increases its vulnerability in case of natural disasters (flood, tsunami, etc.). For example, a 2D flood simulation will only identify the areas that are potentially to be affected. However it will not be able to determine the exact floors to be ultimately affected. 3D city model therefore is needed for an effective disaster management in order to reduce the losses and minimize damages to assets prior and after disaster. Existing 3D city models can easily be generated through several automatic and semi-automatic mechanisms involving expensive equipment and manual corrections which are most time consuming and also require operator's expertise. These mechanisms may not be befitting to produce great area of city model in a short period of time especially in emergency. A simple block model (e.g. CityGML LOD1) can even be an effective mean of countering disasters such as flood etc. as long as it does not only provide appropriate 3D visualization but also maintains geometry and topological accuracy. 3D building models have to be produced in a relatively short period of time and have to be easily accessible in the market. This research work has developed a methodology for automatic extrusion of 2D footprints to 3D building model having accurate topology and implicit geometry. An appropriate geometric model called Alternate Hierarchical Decomposition (AHD) is used for decomposing the regions in to convex components that facilitates extrusion. The 3D disaster model can be used to help the risk managers through optimum utilization of resources in decision making they need for natural disaster management. KEY WORDS: Disaster Management, 3D City Model, LOD, Automatic Extrusion, Topology, Geometry, Alternate Hierarchical Decomposition (AHD) 1. INTRODUCTION: which besides visualization also allows spatial Cities are dynamic living organisms that are evolving analysis. through interplay of regulatory and entrepreneurial activities [Hamilton, 2005]. Improvement of various 1.1 Purpose of Study: infrastructures of our living space has always been a 2D modeling gives results that do not take into worrying factor along with their growth. Advent of consideration the third dimension. For example, a 2D Geographical Information Systems (GIS) and flood simulation will only provide us with areas different mapping services through latest which are vulnerable to be affected. However unless technologies have given a tremendous momentum to we don‟t have the third dimension component built different stakeholders while ensuring availability of into our models, we will not be able to determine the up-to-date information thus greatly facilitating specific information such as the exact floors due to decision making process. With significant be affected. Similarly incase of terrorist attacks like improvements in the field of GIS, best management bomb blast etc. it would be much easier with of the cities in true 3D environment is now within availability of 3D component to ascertain the specific realm of possibility. Although different GIS buildings to be evacuated. The implication of this is techniques mainly focusing on visualization are therefore different to skyscrapers and small presently in practice such as city modeling from high buildings. The significance of 3D component in resolution aerial imagery, laser scanner data or semi- today‟s ever volatile environment thus proves to be a automatic modeling from high resolution imagery, necessity rather a mere relevance. Apropos, an we propose developing a methodology for automatic effective city management necessitates an accurate extrusion of 2D footprints to 3D building model 3D city model which is growing and expending quite having accurate topology and implicit geometry rapidly in variety of fields. Reduce Exposure to Reduce Risk 255

In a steady shift from traditional 2D toward 3D 2. METHODOLOGY: model, a great amount of accurate and efficient 3D 2.1 Datasets: city models have become necessary to be produced in a short period of time and provided widely on the  2D building footprints market [Takase et al., 2003].  building heights  disaster related data 1.2 Statement of the Problem: 2.2 Technique: Alternate Hierarchical Existing 3D city models use laser scanner data, 2D Decomposition digital map, aerial image, etc. which need expensive An algebraic approach i.e. Alternate Hierarchical equipment and manual corrections have to be Decomposition (AHD) is used for the decomposition performed. Moreover manual 3D modeling uses of non-convex polytopes into a set of convex hulls softwares, like Google Sketch Up, which besides that are represented hierarchically in a Convex Hull being very time consuming also requires operator‟s Tree, CHT. This technique is robust, efficient and expertise. Therefore, these mechanisms though very scalable to any dimensions [Bulbul et al., 2009]. useful may not be very effective for production of 2.2.2 Extrusion Algorithm: large area of city models in relatively short period of An extrusion algorithm of our own will be time. implemented to extrude the 2D AHD into 3D AHD. For developing countries, like Pakistan, where data 2.2.3 3D Disaster Model: acquisition may take considerable time, there is a Finally a GIS tool will be applied on 3D AHD and need for an automatic system which not only disaster data to generate an accurate assessment of precludes heavy processing but also produces risks. geometrically and topologically accurate building 2.3 Extrusion process: models very efficiently. Figure 1: Extrusion process for Geometric extrusion 1.3 Objectives: of 2D footprints to 3D Building Model for Disaster In this paper we will survey the state of the art for Management extrusion of buildings from footprints to:  generate 3D building model from 2D building footprints  3D disaster model The results will form on deriving a hypothesis as under: 1.3.1 Hypotheses: There exists a solution for the automatic geometric extrusion of 2D footprints to 3D building model, which has the following characteristics:  cost effective  efficient  implicit geometry 1.4 Study Area 3D city modeling has always been a neglected field in Pakistan, however over a period of time massive devastation caused by natural calamities across the world and vulnerability of numerous places in general and our coastal areas in particular to such disasters such as Tsunami, flood etc. has developed an unprecedented urge and desire into Government and public for effective fore-planning and in depth study of the threatened areas. The issue is more pronounced and relevant to big cities such as metropolis of Karachi owing to presence of large buildings and ongoing rapid developments and therefore merits being my area of study. Reduce Exposure to Reduce Risk 256

3. CONCLUSION: Brenner, C., 1999, Interactive modelling tools for 3d Owing to massive destruction caused by natural or building reconstruction. human motivated disasters, importance of effective recovery and rehabilitation increases manifold. Cornells, N., Leibe, B., Cornells, K., and Van Gool, Advent of 3D component has augmented the efficacy L., 2006, Visualization, and Transmission “3D of city management against such vulnerabilities and city modeling using cognitive loops”. Third threats. This is more relevant to developing countries International Symposium on 3D Data like Pakistan where this system is still in the process Processing. of maturity. In this paper, we mainly focused on automatic extrusion of 2D footprints to 3D building Ehlers, M. and Hijazi, I., 2009, Web 3d routing in model at a level of detail of LOD 1 in which we and between buildings. intend implementing concept of AHD, using convex decomposition for representation of the objects in Elberink, S. and Vosselman, G., 2005, 3d modelling CHT which is very useful for achieving specified of topographic objects by fusing 2d maps and level of detail. This fundamental geometric primitive lidar data. provides simple, comprehensive, highly efficient and consistent topological processing for spatial queries Groger, G., Kolbe, T., Czerwinski, A., and Nagel, C., to analyze the disaster affected areas in three 2008, Opengis city geography markup language dimensions. (citygml) encoding standard. Open Geospatial In today‟s fast paced world where high technology Consortium. coupled with state of the art equipment and impressive human skills dominate every field of life, Groger, G. and Plumer, L., 2011, How to achieve the concept of high causalities and massive damages consistency for 3d city models. GeoInformatica, from disasters as a fait accompli is simply outdated. 1-29. In paradox, effective fore-planning and preparation such as 3D modeling enable us to reduce our Gruen, A., 2008, Reality-based generation of virtual vulnerabilities and exposure to such disasters, environments for digital earth, International thereby reducing the risk considerably. Journal of Digital Earth. REFERENCES: Haala, N. and Brenner, C., 1999, Extraction of buildings and trees in urban environments, Hamilton, A., 2005, Urban information model for ISPRS Journal of Photogrammetry and Remote city planning, Journal of Information Sensing. Technology in Construction. Panchal, H.and Khan, R. and Sengupta, S., 2011, Takase, Y., Sho, N., Sone, A., and Shimiya, K., 2003, Gis-based smart campus system using 3d Automatic generation of 3d city models and modeling. related applications. International Archives of Photogrammetry, Remote Sensing and Spatial Pu, S. and Vosselman, G., 2006, Automatic Information Sciences. extraction of building features from terrestrial laser scanning. International Archives of Bulbul, R. and Frank, A.U. 2009, AHD: The alternate Photogrammetry, Remote Sensing and Spatial hierarchical decomposition of nonconvex Information Sciences. polytopes (generalization of a convex polytope based spatial data model). IEEE Rottensteiner, F. and Briese, C., 2003, Automatic Geoinformatics, 17th International Conference, generation of building models from lidar data 1-6. and the integration of aerial images. International Archives of the Photogrammetry, Ledoux, H. and Meijers, M., 2009, Extruding Remote Sensing and Spatial Information building footprints to create topologically Sciences of the ISPRS. consistent 3D city models. Journal of Urban and Regional Data Management, UDMS Schwalbe, E., Maas, H., and Seidel, F., 2005, 3d Annuals, 39-48. building model generation from airborne laser scanner data using 2d gis data and orthogonal Baillard, C., Schmid, C., Zisserman, A., and point cloud projections. Fitzgibbon, A., 1999, Automatic line matching and 3d reconstruction of buildings from Zlatanova, S., Van Oosterom, P., and Verbree, E., multiple views. International Archives of 2004, 3d technology for improving disaster Photogrammetry and Remote Sensing. management: Geo-dbms and positioning. Reduce Exposure to Reduce Risk 257

IMPACT OF MONSOON FLOODING ON RAMSAR WETLAND IN NORTH EAST INDIA Chitrini MOZUMDER and Nitin K TRIPATHI Remote Sensing and GIS, School of Engineering and Technology Asian Institute of Technology, Khlong Luang, Pathumthani 12120 Thailand E-mail: [email protected] ABSTRACT: Flood, the other word for devastation is also responsible for establishing the lateral connections which are important to maintain biodiversity and key ecological processes in freshwater environment. It is important to consider these lateral connections before designing an efficient and effective conservation area since disturbances such as pollution, flow alteration and the spread of introduced species are easily propagated through the hydrological networks established during flooding. Deepor Beel, one of the twenty five Ramsar wetlands in India is considered as largest storm water basin in the Brahmaputra valley of lower Assam, India, but highly eutrophicated due to excess nutrients from last 10 years. In this study, four water indices from Landsat image with ASTER DEM has been used for thresholding the inundated area from the other land use/land cover. Nine Landsat scenes from June 2000 to December 2001 were collected for the spatio-temporal variation analysis. It was found that the wetland, which barely covers an area of 927 ha, inundates area over 6000 ha during monsoon flooding every year. From the classification map of the Deepor Beel sub-basin it was found that about 4400 ha of the inundated area is covered by rice fields which are fed fertilizers (P, K, N) 1-2 times per year. Being the lowest elevated of the basin, Deepor Beel receives the inputs from rice fields and other inundated area during flooding. Along with other causes, these inputs act as catalyst for eutrophication in the freshwater wetland Deepor Beel. KEY WORDS: Wetland, Remote Sensing, Flood, Connectivity, Eutrophication 1. INTRODUCTION In this study an attempt has been made to portray the Wetlands, specifically those which are dependent on seasonal variation of the wetland Deepor Beel. We aquifers are highly affected by the large fluctuations developed rules for classifying the habitats within the in climatic variables leading to alterations in water wetland with help of 4 water indices and NDVI levels and aquatic vegetation which in turn influence (Normalised Difference Vegetation Index). We also the functioning of ecosystem processes (Gorham investigated the lateral connections those are 1991; Bauder 2005; Wang et al. 2012; Paillisson established during high flood period with the upland 2003; Liu et al. 2011; Ventelä et al. 2010). The in greater Guwahati city, Assam, India. spatial connections established during high flood period also plays crucial role in maintaining the 2. DATA AND METHODS ecological processes and biodiversity in freshwater 2.1 Data inland wetlands(Hermoso, Kennard, and Linke 2012; The Landsat (TM, ETM+) multi temporal data of 3rd Pringle 2001). During these interactions between the June, 25th October, 26th November 2000, 29th wetland and riparian zones/ uplands disturbances January, 2nd March, 3rd April, 8th July, 28th October such as pollution, flow alterations and spread of 2001 and 1st February 2002 (hereafter Jun 2000, Oct interacted species are easily propagated though the 2000 etc.) were used for the seasonal wetland change drainage networks (Pringle 2001). Deepor Beel, analysis. The system is designed to collect 15 m located about 10 km southwest of Guwahati city, resolution panchromatic data and six bands of data in Assam, India is considered as one of the large and the visible, NIR and SWIR spectral regions at a important riverine wetlands in the Brahmaputra resolution of 30 m. The DN values of the images valley of lower Assam (figure 1). It is surrounded by were converted to radiance based on a calibration the Bharalu basin in the east, the Kalamani in the curve of DN(Chander and Markham 2003; Negi, west, Jalukbari in the north and Rani and Garbhanga Kulkarni, and Semwal 2009). The radiance were reserve forests in the south. It lies between 91º35′ to simulated to reflectance values using 6s (Second 91 º 43′ E longitudes and 26º05′ to 26º11′N latitude. Simulation of a Satellite Signal in the Solar It is recognized as one of the most significant Spectrum) radiative transfer code (Vermote et al. wetland systems (10,000 acres) in the world under 2006). The ASTER obtained elevation data with 30 the Ramsar International Convention on Wetlands. Reduce Exposure to Reduce Risk 258

m spatial resolution was used which have estimated The DEM was reconditioned using ancillary drainage accuracies of 20 meters at 95 % confidence for maps prior to prepare the slope maps using ARC vertical data and 30 meters at 95 % confidence for Hydro Tools in ARC GIS 10. horizontal data (ASTER, 2009). Figure 1: Location map of study area 2.2 Wetland detection 2.3 Feature Extraction and Classification Several water indices have been used in this study to For the seasonal analysis, an object based approach delineate the wetland from the satellite images. These was adopted in this study to extract the “wet” indices are derived from arithmetic operation of two features of the images in ENVI EX. The ENVI EX bands of the satellite which also diminishes the noise uses an edge based segmentation algorithm based on components of the wavelength. The arithmetic a user selected scale level (Xiaoying 2009). Although representations of the vegetation and water indices we used the reflectance image for segmentation, are as follows(DeAlwis et al. 2007; Ji, Zhang, and image was classified by building rule sets based on Wylie 2009): the combination of water and vegetation indices which were applied to each scene to group objects Equation 1 into 5 cover types within the wetland such as open water, transition zone, light aquatic macrophyte, Equation 2 dense aquatic macrophyte and mudflat. For the lateral connectivity analysis, the Deepor Beel Equation 3 catchment was used for LULC (Land Use/ Land Cover) classification. The basin was delineated from Equation 4 the ASTER DEM. The study basin consists of hilly areas which leave shadow on the images. One major Where, NDWI stands for Normalised Difference challenge was that the classification scheme Water Index, MNDWI for Modified NDWI, NDPI developed here, over classified many of these shadow for Normalised Difference Pond Index and NIR, R, regions as water logged areas. SWIR, Green are the reflectance values of Landsat satellite. Reduce Exposure to Reduce Risk 259

Figure 2: Wetland cover types plotted with rainfall and temperature Patterns, 2000-2002 To overcome this, the slope layer prepared from “high aquatic macrophytes” extend. Mudflat class DEM was included to the ruleset with a value of was relatively static than other cover types Slope ≤ 5 °. The basin LULC was classified to 9 throughout all seasons. The areal extent of mudflat classes, water, forest cover, open hill, high dense ranged from 0 to 80 ha in Oct01 and Jan01 vegetation, medium dense vegetation, crop land, respectively. fallow land, barren land and built up. 3.2 Seasonal Lateral Connections 3. RESULTS AND DISCUSSION Connectivity, generally defined by Pringle (2001), is 3.1 Seasonal Changes in Deepor Beel the water-mediated transfer of energy, materials, and Deepor Beel, being mainly rain-fed, has remarkable organisms across a hydrologic landscape. Flooding in-lake changes in each season with very short and interactions between the aquifer and surface run- residence time. Figure 2 shows the trend of area off underline that the river itself is only one part of a changes of the cover types within the wetland (in- complex overall system which also comprises strong lake) in Jun00, Oct00, Nov00, Jan01, Mar01, Apr01, vertical exchange processes (Standford and Ward Jul01, Oct 01 and Feb02. The results show that the 1993). Although Deepor Beel is longitudinally wetland is highly dynamic with the cover types and connected to river Brahmaputra and city Guwahati changes almost on every month. The peak of open through Khanajan and Bahini (figure 1), it also water is in Oct-Nov and lowest in May-Jun. The experience every year flooding during monsoon highest extent of open water reached in Jan01 with an building the lateral connectivity. Hence, besides the area covering 365 ha. Interestingly, the open water nutrients and materials interchange through the areas obtained in Oct00 and Oct01 were almost same longitudinal connection, a lot of exchange process is covering 301 and 300 ha respectively. The lowest going on through the lateral connection every year. areal extent for open water was in Apr 01 with 33 ha As a consequence, very low or very high nutrient only. On cross checking it was found that rainfall levels should decrease species richness by selecting shows a similar pattern in these 2 years with a lag of specialized species, whereas intermediate nutrient almost 4 months. The maximum rainfall in 2000 was levels should favor the co-occurrence of species with in August intensity of 373 mm, whereas in 2001 the contrasting nutrient requirements. highest rainfall was observed in June with intensity of 343 mm. The curve of transition zone areas The lateral connections made by the wetland due to follows a trend almost opposite to that of open water change of season within a year are important to and rainfall (figure 2). The peak area was observed in understand “what is coming to the wetland” every the month of March with 346 ha and lowest values in year. Figure 3 shows the spatial pattern of October ranging 7 to 21 ha. The cover types low inundations of the wetland and upland for several aquatic macrophytes and high aquatic macrophytes months in 2 years. After the monsoon starts hitting complimented each other by opposing the peaks and this part of India in May-June, the maximum falls. The vegetation starts growing in Sep-Oct (the inundation was observed from Jun-Aug with area cover type “low aquatic macrophytes” reaches its ranging from 5307 ha to 5581 ha. peak) and full bloom in Mar-Apr when the peaks of Reduce Exposure to Reduce Risk 260

Jun 2000 Oct 2000 Jan 2001 Mar 2001 Apr 2001 Aug 2001 Oct 2001 Feb 2002 Figure 3: Wetland inundation patterns in several months of 2000-2002 Although there is a perennial direct connection respectively. It should be noted that these croplands between the wetland and river Brahmaputra, several are mainly rice fields which have 2 cropping cycles other surface connections are visible transferring the per year. The statistical handbook of Assam, India water and materials between them. The lowest wet shows that the fertilizers Nitrogen (N), Phosphorus area of 725 ha was observed in month of Mar01. We (P) and Murate of Potash (K) are added in the both have mapped the inundated areas and shown in cropping cycle which in turn reaches the wetland results that a total of 6850 ha area gets inundated during flooding. This is because this wetland is at the around the wetland (figure 3). From the classification lowest elevation of the basin. The croplands are map of the Deepor Beel sub-basin it was found that mainly rice fields which produce nitrogen and the major area is covered by light vegetated crop land phosphorus (Govt. of Assam 2009) which in turn and fallow land with 3265 ha and 1251 ha reaches the wetland causing the eutrophication. Reduce Exposure to Reduce Risk 261

5000 1989 Series of Remotely Sensed Images.” Hydrology 4000 2001 and Earth System Sciences Discussions 4 (3): 2011 1663-1696. Gorham, E. 1991. “Northern Peatlands: Role in the Area (ha) 3000 Carbon Cycle and Probable Responses to Climatic Warming.” Ecological Applications 1 2000 (2): 182-195. Govt. of Assam. 2009. Statistical Handbook of 1000 Assam. Guwahati. Hermoso, Virgilio, Mark J Kennard, and Simon 0 Linke. 2012. “Integrating Multidirectional Connectivity Planning for Freshwater Systems”: WaVtHeegrigehtaDteednsHeillVegetaVteiognetated land Crop landFaWlloawtelranLdogged Area Built Up 448-458. Cover Types Ji, Lei, Li Zhang, and Bruce Wylie. 2009. “Analysis of Dynamic Thresholds for the Normalized Figure 4: Land Use/ Land Cover types in Difference Water Index” 75 (11): 1307-1317. Liu, Hongjuan, Rencang Bu, Jintong Liu, Wenfang yearly inundated areas in wetland upland in Leng, Yuanman Hu, Libing Yang, and Huitao 1989, 2001 and 2011 Liu. 2011. “Predicting the Wetland Distributions Under Climate Warming in the Great Xing‟an 4. CONCLUSION Mountains, Northeastern China.” Ecological In summary, we have explored several aspects of the Research 26 (3): 605-613. wetland by using remote sensing and GIS tools and Negi, H. S., A. V. Kulkarni, and B. S. Semwal. 2009. some ancillary data. In this paper, an attempt is made “Estimation of Snow Cover Distribution in Beas to show the implication of yearly flood in a Basin, Indian Himalaya Using Satellite Data and freshwater Ramsar wetland. The rainfall is one of the Ground Measurements.” Journal of Earth System major factors of the seasonal variations of the inland Science 118 (5): 525-538. cover types of the wetland. Every year, most part of Paillisson, Jean-marc. 2003. “A Mass Balance the floating aquatic vegetation starts blooming from Assessment of the Contribution of Floating- end of January, reaches peak in April and gets leaved Macrophytes in Nutrient Stocks in an disturbed during the monsoon due to the flood water. Eutrophic Macrophyte-dominated Lake” 75: 249- Hence, from October to January, the highest amount 260. of open water can be seen from the space born Pringle, C.M. 2001. “Hydrologic Connectivity and datasets. However during the monsoon, a vast area the Management of Biological Reserves: a Global gets inundated and the wetland receives inputs from Perspective.” Ecological Applications 11: 981– the agriculture lands, built up areas, forests etc. 998. Although lateral connections are necessary for Standford, J.A., and J.V. Ward. 1993. “An transfer of materials and nutrients, over connectivity Ecosystem Perspective of Alluvial Rivers – may cause eutrophication as happening to the present Connectivity and the Hyporheic Corridor.” study area. Although, flooding and hence the lateral Journal of the North American Benthological connection is not only the reason for eutrophication, Society 12: 48-60. it has a major contribution in increasing the nutrient Ventelä, Anne-Mari, Teija Kirkkala, Amaury amounts as most part of the flooded upland consist of Lendasse, Marjo Tarvainen, Harri Helminen, and agriculture lands which gets 2 time fertilization per Jouko Sarvala. 2010. “Climate-related Challenges year. Hence, there must be also some threshold of the in Long-term Management of Säkylän Pyhäjärvi lateral connectivity which is indicated necessary by (SW Finland).” Hydrobiologia 660 (1): 49-58. most of the researchers on water resources. Vermote, E, D Tanre, J L Deuze, and M Herman. 2006. Second Simulation of a Satellite Signal in References the Solar Spectrum - Vector (6SV) User Guide Version 3. USA. Bauder, E. T. 2005. “The Effects of an Unpredictable Wang, Lin, Iryna Dronova, Peng Gong, Wenbo Precipitation Regime on Vernal Pool Hydrology.” Yang, Yingren Li, and Qing Liu. 2012. “A New Freshwater Biology 50 (12): 2129-2135. Time Series Vegetation–water Index of Phenological–hydrological Trait Across Species Chander, G., and B. Markham. 2003. “Revised and Functional Types for Poyang Lake Wetland Landsat-5 Tm Radiometric Calibration Ecosystem.” Remote Sensing of Environment 125: Procedures and Postcalibration Dynamic 49-63. Ranges.” IEEE Transactions on Geoscience and Xiaoying, Jin. 2009. “Segmentation-based Image Remote Sensing 41 (11): 2674-2677. Processing System.” http://www.faqs.org/ patents/app/2009012307. DeAlwis, D A., Z M Easton, H E Dahlke, W D Philpot, and T S Steenhuis. 2007. “Unsupervised Classification of Saturated Areas Using a Time Reduce Exposure to Reduce Risk 262

Study of low-latitude Ionospheric Scintillations during increasing solar activities using GPS – SCINDA data V Rajesh Chowdhary , Sanit Arunpold and N. K. Tripathi Asian Institute of Technology, Thailand Email: [email protected] [email protected] and [email protected] ABSTRACT: As sun is approaching its solar maximum, the solar activities are observed to be increasing rapidly due to its 11 year solar cycle. The increase in solar activities adversely affects Earth‟s Ionosphere and onto GPS signals reaching ground based receivers. That can be a major threat to GPS users worldwide. This paper presents a study of effects of solar activities on Ionospheric scintillations and Total electron content using GPS-SCINDA data from Thailand region. The data used are the scintillations index (S4) and vertical TEC (vTEC) observed at the AIT, Bangkok GPS-SCINDA station, 14.079 N Lat & 100.612 E Long. This study is based on the solar activity information during February to August, 2012. During these period new sunspots regions were observed to be forming on the surface of the sun. Also the geomagnetic indices have exhibited an intense phenomenon due to these solar activities. The data of such solar activities have been taken into account for this study. The result shows that scintillations were observed maximum during 8:00 to 22:00 hour‟s universal time. Scintillation index (S4) value reached upto 0.4 and very intense variations in TEC have been observed. Due to this solar activity, this data shows quite complimentary to assumption of intense scintillations during night time. KEY WORDS: GPS-SCINDA, Solar cycle, S4, vTEC, Geomagnetic Indices. 1. INTRODUCTION Technology, Bangkok, Thailand. The raw data Recent increase in solar activities has developed a obtained at the station has been processed by using panic across the scientists worldwide. These solar GOPI software to acquire the plots of TEC and S4 activities can lead to the major catastrophic events and the data thus obtained are compared with Space and can cause big threat to the life on our planet Weather Prediction Centre (SWPC) retrieved Earth. Every 11years or so, the activity on the Sun Geomagnetic Indices and GOES-15 datasets. reaches a peak. During this solar maximum, which Synchronized effects were observed during the every can extend to several years either side of the actual extreme event during the period of observation from peak, the Earth gets hammered by intense space February 2012- August 2012. The below figure weather [1]. When these storms released due to shows the current solar cycle 24 approaching to its intense solar activities are punched to Earth, it may peak in early 2013 based on the increasing number of lead to various phenomenal activities as well as sunspots being observed. major disasters leaving Earth out of electric power for years to come [2]. Last solar cycle observed to Figure 1: Solar cycle 24 sunspots prediction number occur during the year 2001. Thus, scientists expect [3]. early 2013 can be the solar maximum for this time. Extreme space weather caused by such intense solar storms or activities affects Ionosphere by forming it more ionized and expanding. Which means the ionosphere will contain more charged particles than before. These free charged particles can cause major disruption in GNSS signals worldwide. It is a known fact that ionosphere is a source of error in GNSS observations. But additional inclusive of free electron can cause tremendous errors in GNSS observations for most of the real time applications. This paper presents solar activity effects on GPS-SCINDA derived parameters Total Electron Content (TEC) and Scintillation Index (S4) station at Asian Institute of Reduce Exposure to Reduce Risk 263

One of the first known effects of space weather is the which encircles the Earth close to the magnetic fluctuations in the amplitude and phase of radio equator in the Van Allen (or radiation) belt of the signals that transit the ionosphere caused by the magnetosphere. The Dst is a geomagnetic index irregularity of electron density in the ionosphere [4]. which monitors the world wide magnetic storm level. The principal manifestation of a disturbed ionosphere It is constructed by averaging the horizontal on GPS signals is ionospheric scintillation, which if component of the geomagnetic field from mid- sufficiently intense to degrade the signal quality, latitude and equatorial magneto grams from all over reduce its information content, or caused failure of the world. Negative Dst values indicate a magnetic the signal reception. The word \"scintillations\" storm is in progress, the more negative Dst is the typically refers to rapid amplitude and phase more intense the magnetic storm [6]. fluctuations in a received electromagnetic wave. The cause may be diffractive when electromagnetic Also the plots provided from National Geophysical waves are scattered in an irregular medium composed Data Centre (NGDC) of NOAA about the GOES-15 of many small changes in the refractive index. The satellite were utilized to screen the initial data fluctuations in the signal intensity are quantified by selection of extreme events detection. Goes-15 avails the scintillation intensity index S4 which is defines as itself with the X-ray flux along with electron, proton follows: fluxes and magnetometer readings for whole month in real-time. From these plots one can easily extract S4 = {[I]2 – [I2]}1/2 /[I] the kind of solar-flare released onto the surface of Where: I = the signal intensity. Earth. And how intense these X-rays flares are and to which extent their values are varying. Thus, it can be S4 = the ration of the signal intensity standard utilized for screening process of data gathering for deviation by the signal intensity mean. extreme events. The Total Electron Content (TEC), one of the most 3. PREDICTION OF EVENTS important parameter to study the characteristics of In this paper, author has taken account into all the the ionosphere is defined as the integral of the major events that occurred during February 2012- electron number density along the signal path from a August 2012. Out of which 5 events can be satellite to a receiver [5]. considered as Extreme Events based on the X-ray graphs, Proton and Electron flux and intense impact 2. DATA UTILIZED AND METHOD OF on geo-magnetic indices of the Earth. Thus the ANALYSIS resultant effect on GPS –SCINDA station derived This work covers the period from Februray-August, Total Electron Content (TEC) and Scintillation Index 2012 of increasing solar activities as illustrated by (S4) were plotted and analyzed. First two extreme figure 1. The data used have been recorded by a events were observed in the month of March. Here NoVatel GPS receiver (model GSV 4004B) installed there has been intense disturbance in the X-ray flux in the framework of the SCINDA project in Asian along with proton and electron fluxes. A solar flare of Institute of Technology, Bangkok,14.079 N Lat & X-class was ejected from the surface of the Sun. Its 100.612 E Long. The parameters are recorded every effects were observed on TEC and S4 and were second and an average is made over each sixty (60) analyzed. Within just 3days again a solar flare of M- seconds and stored in a compressed file. The types of class was ejected into space from the surface of sun. files generated by a GPS SCINDA are differentiated The corresponding variations and disturbance in the according to their filename extensions. The TEC and scintillations were recorded and analyzed parameters for this study are the S4 and the TEC for this study. Third major event was observed during among the so called ionospheric statistics stored in the month of April. Where Dst Index has shown a the compressed files with the extension *.scn. The sudden change in its phenomenon. During 23rd-24th datasets chosen as extreme events were supported the index value was observed to be „-100‟, indicating from Space Weather Prediction Centre (SWPC) data the severity of the geomagnetic storm. Thus this of National Oceanic and Atmospheric event was considered for this study. Fourth & fifth Administration, for various geo-magnetic indices and events which were considered for this study are again solar flux, proton flux and electron flux datasets. due to the observed intense disruptions in Dst-index SWPC provides a real-time update of all these values during the month of June and July parameters for space weather prediction. The data respectively. In June, a sudden change the plot was provided from this centre was utilized to sort out all observed during 17th day of the month. The below extreme events along with the World Data Centre for graph shows the disturbance occurred due to Geomagnetism, Kyoto. WDC provides online real- geomagnetic storm. In July an intense downfall of the time plots of various geomagnetic indices, of which Dst-index value was observed during the 15th day of Disturbance storm time was used during the the month, which lead to include this day into the list preliminary elimination of the large data provided by of considered extreme events during this year. GPS-SCINDA station. The Dst index monitors the variations of the globally symmetrical ring current, Reduce Exposure to Reduce Risk 264

Figure 2: GOES-15 summary plots indicating Figure 6: Variations in TEC and S4 by the release of various solar fluxes [7]. X-class flare due to intense solar activity during 7th- 9th March 2012. Figure 3: Dst-Index during the month of April 2012, showing a sudden change in its phenomenon [8]. Figure 7: Variations in TEC and S4 by the release of M-class flare due to intense solar activity during 10th-11th March 2012 Figure 4: Dst-Index during the month of June 2012, representing a sudden drift in the plot of Dst-index. Figure 8: Variations in TEC and S4 observed for a warning issued due to K-Index (K>=7) during 23rd - 26th April 2012. Figure 5: Dst-Index during the month of July 2012, showing an intense downfall of Dst-index. Reduce Exposure to Reduce Risk 265

Figure 9: Variations in TEC and S4 observed due to During the month of April, due to geomagnetic storm sudden downward drift in Dst-Index during 16th -17th disturbance the TEC peak touched the value of above 75 TEC and very normal scintillations leading to June 2012. S4>0.3 were seen. Those minor scintillations were seen during 6:00 to 10:00 hours UT. The event was Figure 10: Variations in TEC and S4 observed due to regarded as G3 alert by NASA‟s SWPC that means sudden downward drift in Dst- Index during 16th - this event was intense and strong. But as datasets 17th July 2012. seen from AIT station retrieved were not that supportive for the existence of such an intense 4. VARIATIONS IN TEC AND S4 geomagnetic disturbance. Thus it can noticed that the In this section, we present the variations occurred due effect would have been more in the other of the to these extreme events in the GPS-SCINDA derived globe. parameters TEC and S4. Plots have been generated for considered extreme events by using the raw data During the month of June a sudden drift in Dst plot of GPS-SCINDA and by post- processing of RINEX was observed. Apparent to this change, neither such format files by using Rinex GPS-TEC program Version effective TEC value nor intense scintillations were 1.45 software developed and provided as an open observed at AIT station. The TEC value has touched source by the Gopi Seemala of Boston University [9]. a lowest notch of 50TECU and Scintillation of During the ejection of X-class solar flare, the TEC S4>0.2 were observed from the plots generated after value went to its peak of 100 TECU and minor post-processing. And those minor scintillation were disturbance of scintillations leading S4>0.3 were seen during 16:00-18:00 Hours UT. This means that observed during 4:00-10:00 hours UT. This X-class effect of geomagnetic storm was not felt by this solar flare would have left with numerous amount of station of GPS-SCINDA. electron into the space because of which TEC value During the month of July, hype was developed world touched an unusual peak and a sudden and drastic wide over the solar storm erupted from the surface of change has been witnessed from 6:00- 12:00 hours the sun and July 15th event was considered of greater UT. Where a drift of ups and downs can be easily importance due to its expected post effects on our understood. planet Earth. But no harm to satellites and power grid were observed during this whole period. Little During the emissions of M-class solar flare during scintillations were seen during 8:00 – 16:00 hours 10th of March, 2012. An intense turbulence in UT. Which refers that the geomagnetic disturbance scintillations was observed during the 6:00 to 16:00 developed was not much effected in the ionosphere hours UT. A peak of S4>0.4 and TEC value touching above AIT‟s GPS-SCINDA station. 75 TECU were retrieved. Thus, in total we have been faced with two solar flares of X and M class 5. ANALYSIS AND CONCLUSIONS respectively within a short span of 2-3days. As observed from the plots of all extreme events, it is quite evident that two major events occurred during the month of March, 2012, that too in a very short span. During this TEC went to its peak by touching 100TECU. Which is the highest peak ever observed during other events as well. It can be understood that this particular event has very high impact and the corresponding plots totally agrees with GOES-15 satellite data , representing the release of X-class solar flare, which is the most severe one of all the remaining classes namely M and C. Also, for every event there seems to be a kind of non-linearity and intense disturbance in ionosphere which can be well observed in the corresponding plots. Of all the events NASA‟s Space Weather Prediction centre has issued alerts as G3(which means strong and Intense) for the first three events occurred during March and April of 2012. And G2(Moderate) for the remaining two events occurred during June and July of 2012. Also, an uncorrelated relationship has been observed with the GPS-SCINDA derived parameters TEC and S4. It can be analyzed that not every geomagnetic storm has effect over the current station. For e.g. the later 3 events have not seen much or intense changes in the recorded TEC and Scintillation index at AIT station Reduce Exposure to Reduce Risk 266

due to its spatial and temporal differentiations. Most [5]”Study of Equatorial Ionospheric scintillation and of the events could have occurred on the opposite TEC characteristics at Solar minimum using GPS- side of the Globe, proving little scintillation at this SCINDA data” by J.-B. Ackah, O. K. Obrou1, Z. site of observation. Thus to understand in detailed Zaka, M. N. Mene, K. Groves, Sun and Geosphere, effects of increasing solar activities, we need to 2011, pg. No. 23-26. analyze the data from various parts of the world, which can lead us to much better and clearer vision [6]. Website Available at Space Physics Research about the effects of these extreme events on to TEC Group, University of California, Berkeley, and S4 and its impacts on life on Earth. “http://sprg.ssl.berkeley.edu/dst_index/Dst- index.html”. REFERENCES [7]. Website Available at National Geophysical Data [1]. “Return of a GNSS villain: the ionosphere strikes Centre, National Oceanic and Atmospheric again” by Volker Janssen and Simson Mcelroy Administration, August/September 2012 edition of Position, pg40 “http://satdat.ngdc.noaa.gov/sem/goes/data/new_plot s/2012/goes15/summary/” [2]. Article published in the national daily, “The Nation”, in Bangkok on July 15th 2012. [8]. Website Available at World Data Centre, Kyoto, Japan, “http://wdc.kugi.kyotou.ac.jp/index.html”. [3]Website Available at Marshall Space Flight [9]. Website Available at blogspot of Gopi Seemala , “http://seemala.blogspot.com/2010/05/rinex-gps-tec- Centre, Solar Physics page program-version-145.html” ,“http://solarscience.msfc.nasa.gov/predict.shtml”. [4]. Hey, J. S., S. J. Parsons, and J. W. Phillips: Fluctuations in cosmic radiation at radio-frequencies, Nature, 158, 234, 1964. Reduce Exposure to Reduce Risk 267

CLIMATE VARIABILITY IN PING RIVER BASIN, NORTHERN THAILAND Abbadi Girmay Reda1 , Nitin K. Tripathi2, Peeyush Soni3, Taravudh Tipdecho2, Chaya Vaddhanaphuti4 1Tigray Agri Research Institute, Ethiopia and RSGIS, AIT, Thailand , ([email protected]) 2RSGIS, School of Engineering and Technology, AIT ([email protected]) 3 Agricultural systems and Engineering, SERD, AIT ([email protected]) 4 Chiang Mai University, Department of Geography ([email protected]) ABSTRACT Climate variability refers to deviations of shorter term climate statistics (daily, seasonal, annual, inter-annual, several years) from the long-term climate statistics relating to the corresponding calendar period and measured by those deviations, which are usually termed anomalies and variations in climate, including the fluctuations associated with El Niño (dry) or La Niña (wet) events but climate change refers to long-term (decades or longer) trends in climate averages such as the global warming that has been observed over the past century, and long- term changes in variability such as frequency, severity and duration of extreme events. Climate variability has always had a significant impact on food production and currently places significant stresses on food production and availability. The climate of Thailand is under the influence of Monsoon, Inter Tropical Convergence Zone (ITCZ) and Tropical Cyclone. Summer monsoon rains are a critical factor in Thailand‟s water resources and agricultural planning and management. This study was undertaken in Ping River Basin, Northern Thailand which is one of the major watershed in northern Thailand, including parts of 5 provinces (Chang Mai, Lamphun, Tak, Kamphaeng Phet & Nakhon Sawan). Climate data of Ping River Basin (1961-2010) acquired from Thailand Meteorological department (TMD) were analyzed at 3 catchment zones (upper, middle, and lower Ping catchments) and at Ping basin level for all variables for climate variability, trend and anomaly. The main climate parameters analyzed include temperature (5 parameters), rainfall (2 parameters), relative humidity and evaporation. The period 1961-90 was taken as a climatological base year. Monthly, annual, decadal and long term climate data descriptive statistics, graphs and summary were prepared. The Ping climate exhibited spatiotemporal variability. Climate variables showed high inter- annual and low inter- decadal fluctuation, inconsistency and anomalies as indicators of climate variability in the past 50 years period (1961-2010). The selected climatological base year period (1961-90) showed clear evidence of climate variability which is in line with the global facts. Extreme temperature values (Extreme maximum of 44 0 c and minimum of 40 c) might have had negative impacts on agricultural production, environment and health. The year 1999 was a typical year for rise of temperature and is also a global indicator of sharp climate shifts. The first two decades of time period 1961-1979 had higher maximum temperature record for the basin and from 1980 onwards maximum temperature shows a declining trend indicating the evidence that there was observable climate change and variability in the baseline period of 1961-90. Mean temperature was stable during the first 2 decades (1961- 1979) but increased by 10C from 1980-2010. This research output of historical climate database of Ping River Basin (1961-2010) could be input for future research and development plans. KEY WORDS: Climate variability, Ping River Basin, anomaly, trend, climate parameters 1. INTRODUCTION and environment (1,2,12). The term climate Climate is the measure of weather patterns over a variability is often used to denote deviations of long period of time, and inherent to climate are climate statistics over a given period of time (such as changes, both long-term and short-term. Short-term a specific month, season or year) from the long-term climate changes represent periodic or intermittent climate statistics relating to the corresponding changes that occur, and this is termed climate calendar period. In this sense, climate variability is variability. These short-term changes might include measured by those deviations, which are usually floods, drought, temperature changes or oscillating termed anomalies. Climate variability refers to weather patterns such as the effects of El Niño or La shorter term (daily, seasonal, annual, inter-annual, Niña. In the most general sense, climate variability is several years) variations in climate, including the thought of as the deviations in climate statistics over fluctuations associated with El Niño (dry) or La Niña a long period of time. Accurately identifying and (wet) events. Climate change refers to long-term understanding climate variations is important to (decades or longer) trends in climate averages such as recognizing and understanding their effect on humans the global warming that has been observed over the Reduce Exposure to Reduce Risk 268

past century, and long-term changes in variability to define a “current climate baseline”. This set of (e.g. in the frequency, severity and duration of years can be used to calibrate impacts models and to extreme events). The climate system is complex and quantify baseline climate impacts, e.g., crop yields chaotic – we will never have 100% accurate climate under current climate. A 30-year continuous record forecasts. The climate variability impact at regional/ of recent climate data is widely used for creating a sub-regional and ecosystem levels is likely to be baseline climate (e.g., Rosenzweig and Parry, 1994). uneven and unpredictable. South east and south Asia The current standard WMO normal period is 1961- are such vulnerable regions wherein impacts on food 1990. But it is often desirable to compare future security would be considerable.. The vulnerability of impacts with current rather than some past condition. rice-based agro-ecosystems is reflected in terms of A 30-year period is likely to contain wet, dry, warm, varying yields due to increase in the levels of CO2 and cool periods and is therefore considered to be and temperature and fluctuation of rainfall sufficiently long to define a region‟s climate. The 30- (5,6,12,13). Climate variability associated with the year “normal” period as defined by the World ENSO cycle has a range of implications for different Meteorological Organisation (WMO) is socio-economic sectors in Southeast Asia. El Niño– recommended by the Intergovernmental Panel on Southern Oscillation (ENSO) has been widely known Climate Change (IPCC) for use as a baseline period to modulate the Asian summer monsoons on (Carter et al., 1994). Climate variability contributes interannual and interdecadal time scales(2,8,10). significantly to poverty and food insecurity. Thailand hydroclimatology exhibits a strong trend Proactive approaches to managing climate variability and interdecadal variability. The variability in recent within vulnerable rural communities and among decades (post-1980) seems to be strongly linked with institutions operating at community, sub-national, ENSO. There is a clear indication that Tmax, Tmean and and national levels is a crucial step toward achieving Tmin in Thailand tend to be warmer (colder) than the Millenium Development Goals (MDGs). normal during the El Niño (La Niña) phase of the Monsoon rains are critical factor in Thailand‟s water ENSO. From a long-term perspective, the data resources and agricultural planning and management suggest that Tmin in Thailand has been on the rise at (2,9). an unprecedented rate since the early 1950s, and has occurred at faster rate than Tmax.. Similar long-term 2. METHODOLOGY changes in Tmin over the past few decades have also 2.1. Study Area: Ping River Basin, Northern been reported for other regions of the world. The Thailand climate of Thailand is under the influence of The Ping River along with the Nan River, is one of Monsoon, Inter Tropical Convergence Zone (ITCZ) the two main tributaries of Chao Phraya River. With and Tropical Cyclone(4,7,12,13). a catchment area of about 35,000 km2, the Ping River Basin covers about 22 percent of the larger Season In Thailand are: Rainy season or Southwest Chao Phraya river system within which it is nested Monsoon Season (mid-May to mid-October), Winter and contributes about 24 percent of the system‟s or northeast monsoon season (mid-October to mid- average annual runoff. The Ping river basin is the February) and Summer or pre-monsoon season (mid- major watershed in northern Thailand. The Ping river February to mid-May) (7,11,13). Climatological basin is one of the four upper tributary basins baseline period is typical in impacts assessment to forming the Chao Phraya river system, the most use a period of years of observed meteorological data important river basin in Thailand. Table1. Geography of Ping River Basin (Summary) S.N. Data Type Data 1 Political administration Northern Thailand, including parts of 5 provinces (Chang Mai, Lamphun, Tak, Kamphaeng Phet & Nakhon Sawan) 2 Area 35,000 km² 3 Geographic Location  Latitude: 15.719900 to 19.821400  Longitude: 98.071700 to 100.166000 4 Altitudinal zonation  Upper Ping (Chiang Mai and Lamphun) of Ping catchments  Middle Ping (Tak)  Lower Ping ( Kamphaeng Phet and Nakon Sawan) 5 Population 2,384,946 6 Annual average runoff 9,073Mm3 7 Dam (reservoir)-Bhumiphol 9,662Mm3 storage capacity 8 Topography Valley, mountains 9 Rice farming Rainfed and irrigated Reduce Exposure to Reduce Risk 269

Figure 1. Location map of Ping River Basin 2.2. Data Analysis The 1961-90 period was selected as baseline Upper Ping climatological period for Ping River Basin for climate variability, trend and anomaly comparison of Middle Ping 50 years climatological period (1961-2010). Time Lower Ping series climate data of 50 years (1961-2010) from Thai Meteorological Department (TMD) was ready Table2. Selected stations for analysis. Analysis was done at 3 catchment zones (upper, StCode Station Lat Long Altitu Ping middle, and lower Ping catchments) and at Ping basin name de Catch level for all variables. Ping basin level statistics as 18.48 98.99 (masl) ment average of the 3 catchments. Chiang 16.9 99.7 Upper Analysis includes: Descriptive statistics of long 327501 Mai 15.7 100 312 Ping term climate data , decadal, annual and monthly and Middle summary; Climate variability (annual, inter annual 376201 Tak 124 Ping and decadal variability and index);Time series trend Nakhon Lower and Climate anomaly 54 Ping 400201 Sawan Climate variables for analysis: Monthly data of 9 variables of temperature, rainfall, relative humidity and evaporation were screened, checked up and coded Selected climate variables: Temperature (5 variables: Maximum (TMax), Minimum(TMin), Mean (TMn), Extreme Maximum (ExtMax), Extreme minimum(ExtMin); Rainfall (Total rainfall and Rainy days); Relative Humidity (RH-%) and Evaporation (Evap.) Climate data of Extreme Maximum and Extreme minimum Temperature and evaporation for Ping River basin were recorded from 1980 onwards and only 30 years period (1980-2010) was considered for these variables. Input Data (monthly data):  50 years (1961-2010): Temperature (Max, Min, Mean), Rainfall and RH  30 years data (1980-2010): ExtMax, Extmin temperature, and evaporation  Climatological base year (1961-90): Temperature (Max, Min, Mn), Rainfall, and RH 3. RESULTS AND DISCUSSION 3.1. Time series average climate in Ping River Basin Altitudinal variation of climate variables: Upper Ping received the highest rainfall followed by Lower Ping and Middle Ping, respectively. Temperature and evaporation increased with decreasing altitude. Reduce Exposure to Reduce Risk 270

Reduce Exposure to Reduce Risk 271

3.4. Climate trend rise of maximum temperature and this year is also a Maximum Temperature: global indicator of sharp climate shifts. Gradient trend: Decreasing trend with increasing altitude and Minimum Temperature maximum temperature declines when we go up from From 1980 onwards, unlike maximum temperature, lower Ping to upper Ping catchment minimum temperature had an increasing trend showing that seasons are getting hotter in recent Temporal Trend: years. This is also in line with global climate change The first two decades of time period 1961-1979 had and variability trends that some places are getting higher maximum temperature record for the basin hotter in which cold periods of the year are getting and from 1980 onwards maximum temperature hotter. This variability is manifested either by rise of shows a declining trend indicating the evidence that maximum temperatures or increment of minimum there was observable climate change and variability temperatures so that years and seasons become hotter in the baseline period of 1961-90. This finding is in than their past climatic condition. line with global trends justifying that the standard base year of 1961-90 could be taken as climatological Extreme Maximum Temperature base year and local maximum temperature It was highly fluctuating and ranged between 38- phenomenon is observed and global trend validated 440C. Very high records of 43-440C were observed in at local level. The year 1999 was a typical year for recent years (2004-2008). These recent inter annual extreme maximum temperatures would have had Reduce Exposure to Reduce Risk 272

negative implications on agricultural production and Figure5. 2011 monthly rainfall as compared to environment. normal (Source: TMD) Extreme Minimum Temperature 4. CONCLUSION Extreme minimum temperature dropped sharply in The Ping climate exhibited spatiotemporal 1999 recording 40C in the upper Ping catchment variability. Climate variables showed high inter- whereas 60C in middle and lower catchments in the annual and low inter- decadal fluctuation, same year. Years before and after 1999 had similar inconsistency and anomalies as indicators of climate records between 8-160C indicating that the year 1999 variability in the past 50 years period (1961-2010). had extreme low temperature which is in line with The selected climatological base year period (1961- global extreme events of the year 1999. 90) showed clear evidence of climate variability which is in line with the global facts. There was time Rainfall series trend in maximum and minimum temperatures There was high inter-annual variability deviating but no trend in rainfall based on Mann Kendall trend upto 450mm above long term average and upto test.Rainfall followed normal distribution but 300mm below normal average in some years. There maximum and minimum temperatures were not was little decadal anomaly of 19.46mm in 1961-70 normally distributed and skewed. Extreme and -7.410C in the next 2 decades (1971-90). The temperature values (Extreme maximum of 44 0 c and decade 1991-2000 showed decline of rainfall (- minimum of 40 c) might have had negative impacts 75.45mm) while the recent decade (2000-2010) was on agricultural production, environment and health. evidenced by increment of 70.80mm indicating that the 1990s were deficit years and 2000-2010 as The year 1999 was a typical year for rise of increased rainfall. temperature and is also a global indicator of sharp climate shifts. The first two decades of time period Relative Humidity (%) 1961-1980 had higher maximum temperature record Years 1966 and 1977 were anomaly years for high for the basin and from 1980 onwards maximum humidity record in upper and lower Ping catchments, temperature shows a declining trend indicating the respectively. Other years and decades have little evidence that there was observable climate change variation from the normal average and variability in the baseline period of 1961-90. Mean temperature was stable during the first 2 Evaporation (1980-2010) decades (1961-1979) but increased by 10C from Evaporation increases as we go down from upper 1980-2010. There was little inter annual and inter catchment to lower Ping catchment which is directly decadal variability. Inter-annual and inter-decadal proportional to temperature increment when temperature variability was observed in terms of descending from upper to lower Ping which is fluctuation, inconsistency, trend and extreme values attributed to distinct altitudinal variation among the 3 Ping sub zones. 3.5. The 2011 Flood in Thailand In October 2011, the late rainy season, the northeast monsoon prevailed over Thailand since around mid- month. Meanwhile, the monsoon trough moved southward to lie across central Thailand, bringing occasional rain. The monthly rainfall amount has decreased from the previous month especially in central part, at 2% below normal. However, associated with the high tide, severe flood and extensive damage in the lower northern part has extended to the central part including the Chao Phraya River Basin, most areas of northern, eastern and western Bangkok and its vicinity. The rainfall amount of Thailand since 1 January to 31 October was 1822.4 millimeters, about 28 percent above normal and the October rainfall was 201.8 millimeters, 10 percent above normal. Seasonal rainfall from May to October in 2011 was above normal of 20 – 60% for most Meteorological Station in northern part and of 10 - 40% with below normal in some areas in central part (TMD 2012). Reduce Exposure to Reduce Risk 273

and events, and anomalies. In the first 2 decades 10. Pedram Rowhani, David B. Lobel, Marc (1961-80), maximum temperature increased by 1.50C Linderman and Navin Ramankutty. Climate and later decreased by 10C in the later decades (1981- variability and crop production in Tanzania. 2010). Minimum temperature dropped by 2.30C in Agricultural and Forest Meteorology 151 the period 1961-80 and increased by 1.530C in the (2011)449–460. years 1981-2010 as compared to the long term 50 http://dx.doi.org/10.1016/j.agrformet.2010.12.0 years normal temperature of 1961-2010. 02 There was high inter-annual variability of rainfall 11. Regional Climate Outlook: Southeast Asian deviation of upto 450mm above long term average Applications by A. R. Subbiah and Kamal and upto 300mm below normal average in some Kishore, 2000. C.K. Folland, T.R. Karl, et‟al, years. There was little decadal anomaly of 19.46mm (book) Observed Climate Variability and in 1961-70 and -7.410C in the next 2 decades (1971- Change 90). The decade 1991-2000 showed decline of rainfall (-75.45mm) while the recent decade (2000- 12. Singhrattna, Nkrintra, Balaji Rajagopalan, K. 2010) was evidenced by increment of 70.80mm Krishna Kumar, Martyn Clark, 2005: indicating that the 1990s were deficit years and 2000- Interannual and Interdecadal Variability of 2010 as increased rainfall. Thailand Summer Monsoon Season. J. Climate, 18, 1697–1708. REFERENCES 13. Thai Meteorological Department (TMD). 1. ADB, 1994: Climate Change in Asia: Executive Meteorological records (1961-2011). Summary. Asian Development Bank, Manila, Philippines, 122 pp 2. Adger, W. N., and Coauthors, 2001: Asia. Climate Change 2001: Impacts, Adaptation and Vulnerability, J. J. McCarthy et al., Eds., Cambridge University Press, 535–581. 3. Climate Variability and the Millennium Development Goal Hunger Target (JAMES W. HANSEN et‟al, 2004, International Research Institute for Climate Prediction, The Earth Institute of Columbia University) doi: http://dx.doi.org/10.1175/JCLI3364.1 4. Easterling, D. R., and Coauthors, 1997: Maximum and minimum temperature trends for the globe. Science,277, 364–367. 5. Hamada, J-I., M. D. Yamanaka, J. Matsumoto, S. Fukao, P. A. Winarso, and T. Sribimawati, 2002: Spatial and temporal variations of the rainy season over Indonesia and their link to ENSO. J. Meteor. Soc. Japan, 80, 285–310. 6. Indian summer monsoon: An observational study. J. Climate, 12, 3117–3132 7. Interannual and Interdecadal Variability of Thailand Summer Monsoon Season (Singhrattna et al., 2005) 8. Intergovernmental Panel on Climate Change (IPCC): Synthesis Report 2001. Contribution of Working Group I, II, and III to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press; 2001. 9. New, Mark, Mike Hulme, Phil Jones, 1999: Representing Twentieth-Century Space–Time Climate Variability. Part I: Development of a 1961–90 Mean Monthly Terrestrial Climatology. J. Climate, 12, 829–856. Reduce Exposure to Reduce Risk 274

SOCIAL COMPETENCIES RELATED TO CLIMATE CHANGE ADAPTATION AND DISASTER RISK REDUCTION M.N.S. Edirisinghe and Patrizia Bitter National Institute of Education (NIE), Maharagama, Sri Lanka E-mail: [email protected], [email protected] ABSTRACT The study suggests a Learning Management System (LMS) for climate change adaptation and disaster risk reduction (DRR) information as an innovative model for national level capacity building for the school community according to a pilot study conducted in 2011. It is heartening that DRR / disaster safety education (DSE) has come to be recognized as a feature of serious significance within the ambit of general education in Sri Lanka. This is all the more vital in that children, whether school going or otherwise, have proven to be the most vulnerable population in the instance of disasters. Therefore, capacity building of the next generation on disaster safety education and climate change issues through media has a spillover effect too. However, social competency DRR and climate change adaptation does not amount to mere awareness of the situation but to the application of the synthesized information regarding the depth, gravity and negative potentials of the situation to be applied in taking preventive measures to ensure risk reduction to oneself and the community. Placement of LMS linked to Geo information systems can play a major role in this age of globalization in providing information to the local community at least through mobile. Therefore this model (LMS) suggests not only updating of information with Geo information but providing the option for the sharing of indigenous knowledge on disaster risk reduction through school net, computer resource centers (CRCs) available at remote locations. Geo information technology has the potential to support decision making relevant events awareness though LMS will benefit for Sri Lankan population through school community. In Sri Lanka, DDR and climate change adaptation competencies are provided under subject areas of the sciences, social studies in the secondary school curriculum. In the above regard GIZ(Deutsche Gesellschaft fuer Internationale Zusammenarbeit - GIZ), in collaboration with National Institute of Education (NIE), has developed a DSE media kit (DVD and web version) comprising a collection of multimedia resource materials for the purpose of capacity building stakeholders. Teachers as well as teacher trainers were provided specific training on media kit in the use of multimedia as a teaching tool as well as a self learning tool. The instructional design module for learning with multimedia is an introductory step of this trainer training program. This was followed by a session for the provision of hands on experience on retrieving and synthesizing information using the media kit. These trainers conducted eight capacity building workshops for purposive sample of 168 teachers from 87 schools, located in 15 zones in four provinces of Sri Lanka. Teachers so trained were encouraged to draw up a lesson plan relevant to own context. A random sample of 10 schools from the 87 above was identified for classroom observation. Finally, the instructional training module was modified and a new model is design through a needs assessment in terms of trainee perceptions. According to Friedman non parametric test “ “pedagogy relevant to context synthesis” was considered to focus on the most effective aspect of the instructional training module where the p value less than 0.05 was statistically significant at that level. As such, the implication here is that learning event based training was the most effective mode for the development of social competencies on DDR and climate change adaptation issues. Based on the outcome of the needs assessment the instructional module based on social competencies was modified and an innovative LMS model for the implementation of DDR and climate change adaptation issues at national level was identified. Reduce Exposure to Reduce Risk 275

NATURAL DISASTER DUE TO EARTHQUAKES AND THEIR COMPREHENSIVE MANAGEMENT Mrinal K. Ghose Asian Development Bank in Vietnam, 113 Regent Estate, Kolkata -700092 E-mail: [email protected] ABSTRACT The natural disaster due to earthquake is not rare or unusual phenomenon in India. This paper examines the cause of earthquakes, which accounts for more loss of life and property than any other natural phenomena. Rising population and urbanization have been discussed in respect of natural disasters. It has given some statistics of disasters, disaster prevention potential of societies and appropriate disaster prevention standards. It analyses the hazard vulnerability in respect of earthquakes and discusses the usefulness of the Vulnerability Atlas, which is being developed by different organizations in India, for formulating proactive policies to face the threat due to natural hazards. It discusses the different aspects of natural disasters and gives a brief account of the statistics on disasters. Natural disasters in Asia account for about 40% the world total, while the death toll ratio is about 50% and the number of people affected about 90%. In India 54% of land is vulnerable to earthquakes. This paper discusses the advances in building sciences in designing seismic resistant structures to combat earthquakes and hazard vulnerability in India. It focuses on the fact that increasing urbanization and degradation of the natural environment on a global scale are having the effect in increasing the frequency and severity of disasters around the world. It discusses the statistics of disasters, prevention potential of disaster by societies and appropriate disaster prevention standards. It suggests for the designing seismic resistant buildings and structures, and it should be a part of sanctioning building plan Keywords: Hazard, urbanization, prevention, tectonic, vulnerability, awareness Mkg541 has increasingly occurred in the mountainous areas in the country in the world, especially in river basins located in the tropical climate zone influenced by moonsoon and storms. Flash floods bring great damages, not only directly on people and property but also have large consequences in habitat destruction. Therefore the world and Viet Nam have come to understand the importance and urgency of the flood forecasting purposes to reduce the consequences of this disaster caused. In recent years, with information technology’s development in the world, remote sensing (RS) technology and geographic information systems (GIS) have been developed from different scientific disciplines and application on the field. There has been a long-term interest in merging these two technologies, because remote sensing provides relatively accurate data on the status of land use, terrain elements, geology, vegetation cover…then linked to GIS. GIS is not only use to represent and process data but also to draw the object class on geographical maps. This article is an introduction to integrate solutions of RS and GIS to establish flash flood level classification map. KEY WORDS: Flash flood, GIS, RS, level classification GIS DATA APPLICATION WITH MOBILE PHONE REMINDING FOR HIGHLY MUDSLIDE PHENOMENON IN THAILAND: NORTHERN PA-SAK WATERSHED AREA Vansarocahana, A., and Piyathamrongchai, K. ABSTRACT Northern Pa-Sak watershed area is the highly mudslide risky zone of Thailand., as appearance in 10 August 2001, people die more than 150 persons, and over 515 houses were destroyed. With risky hazard phenomenon, this area is selected to receipts daily rainfall telemeter stations for serving rainfall data to predict mudslide hazard. This study represents the application of GIS techniques, related spatial data, and rainfall telemeter data to generate dynamic mudslide risk area, and convey the risks to remind on mobile phone. KEY WORDS: Daily rainfall telemeter stations, Dynamic mudslide area. Reduce Exposure to Reduce Risk 276

Flood Plain Encroachment and Mapping using LiDAR derived High Resolution Digital Elevation Model and HEC-RAS for Adyar River D. Thirumalaivasan1, Dr. M. Ramalingam2 and Dr. Bhoop Singh3 [email protected], [email protected] and [email protected] ABSTRACT Flooding is more frequent in urban areas due to rapid urbanization, encroachment in flood plains and shrinking lakes or water bodies in a typical urban setup. In this study, the landuse/cover of city of Chennai, then known as Madras, was analysed using old historical data and compared with current high resolution data. The 1903 Map of Chennai was mapped and the changes with specific reference to waterbodies present were compared with CORONA Spy Satellite declassified image, Quick Bird High Resolution data. For the purpose of analyzing the impact of encroachment on waterbodeis, a pilot study area, consisting of four consecutive stretches of Adyar River Nandampakkam bridge to Jafferkhanpet bridge (2.8 km), Jafferkhanpet to Saidapet bridge (2.5 km), Saidapet Bridge to Kotturpuram bridge (2.8 km) and Kotturpuram bridge to Adyar mouth,(4.2 km) within Chennai City. Lidar derived Digital Elevation Model with 1m resolution available for the project study area has been used to derive the cross-sectional profiles and HEC-RAS model was used to model the flood levels and the same has been compared with observed flood levels for 2005 floods. Site specific flood mitigation measures and recommendations have been suggested in terms of augmenting the existing storm water drains based on their current capacity and removal of encroachments in the flood plains of Adyar River. MONITORING AND MODIS DATA SYNTHESIS FOR DETECTING SPATIO-TEMPORAL HAZE SMOKE DYNAMICS IN NORTHERN THAILAND Nguyen Luong Bach and Nion Sirimongkalertkal Mae Fah Luang University (MFU), Chiang Rai, Thailand E-mail: [email protected], [email protected] ABSTRACT Haze and smoke problems have become more serious disaster issues over the last few years, especially in Northern Thailand. The main cause is the similar landuse culture of forest fires and open land burnings in the burning season typically from January to April, from both within Thailand and its neighboring countries such as Myanmar and Laos. Although remote-sensed data have been intensively used for fire hotspot detection at the source, there is an equal important need to investigate spatio-temporal dynamics of haze smoke at receptors, so that more suitable prevention and mitigation strategies can be designed accordingly at relevant scales. This study aims first to analyze available PM10 data to detect spatio-temporal dynamic patterns of haze levels and haze spots in Northern Thailand where data are available. Our initial findings reveal that, over the last several dry years from 2007 to 2012, there have been substantial elevations of PM10 levels in the burning season and also significant high deviations from the monthly averages within the burning season. More interestingly, both haze levels and spots in terms of top peaks in monthly PM10 averages during the burning season are detected to have systematically increased and changed in Northern Thailand, starting from the lower part to the upper, then diverting to the west, and finally moving further up north in the two Thai-Burmese provinces of Chiang Rai and Mae Hong Son. On the contrary, haze tendencies have been declining more or less to the medium level in the lower part of Northern Thailand. Further attempt to explain why such dynamic changes can occur is being made by taking into account other related dynamic behaviors of fire hotspots and meteorology. Our preliminary analysis indicates that there is only an overall match of the high level between PM10 values at a site and burning hotspot counts in its vicinities, while large variations of PM10 levels are more related to other factors such as topography, meteorology and long-range burning hotspots. This study continues to reinforce our earlier finding that fire and pollution risk mapping or zoning must be made dynamically both in time and space. Policy Reduce Exposure to Reduce Risk 277

implications for haze disaster reduction are also discussed along with further research needs including additional monitoring and remote-sensed data requirements at various geographical scales. KEYWORDS: Spatio-temporal patterns of PM10, Dynamics of burning hotspots, Temporal dynamics of haze hotspots, Meteorological patterns, Northern Thailand and neighboring countries. INTEGRATED FOREST FIRE MANAGEMENT STRATEGIESTHROUGH PARTICIPATORY SPATIAL PLANNINGFOR SRI LANKA H.M.B.S. Hearath Department of Geography, University of Sri Jayewardhenepura, Sri Lanka E-mail: [email protected] ABSTRACT Forest fires are one of the main reasons for forest degradation in Sri Lanka. It has caused number of environmental and socio-economic damages to the country. The magnitude and the frequency of this problem have shown an increasing trend all over the country along with the climate change driven unexpected weather conditions. Therefore, control of forest fires is important to conserve forest resources for the prosperity of the nation. Forest fires in Sri Lanka are fundamentally due to anthropogenic activities and they are influenced by social economic conditions of the people who live around the forests. Further, vegetation characters of the forest and climatic conditions are also influenced to aggravate this problem.Forest department of Sri Lanka adoptsvarious fire control measures such as establishment of fire lines, deploy of fire guards and create awareness among people during dry season. However the problem still arises frequently during dry season in various places in Sri Lanka. Understanding the spatial aspects of above conditions is important in control of forest fires. Since forest fires are manmade and influenced by spatial aspects, in order to control it participatory spatial planning with various actors is a must. Therefore, an integrated fire control plan should be developed with adjoining communities using their indigenous spatial knowledge and concerning traditional rights of the communities. Participatory GIS is good tool to develop participatory spatial plan for forest fire control in Sri Lanka. This paper describes how Participatory GIS can be used in this context. Therefore, the objectives of this research study are to identify spatial aspects of the forest fire problem , to identify the local knowledge and its significance in fire hazard management and to present a set of integrated spatial model consist with strategies’ to manage fire hazard in Sri Lanka. A literature survey has been carried out to identify appropriate tools to analyze the spatial aspects of the forest fire problem. The application of participatory GIS to fire hazard management was evaluated extensively to present an integrated spatial planning framework consist with strategies for forest fire management in Sri Lanka. Reduce Exposure to Reduce Risk 278

AN IDENTIFICATION OF SPATIAL AND TEMPORAL PATTERNS OF FOREST FIRES USING GIS TECHNIQUES :SPECIAL REFERENCE TO BADULLA AND NUWARAELIYA DISTRICTS R.L. Ratnayake, K.A.S.S. Wijesekara, and S.P.D.R. Senarathna Department of Geography, University of Kelaniya, Sri Lanka E-mail: [email protected], [email protected], [email protected] ABSTRACT Identification spatial information of forest fires is needed to better understand and manage emerging environmental problems and those are very useful for planning purposes. Increasing forest fires are likely to have a harmful impact on the vegetation of catchment areaswhich may cause soil erosion and landslides. Sri Lanka has witnessed number of forest fire hazards during last few years, the worst drought that prevailed in 2012 in the central highlands. It causes human activities and enhance rainless period during the south west monsoon. Disaster Management Center in Sri Lanka is reportedapproximately 1000 acres of forest rich with indigenous fauna and flora have been destroyed due to forest fires. The aim of this study is identify and analyzespatial and temporal changes of forest fire in Sri Lanka using GIS techniques. Primary data for the research were gathered through field observations, questionnaires and interviews. Secondary data were collected from data sources fromdifferent institutions and agencies. The information gathered was simplified, classified and summarized and the composition and analyzing of the possessed data is displayed different visualizing methods. Geospatial techniques are proving to be powerful tools to assess the forest fire risk. The principle component analysis was used to sort out the relationships between forest fire potentials and environmental factors. The classifications of these factors were performed with GIS, generating three maps: a topography- based fire risk map, a climatic -based fire risk map and an anthropogenic-factor of fire risk map. These three maps were then synthesized to generate the final fire risk map. The linear regression method was used to analyze the relationship between an area-weighted value of forest fire risks and the frequency of historical forest fires at each forest fires areas. This study revealed that human influence on the forest areas has been growing over recent decades, mainly as a consequence of human-induced fires. KEY WORDS: Forest Fires, Spatial and temporal changes, Geospatial techniques, Fire risk maps INTELLIGENT ROUTE SELECTION SYSTEM IN CASE OF DISASTER FOR MOBILE USER Faryal Safdar Institute of Geographical Information System-NUST, Islamabad, Pakistan, [email protected] Abstract In the past few years, the occurrence of natural disasters has greatly affected Pakistan. Location based services can play an important role in disaster mitigation and management. All the current solutions are for the savers/rescue teams to get to the spot as quickly as possible. But what if a user wants an intelligent route planning in case of emergencies? For example in 2005 when earthquake hit the northern areas, most communication roads were blocked. People did not have any information of open roads or road with good condition or safe roads. For someone to reach home safely he needs a safe path. In response to this situation, a new solution needs to be worked out and framework for the disaster should be built. The simplest solution is a system/prototype for the mobile devices i.e. cellular phone which provides user smart routes with the help of real time information and inform users by generating disaster alerts before time. In this paper a conceptual model of an intelligent route selection system is presented, which facilitates mobile user in case of emergency. The paper explains the components of Location based service, real information needed in planning efficient Reduce Exposure to Reduce Risk 279

routes and generating alerts. There are two main components of the purposed system: first is text based information for the emergency and disaster alerts, second is map based for routing. The result of this study will offer a prototype which will enable regular cell phone users to get beforehand information of disasters and natural calamities and also guide them to the safe place/home in such situations. KEY WORDS: Disaster mitigation and management, LBS, Routing ECOLOGICAL PROTECTION WHEN DEVELOPING OIL ON A SHELF BY MEANS OF GIS TECHNOLOGIES Blinovskaia Iana Maritime State University named after Admiral G. I. Nevelskoy, e-mail: [email protected] ABSTRACT Coastal areas are among the most developed places. The most prominent issues are oil pollution. We developed GIS system for accidents prevention and taking response measures in oil spill risk areas. Such system considers the environment sensitivity to oil pollution, which allows managing the situation and making efficient decisions. The informational-analytic system core is the coastal sensitivity maps to oil pollution. The sensitivity indices make it possible to choose the technology for response measures in risk areas. For choosing the most appropriate technology we suggest to use the matrix as it allows selecting combination of optimal response measures based on local conditions. This is how the process of operational modeling works. It is not possible to predict precisely the place, time, scale of oil spills. We could use different scenarios that we put into the system. Usually we consider the typical and worst conditions in main seasons. To sum up our GIS mapping system based on the coastal areas sensitivity to oil pollution is used for planning, obtaining reference data, making predictions and recommendations. The available algorithms allow setting priorities and form complex environmental safety system. KEY WORDS: Geoinformation technology, oil pollution, oil accident prevention, sensitivity map DISASTER MAPS TO SUPPORT WEST LOMBOK'S TOURISM Mone Iye Cornelia Marschiavelli, Lalitya Narieswari, Sri Lestari National Coordinating Agency for Surveys and Mapping, [email protected] ABSTRACT West Lombok Regency, which is located in West Nusa Tenggara, becomes a famous tourism destination besides Bali in Indonesia, with its beautiful beach and its unique heritage and culture. In addition to its presence as a tourist area, West Lombok also classified as a disaster-prone areas, such as tsunami, floods and landslides. Therefore, disaster risk management approach within the framework of tourism is needed to support Disaster Risk Reduction in West Lombok. This study is aimed to conduct a study as well as mapping the disaster risk in West Lombok, particularly in Batulayar Sub-district. The method used in this research is through Focus Group Discussion (FGD), primary and secondary data collection and GIS analysis to produce thematic maps. Furthermore, the results of this study is expected can be used as an input to the government and the people engaged in the tourism sector in the context of disaster risk reduction in West Lombok.. KEY WORDS: West Lombok, Risk Map, Tourism, GIS analysis Reduce Exposure to Reduce Risk 280

GIS BASED VILLAGE MONITORING AND EVALUATION SYSTEM–A CASE STUDY OF SURAJPURA VILLAGE Padmasree Merakanapalli *, Vemu Sreenivasulu *, V Madhava Rao# * Jawaharlal Nehru Technological University Kakinada, Kakinada, India-533003. # Centre on Geo-informatics Application in Rural Development (C-GARD), National Institute of Rural Development, Hyderabad, Email: [email protected] ABSTRACT The decision makers along with planners, policy makers and administrators feel helpless while planning the natural and social wealth of a region, in the absence of accurate information about all kinds of resources at village level. On the other hand planning is now widely accepted as a way to handle complex problems of resource allocation and decision making. For this purpose both spatial and non-spatial data is required. A suitable information system is required at village level to serve all these requirements. Such information system can provide a more effective and meaningful direction to the planning and development of rural settlements. A village Surajpura has been choosen for this study. In this study GIS based Village Monitoring and Evaluation System is developed for SURAJPURA VILLAGE. The importance of this system is not only integrating the data at household and cadastral level, but also monitoring spatial and non-spatial changes temporally and the showcasing the possibility of evaluating some Rural Development Programs like MGNREGS and INTEGRATED WATERSHED programs. The Thematic Layers are generated using ARCGIS and then converted into shape files which are used as an input to the developed Customized Application using Visual Basic environment and a Map Objects component. The operational advantage of such a system design is that there is no specialized commercial GIS software is required. The developed software is a customized initiative, which runs as an executable mode and distributable. All the data formats required at village level are incorporated in this system to have greater use and decision making and tremendous capabilities of graphical analysis directly from attribute data has been an inbuilt feature of this system. KEY WORDS: Village Information System, GIS, Visual Basic, Map Objects. ASSET MAPPING AND CONSUMER INDEXING USING GEO-INFORMATION TECHNIQUES – A CASE STUDY OF BELLAMPALLI, INDIA Pandruvada Sreekanthi*, Vemu Sreenivasulu*, S Devi Prasad# * Jawaharlal Nehru Technological University Kakinada, Kakinada, India-533003 # Bharathi Engineering Spatial Technology, Hyderabad, India, E-mail: [email protected] ABSTRACT GIS ties together all the pieces of the electric distribution system for improved customer service, better management of assets and outages, and increased accuracy of data. GIS technology is being widely used as the platform for network operations and asset management. Establishment of reliable and automated systems for sustained collection of accurate base line data, and the adoption of Information Technology in the areas of energy accounting will be essential before taking up the regular distribution strengthening projects. This is achieved through GIS mapping to pre-defined scale, generation of intelligence electrical network maps and super imposing them on the land base GIS maps. A town Bellampalli in the Adilabad district, in the state of Andhra Pradesh, India has been chosen for this study. Various thematic layers such as LT and HT lines, LT and HT poles, Distribution structures, LT and HT cable joints of Bellampalli Town are generated using geo- information techniques. The present work gives the better understanding of the Electrical Utility Networking and usage of GIS tool for improvement of electrical power distribution and quality service to the consumers. KEY WORDS: DGPS, Electrical Utility Networking, consumer indexing, GIS. Reduce Exposure to Reduce Risk 281

DETERMINATION OF SUITABLE SITES FOR MASONRY CHECK DAMS USING SDSS AT THE BASIN SCALE: CASE STUDY: DARE BID, TABAS, IRAN Soraya Ardakany,hamed1, Ahmadi,hasan2, Farahpour,mehdi3, Jamali, Ali Akbar4 1-MSc of. Watershed Mansgement. Islamic Azad University, Research and Science Branch, Tehran, Iran 2Faculty of Natural Resource, University of Tehran, Iran 3 Research Institute for Forests & Rangeland (RIFR), Tehran, Iran 4 Watershed Management Dept., Islamic Azad University - Maybod Branch, Iran Abstract: Masonry dams are walls or any other barriers that could be constructed in any shape and dimension using stone with cement or other gluing materials. Study area is located at Dare Bid basin geographically near Tabas town north of Yazd province. It is 117.46 km2. Masonry dams in this area mostly is built to stabilize the canal, increasing water concentration period, controlling water velocity and prevention of sediment movements to the storage dam. Variation in criteria, heterogeneity, and needs for simultaneous survey on criteria and their changes are good reasons for using SDSS (Spatial Decision Support System ) and SMCE(spatial multi criteria evaluation) which are imbedded in ILWIS software. This could work as an efficient tool for watershed management and erosion control. Raster maps of slope, geomorphology and geology as spatial limitations and vegetation density, vegetation type, erosion, and distance to roads, to village,to well,to spring and residential areas as spatial sources were introduced to the conceptual tree model. The model was run in the SMCE in ILWIS 3,3. Areas with certain limitations such as stone faces were omitted using 0 or 1 role. Standardization of criteria was also done to make no-unit criteria. This means all criteria would have the value between 0 and 1 and could be compared more easily. For weighting process direct weighting method using empirical assessments were used. Output was the CIM (Composite index map) map, a combination of all criteria maps. On the CIM suitable areas for Masonry dams were shown. Output of the model was compared with controls, i.e. suitable masonry dams constructed in the real wold. Results showed that areas selected by the model are accurate enough to be used for further works. KEY WORDS: suitable sites, Masonry check dams, SDSS LANDSLIDE SUSCEPTIBILITY MAP USING BIVARIATE STATISTICAL ANALYSIS, A CASE STUDY IN BOGOTA A. H. Souri1, A.Abedini2 and Soran Parang3 Department of Surveying and Geomatic Engineering, College of Engineering, University of Tehran, Iran (Souri_rs,aabedini,,soran_parang) @ut.ac.ir ABSTRACT Landslides occur in all of regions of the world due to preparatory, triggering, sustaining and precondition (predisposing) factors. Researching for landslide hazard and risk has become a critical research for the international community over the last decade. In this paper we utilized bivariate statistical method to create landslide susceptibility map based on quantitatively defined weight values. The main and relevant factors for producing susceptibility map are slope, land use and geology maps. After assigning specific weights for each class of maps we combined all weight maps into a single map using certain combination rule KEYWORDS: Landslide, Bivariate statistical, Landslide susceptibility, hazard Reduce Exposure to Reduce Risk 282

MAXIMUM ENTROPY ANALYSIS FOR LANDSLIDE SUSCEPTIBILITY MAPPING IN CHITRAL, PAKISTAN Rashid Saleem1, Sawaid Abbas2, Faisal Mueen Qamer3, Khuram Shehzad4, and Hassan Ali4 1National Engineering Services Pakistan (NESPAK) Pakistan, [email protected] 2Asian Institute of Technology, RS&GIS, Thailand, [email protected] 3International Centre for Integrated Mountain Development (ICIMOD), Nepal, [email protected] 4 World Wide Fund for Nature (WWF) – Pakistan, [email protected], [email protected] ABSTRACT Landslides occurrence is one of the most common natural hazards in the mountainous regions and can consequence in enormous damage to both property and life. Hence, identification of landslide-prone areas is necessary for safer strategic planning of future developmental activities. To assess the threat triggered by the landslides, it requires the preparation of landslide susceptibility map. This study aimed at mapping the likelihood of landslide susceptibility sites in the disaster prone areas of District Chitral. For this purpose, the statistical method like maximum entropy analysis (MaxEnt) and weighting factor methods have been used with GIS to prepare a landslide susceptibility map in two union councils (Koh and Baroz) of district Chitral. Spatial layers used for the analysis comprises of past landslide incidents inventory, land cover / land use map, slope, aspect, elevation, distance to stream density, and distance to steep stream. The stability index map produced by the model was further categorized into three susceptibility classes. The results were verified with the existing landslide ground based observed points and the comparison of the results divulged that the MaxEnt models performed better than conventional methods of weighting factors. . KEY WORDS: Landslide susceptibility, Natural Hazards, MaxEnt, GIS. EVALUATION OF LANDSLIDE HAZARD POTENTIAL OF NORTH TEHRAN DISTRICTS Ali Uromeihy and Mahsa Sharif Dept. of Eng. Geology, Tarbiat Modares University, E-mail: [email protected] ABSTRACT Tehran is the country’s most densely populated district which is located on the southern part of Alborz Mountains. Many sensitive infrastructures such as dams, roads, power lines and housing complexes are located within the area. In this paper the potential of landslide hazard for North Tehran is evaluated and a hazard zonation map based on ARC-GIS method is prepared. Several parameters such as lithology, slope angle, slope direction, distance from faults and seismicity (earthquake deduced acceleration) are considered as main factors. The quality parameters of the defined effecting factors are quantified by the above named software where the data layers are divided into smaller classes. In the next stage, specific weights are assigned to each class and the hazard potential values are determined. Finally they are presented as a landslide hazard zonation map of the area. It was found that slope angle of 16-35 degrees, slope orientation to the west, lithology of sedimentary rocks, distance of fewer 2 km from fault and earthquake acceleration of over 0.5g has great effect on the distribution of landslide in the area. KEY WORDS: landslide evaluation, landslide hazard zonation mapping, Reduce Exposure to Reduce Risk 283

IDENTIFICATION AND HAZARD ASSESSMENT OF POTENTIALLY DANGEROUS GLACIAL LAKES IN HIMACHAL PRADESH Rajeshwar Singh Banshtu and Chander Prakash Civil Engineering Department, NIT Hamirpur, Himachal Pradesh, India, [email protected] Civil Engineering Department, NIT Hamirpur, Himachal Pradesh, India, [email protected] ABSTRACT Mountain glaciers interact sensitively with climate and therefore they are considered as climate indicators. The climate change of the 20th century has had a pronounced effect on glacier environments of the Himalayas. Warmer climates of the past 100 to 150 years have resulted in widespread glacial retreat and the formation of glacial lakes in many mountain ranges.The formation of moraine dammed glacial lakes at the snout of the glacier and outburst floods from such lakes are a major concern in countries such as Bhutan, Tibet (China), India, Nepal and Pakistan. These glacial lake outburst floods (GLOF,) can cause extremely high water discharges as well as large mudflow events. Triggering events for an outburst can be moraine failures induced by an earthquake, by the degradation of permafrost and increased water pressure, or falling of a rock, snow, or ice avalanche into the lake causing a flood wave with a subsequent outburst. The instantaneous discharge of water from such lakes can cause flash floods, enough to create enormous damage in the downstream areas. The hazardous lakes, however, are situated in remote areas and are very difficult to monitor through ground surveys due to the rugged terrain and extreme climatic conditions. Therefore, remote sensing data and GIS are ideal tools for studying and monitoring glacial lakes and assessing their hazard potential. GIS is capable of integrating and aggregating the data acquired from different sources i.e. topographic maps, satellite data, published reports etc. Glacial lakes are identified and mapped from the satellite data using image processing tools. The glacial lakes and surrounding characteristics such as slope, geology, geomorphology, etc are used to identify the potentially dangerous glacial lakes. A comprehensive approach by coupling of remote sensing, geomorphometric analyses aided with GIS modelling for the identification of potentially dangerous and hazard assessment is used for the present study of glacial lakes in Himachal Pradesh. KEY WORDS: Glacial lakes, GLOF, Remote sensing, GIS modelling, Image Processing, Hazard assessment USE OF REMOTE SENSING AND GIS TO DEVELPOE A LANDSLIDE VULNERABLE INDEX FOR BERAGALA-KOSLANDA AREA IN SRI LANKA W.A.K.NAYANA PADMANI*, H.M.K.G.G.BANDARA** Planning Division, Road Development Authority, Sri Lanka. E-mail; [email protected], [email protected] ABSTRACT Sri Lanka, a tropical country, which receive higher intensity of rainfall all over the year especially from the two monsoon periods and two inter monsoon periods. According to the geological structure of the country, rainfall pattern has been directly influenced the occurrence of landslide that is one of most recorded natural disaster in Sri Lanka. Landslide based on seven inherent physical factors such as distribution of historical landslide, combined soil cohesion, slope steepness, hydrology, land use, human intervention (Road Buffer) and type of bed rock and its structure. Landslide mainly impacts on human and their properties. It is obvious that impacts of landslide could be lead to destroying the lives, damage houses; destroy the cultivation, irrigation system and road network in the country. The results of a RS and GIS based approach to identify the landslide vulnerable index are presented by this paper. Six layers such as Hydrology & Drainage, Bedrock Geology, Soil Cover, Slope range & category, Land use and past landslide deposits were converted into grid data layers as first part of Reduce Exposure to Reduce Risk 284

the methodology. Downloaded DEM have been used to prepare a slope range and category layer and land use map prepared by 1:10,000 scale topographic maps. As a second part of the methodology, the features of those layers were ranked and weighted by expert’s knowledge of landslides. The final result of this study shows that the landslides vulnerable index of the area in four categories such as High Vulnerable area, Moderate vulnerable area, Low vulnerable area and Vulnerable free area. The analysis results were compared with the data collected from the field by using GPS for verification. The index area map can be used for any activity such as road constriction. KEY WORDS: Landslides, Vulnerable Index, GIS, Remote Sensing, GPS LANDSLIDE ZONATION MAPPING OF VULNERABILITY AND RISK ANALYSIS IN ELAPHATHA AREA USING STABILITY INDEX MAPPING (SINMAP) TOOL K.V.D. Edirisooriya 1 N. Saranga Vithanage 2 K.Deheragoda2 1Faculty of Social sciences and Languages, Sabaragamuwa University of Sri Lanka 2Faculty of Graduate studies, University of Sri Jayewardenepura2 [email protected], [email protected], [email protected] ABSTRACT Natural disasters bring about extreme human suffering and cause extensive physical and economic damage. The most common definition for the landslide is the movement of a mass of rock, debris or earth down the slope. The occurrence of landslides is the consequence of a complex field of which is active on a mass of rock or soil on the slope. In Sri Lanka, there are many studies and publications are on landslides. They mainly discussed with most physical characters, types and its impacts of landslides.Elaphatha experience multi hazard problem which, has resulted in significant changes to physical environment as well as social and cultural setup of the community. The deterministic slope stability model used in this study to generate the landslide hazard zonation map is called Stability Index Mapping (SINMAP). The model is a slope stability predictive tool, within which is a hydrologic flow modelling component. It uses the surface topography to route flow down slope, assuming that the surface hydrologic boundary parallels the surface, and soil thickness and hydraulic conductivity are uniform. The model requires three groups of input data. Terrain topography in a DEM grid format; soil mechanical and hydraulic properties in a grid or polygon vector format; landslides source areas inventory in a point vector format. Four stability classes were generated through the model as high, moderate, low and safe areas. Because of the variety of terrain types break out the few calibration regions static have calculate for each of the stability classes within each of the calibration region. The results of this study show that 87% of the study area has been identified as stable by the SINMAP model and about 6 % of the study area has been identified as unstable. 7 % of the study area shows marginal instability as this area can become unstable due to minor destabilizing forces. KEY WORDS: SIN Map (Stability Index Mapping), Digital Elevation, Vulnerability, Hazard, Landslide Zonation Mapping Reduce Exposure to Reduce Risk 285

GEOMORPHIC AND HAZARD RISK ASSESMENT FOR SITE SELECTION OF MILITARY CAMP ON VULNERABLE TERRAINSUSING GIS Muhammad Mateen Mahmood E-mail: [email protected] ABSTARCT Alpine hazards such as landslides, floods and avalanches cause a severe threat to the establishment of military camps constructed in vulnerable terrains comprising of combat zone. The evaluation of the susceptibility of the already built locations and the ones to be chosen for construction to these hazards is a topic that is growing in importance due to climate change impacts. This study presents a review of existing methods for vulnerability assessment related to mountain hazards using GIS. Information on topography, hydrology, precipitation, geomorphic processes, bedrock, structural geology, soils, vegetation and land use, can be processed using various forms of spatial data for hazard assessment as earth observation products and GIS have become an integrated approach in disaster risk management. Digitization of geological maps from primary source, compilation of additional data layers from observations, digital models of slope activities, combining rasterized geomorphologic polygons with digital elevation models, aerial photo interpretation and disaster record map of a given area is paramount in predicting high potential for additional hazards in an area. KEY WORDS: Vulnerability – Landslide – Avalanches – Flood – Bedrock INTRODUCTION GIS SYSTEM Information about the landscape characteristics is Geographical Information Systems (GIS) involves important in estimations about changes of the the integrated acquisition, modeling, analysis, landscape due to changing environmental conditions. presentation and management of spatially referenced The key to understand this development lies in data (i.e. any type of data that includes its location on understanding the past. By studying the landscape in earth), to support decision making. GIS is largely a scientific manner we can start building up required, in particular for spatial analysis, geo- knowledge about our surroundings that will help to morphological legend implementation and integration understand past developments and maybe also predict of non-geo-morphological data. It allows several the future. Since the mid-1990s, studies have been levels of information to be, and provides a increasingly dependent on satellite images, maps and methodology of sorting in meta-data.Non- three dimensional terrain visualizations to assess the cartographic data (e.g. flood events, forest quality cause of natural and human-induced disasters around and potential wood floating, geophysical profiles, the globe. In [1], Nicoll assesses the potential effects etc.) can also be included in the database when using of landslide hazard in the Traverse Mountains, where GIS in order to serve as a tool for further analysis. developmentsof “master-planned communities” have With GIS, a range of solutions can be developed for been permitted onlandslide deposits since 2001. The the management of survey and mapping information study contains vulnerability assessment due to covering cadastral data, topographic and mapping landslide at medium range altitudes with domestic data, aerial photographic images and geodetic establishments. These assessments become all the information within location; the end product of the more critical when the areas to be studied are at even GIS can resolve the issues of data integration and higher altitudes with vulnerable terrains, where more delivery from multiple heterogeneous data sources. critical establishments have to be made. In political conflict stricken areas, military troops are posted for SITE SELECTION REQUIREMENT protection and reaction where military camps are There are two aspects of site selection for a military constructed in places of high altitude with vulnerable camp. One is related to safety measures against terrains. Vulnerable terrains have natural conditions enemy’s attack and second is the safety assessment of which make it susceptible to disasters and its the site itself against natural hazards.The aim here is prevention is a must. A number of variables must be to provide GIS ways to address the later. The idea is considered for site selection of such military to do the Geomorphic assessment of the site and establishments. In this paper, this physical further extending the analysis for natural vulnerability has been investigated without taking hazardsestimation. Then, based on both assessment into consideration the size of the setupinstead the and estimation, a hazard index based on the level of focus is primarily on the natural terrain assessment danger ranking of each site can be formed. This and particularly on the impact of natural hazards on the built environment. Reduce Exposure to Reduce Risk 286

hazard index can help in the selection of a minimum photographs taken of the same area in different years hazard site. to track changes in terrain, stream patterns and land- use over time. The new way to achieve the extraction GEOMORPHIC ASSESSMENT of information from a set of aerial photograph called Geomorphology refers to forms of earth’s surface “digital stereo image interpretation”, which consist in and processes creating and reshaping them.It the creation of stereoscopic images using computer incorporates parts of many different scientific genres systems, and special software. Also special designed (i.e. geophysics, sedimentology, geochemistry, glasses to be able to see the real 3D topographic hydrology, climatology, pedology, biology, and aspect of the terrain on the computer screen. engineering) and binds them together in their (Structural Geology) - All geological structures being common effect on our environment. in the scope of structural geology have a geospatial Morphometric indices are powerful tools for location. The basic feature of them is their position investigating effects of active tectonics and climate with reference to the Earth's coordinate system. A change on mountainous landscapes such as places simplest method of their location is to mark them on under consideration for site selection of military a map, considering their position according to camp. Geomorphological maps have been used for regional or global coordinates. Using GIS, a set of the description of the landscape, both in scientific data being characteristic for an individual type and practical purposes. Due to the amounts of data ofstructures is easy to arrange in the tabular form, required to give a full description of the landscape, where each row is matched to one location of the such mapping systems often result in a complex structure, and in columns is listed all information legend and map sheets hard to read and analyze. describing the given structure in the qualitative and However, with GIS, the complexity of how to present thequantitative way.Usefulness of GIS organized a detailed general description of the landscape databases discloses particularly at the stage of their development can be reduced, since the data handling processing, especially when the given data could be capacity in a GIS is the tool solving previous visualized on the thematic layer. This layer can be problems with map sheets presenting too much data. compared with other layers and then a number of A GIS-based geomorphological map of the particular combinations are unlimited. Possibilities of databases vulnerable terrain under consideration also gives the (GIS) usage reveal also in the case of geological geomorphologist the opportunity to describe and modeling. Because in the database (GIS) the analyze a more complex landscape of site that is unlimited amount of data can be recorded, and they under consideration than by using traditional could also be applied for a geological modeling. geomorphological maps of the same area. This is Geostatistical methods are especially useful. made possible by the use of several layers of Kinematic slope stability analysis can be used to geomorphological data in combination with each evaluate the potential for planeslope failure, which is other or in combination with other spatial digital data sliding along a bedrock discontinuity plane such as such as data on vegetation or precipitation. bedding and wedge failure, which is sliding along an intersection line of two intersecting discontinuity HAZARD RISK ASSESSMENT planes such as bedding and a fault. The geometric GIS has been used as a tool to address specific relationship between the orientation of the aspects of risk management. There are obvious discontinuity planes and the orientation of the advantages in developing a fusion between a overlying topography determine the kinematic philosophy of risk management and a power of GIS stability of a slope that is considered to be vulnerable. as a decision support tool. Historical data coupling We can apply a GIS-based kinematic model to with new datasets aids to dohazard risk assessment in address slope stability for the particular vulnerable all its types as explained in the following discussion. terrain. Geologic data is compiled into the GIS to (Aerial Photo Interpretation)- Aerial photograph create a representative model of the geologic interpretation is an accurate and economical method structure of the site terrain. Spatial analysis of the of assessing terrain conditions and natural hazards geologic model and the topographic model determine affecting land and environmental conditions. kinematic slope stability and identify areas of Completed in advance of satellite and aircraft-based potential plane failure or wedge failure. There is reconnaissance, it provides a comprehensive much feasibility and usefulness of performing overview of the prospective vulnerable terrain site kinematic slope stability analysis within the GIS under analysis, the data for which cannot be obtained framework. GIS-based kinematic analysis is superior from ground level. Aerial photograph interpretation to traditional application of kinematic analysis as GIS helps construct and confirm preliminary hazard based analysis provides: (1) greater accuracy of inventories, understand hazard mechanisms, and results, (2) increased efficiency of analysis, and (3) estimate hazard volume and activity.Time series better communication of findings. GIS can also be photo interpretation uses several sets of aerial used to define and delineate geological structures from satellite images. (Hydrology - Flood Prediction) GIS can be used to estimate rainfall accumulations Reduce Exposure to Reduce Risk 287

over a small watershed. GIS concepts involves (1) hydrological hazards in the canton. The objective of Watershed delineation. (2) Rainfall Maps. (3) Fast using GIS for hazard risk assessment is to aid response basin hydrology. Flood prediction involves decision making and problem solving in fields that (1) Mapping rainfall into basin. (2) Rainfall have a bearing on community safety and intensities in space and time. (3) Rainfall extent sustainability. It kind of provides the “analytical” versus basin size. Because water in its occurrence push which drives the natural hazard risk assessment varies spatially and temporally throughout process. It also provides a more potent form of risk the hydrologic cycle, its study using GIS is especially communication through its capacity to provide a practical. Although GIS systems previously were visual representation of risk situations. mostly static in their geospatial representation of hydrologic features before but now, GIS platforms HAZARD MAP have become increasingly dynamic, narrowing the Hazard map is a map that highlights areas that are gap between historical data and current hydrologic affected or vulnerable of a particular hazard. Hazard reality. Considering the vulnerable terrain that is to map provide guidance to help those in authority to be analyzed for site selection, the elementary water address the affects of natural hazards. Hazard map is cycle has inputs equal to outputs plus or minus essential to understanding and addressing risks that change in storage. Hydrologists make use of a can interfere with the task in hand. Its production is hydrologic budget when they study a watershed. A three step processes which are: watershed is a spatial area, and the occurrence of water throughout its space varies by time. In the (1) A phenomena map is produced by using hydrologic budget are inputs such as precipitation, field geomorphological evidence. A color surface flows in, and groundwater flows in. Outputs scheme is given to each morphogenic are evapotranspiration, infiltration, surface runoff, process. Different tones of colors and and surface/groundwater flows out. All of these thickness of symbols allow the quantities, including storage, can be measured or differentiation of the substratum type (rock estimated, and their characteristics can be graphically or deposit formation), the depth of displayedand studied in GIS. (Bedrock Mapping) gravitational processes, and the evidence, Bedrock Geology of an area that is considered for relative age and size of processes. site can be mapped using GIS by determining the distribution, interrelationships, mineral composition, (2) Then base on previous evidence, intensity mode of formation and age (relative and absolute) of maps are produced either by numerical the different rocks, and by identifying any structures modeling and/or expert system mapping. present. Various technical properties of the bedrock of the terrain are also analyzed to assess the (3) The last step, called Hazard map is a much suitability of the different rock types and production more synthetic map, which shows the of Bedrock Quality Maps. (Avalanche Assessment) different degrees of danger and is based on Precise mapping of the area under consideration can two parameters, intensity and probability of be done in order to address the snow avalanche risk hazard. management, surveying and monitoring. Using GIS; slope, aspect, altitude; landforms and roughness Phenomological map must be remade after each new resulted from a high resolute numerical terrain event and all the maps should finally be superposed model, are considered.This GIS framework helps in to have a global view of the flooded area. Avalanche Assessment of the vulnerable terrain. TheConcept of Hazard Map can be further extended (Debris Flow) Landforms related to debris flows over by using it for indexing. Index will be made for all the terrain are very active and may change very the possible sites under consideration for military quickly over a short time and space scale, for this setup. Indexed map will be produced by collection of purpose Debris Flow Mapping can be done using ranks of many degrees of danger from all produced GIS. Such mapping is the integration of past events / hazard maps. This index map will show the complete history of the stream, distinction between punctual picture as it will compare the indices with the and potential sediment alimentation of a debris flow, vulnerability of all the terrains whose hazard maps distinction of the different processes in the deposition are produced taking care of changes over time. The zone or distinction of different deposition landforms. terrain with the optimum index can be chosen as our With this digital form of information using GIS, site for military camp as that will be the least risk more detailed information about volumes of prone. potentially mobilized sediments, especially in densely populated mountain regions with a high CONCLUSION potential of natural hazards. A better cartographic Survey records existed only in paper and films thus recognition of the slope system and especially the were not well-managed. Government projects, sediment transfer processes, linked with a specific national developments and researches using and legend, could improve the knowledge on investigating historical survey data were costly, time consuming, and staff dependent. To better manage and allow more efficient access to survey related data for both staff and customers, it is the need of the hour Reduce Exposure to Reduce Risk 288

to totally accept the task of compiling historic and M. Papathoma-Kohle, M. Kappes, M. Keiler and T. notable survey-related information into digital Glade, Physical Vulnerability Assessment for format. The formatted data can then be easily Alpine Hazards:State of the art and future needs. available across the organization by linking them to a new database or digital file storage. With these United Nations Disaster Management Training interfaces, time to explore historic data can be Program 1994, Vulnerability and Risk reduced relative to manually searching the paper Assessment, 2nd Edition records. GIS uses these digital data files for study and analysis. Although it can be applied on many Zainal A. Majeed and David Parker practical problems, but the task chosen in this paper Geographic Information System (GIS) for Managing is regarding site selection. It has been presented that if used in effectively, GIS in its totality provides a Survey Data to the Development of GIS-Ready comprehensive solution for selection of site at high Information altitude vulnerable terrains. The idea of Geomorphic Tim Mote, Derek Morley, Timothy Keuscher and assessment, Hazard risk assessment and Hazard maps Todd Crampton, GIS-Based Kinematic Slope can be used efficiently to form Indexed maps of Stability Analysis proposed sites. These maps contain the combined Marcus Gustavsson, Development of a Detailed effect of each sites’ historic temporal evolution for all Geomorphological Mapping System and GIS assessments and estimations, which can then lead to a Geodatabase in Sweden, Chapters 1-4: Review of very well informed choice by the military personal. Geomorphological mapping R Weibel and M Heller, Digital Terrain Modeling REFRENCES Jose Antonio Navarrete Pacheco, Digital Stereo Image Interpretation for Natural Hazard Kathleen Nicoll Geomorphic and Hazard Assessment Vulnerability Assessment of recent residential Firoz Verjee The Application of Geomatics in developments on landslide prone terrain: the case ComplexHumanitarian Emergencies study of Traverse Mountains, UTAH, USA. David Theler, Emmanuel Reynard and Eric Bardou, Geomorphological Mapping to Risk Assessment: M. Keiler, R. Sailer, P.Jorg, C.Weber, S.Fuchs, A. A Projectof Integrated GIS Applicationinthe Zischg and S. Sauermoser, Avalanche Risk Western Swiss Alps Assessment – a multi temporal approach, results from Galtur, Austria Reduce Exposure to Reduce Risk 289