Distr.: For participants only 21 April 2011 English only Economic and Social Commission for Asia and the Pacific Committee on Disaster Risk Reduction Second session Bangkok, 29 June-1 July 2011 Item 4 of the provisional agenda Recent trends in disasters and their socio-economic and environmental aspects Summary Addressing disaster risk reduction and development through improved data on disasters* The present information note, prepared by the secretariat with a contribution from the Asia-Pacific Regional Centre of the United Nations Development Programme, provides a concise review of available international and national statistical disaster databases. It points to the need to improve the capacity to collect and analyse disaster-related data and statistics and to improve collaboration between national statistical systems and disaster management authorities. Contents I. Special characteristics of disaster data... 2 II. Internationally comparable data and disaster trends... 2 III. National disaster loss databases... 3 IV. Lessons learned from establishing disaster databases... 4 V. International development goals call for improvement in statistical capacity... 6 VI. Annex Capacity building needed for integrating statistics to national disaster management... 8 I. Recent disaster trends in Asia and the Pacific... 10 II. UNDP approach to establishing national disaster loss databases... 13 * The present document has been issued without formal editing. DMR A2011-000116 TP300511 CDR2_INF5
I. Special characteristics of disaster data 1. Reliable and timely statistics are essential for disaster preparedness planning and response readiness. After a disaster has struck, statistics are used for assessing the initial disaster impact, and for subsequent monitoring and evaluation. The information for pasts disasters are also used in policy planning and decision-making at all levels in the country for disaster risk reduction, preparedness, mitigation, response, relief and recovery. Disaster databases and resulting statistics serve also as tools for measuring the progress and achievements of the policies and plans. 2. The timing of disasters cannot usually be predicted, but advance preparation and contingency planning can reduce disaster risk and mitigate the impact of disasters. Statistics and statistical methodologies are tools to be used in that prior preparation. They also play a critical role in the wake of disasters in providing a correct picture of the scale of losses and the progress of relief and reconstruction. 3. Statistics telling about already occurred disasters have special characteristics as disasters (i) can be localised affecting only a small area and population, (ii) can affect a major part of a country, or even (iii) can expand to the territory of more than one country. The duration of a disaster can vary from momentary (earthquake) to several weeks (flooding) or even years long (prolonged drought). 4. Apart from surviving victims of disasters as respondents of information collection, the reporting on disasters involves players from various national authorities and relief organizations. Consequently, the events are more than likely to become recorded and reported differently in various databases and data sources. As data collection can only be a secondary priority at the time of response and emergency relief, reaching consistency in the reporting on disasters is challenging. Statistical methods in disaster data collection and analysis need to be pragmatic and sound enough, but they hardly can be perfected as methodological prudence is time-consuming and costly. Although the data about disasters and their impacts is collected after the occurrence of the events, the institutional capacities, systems and protocols for collecting the data is required to be built prior to the occurrence of such events II. Internationally comparable data and disaster trends 5. The leading publicly available international database, the Emergency Events Database (EM-DAT), records core data on the occurrence and effects of over 18,000 mass disasters in the world, stretching, in principal, from year 1900 to present. EM-DAT was created with the initial support of the WHO and the Belgian Government and is maintained by the Centre for Research on the Epidemiology of Disasters (CRED) at the School of Public Health of the Université Catholique de Louvain in Brussels. Their website 1 provides up-to-date weekly lists of 1 EM-DAT. The International Disaster Database, www.emdat.be. 2
individual natural disasters and technological disasters (e.g. fires, traffic and mining accidents). 6. The current EM-DAT data recording system uses a unique identifier 2 for each disaster and attempts to cover all disasters meeting at least one of the following criteria (i) ten or more people reported killed, (ii) 100 or more people reported affected, (iii) declaration of a state of emergency, and (iv) all for international assistance. 7. The advantage of this database is its seeming consistency of reporting. The ambitious global coverage means that the number of variables collected is necessarily limited. EM-DAT only records disaster incidents and casualties in terms of the number of deaths and people affected, and economic damage caused. 8. EM-DAT can be used to compare trends of major types of disasters between different countries and areas. The Asia Pacific Disaster Report 2010 analysed disaster trends and disaster risk in detail 3. While an upward trend was observed in the number of reported disasters, the Report noted that it could be attributed to the increase of population exposed to natural hazards and improvements in reporting of disaster to the database. The Report compared EM-DAT data to nationally reported disaster data in selected countries and found discrepancies between the sources. This points out to the need to improve the need to harmonize methodologies and improve the quality of data collection in all sources. 9. Annex I provides an update of the latest disaster trends for different world regions from 1980 onwards. The Asian and Pacific region has consistently had by far the largest number of people affected by natural disasters in the world. Also the region s estimated economic damage and number of deaths have been topping the charts almost every year. III. National disaster loss databases 10. Disaster management at national level requires detailed data, including on small disasters in particular those that are frequently occurring, below the EM-DAT thresholds that can have wide human, economic and environmental impact. A large number of countries in the Asia-Pacific region are vulnerable to disasters and would benefit from more systematic approaches in disaster-related data collection and from better integration of data collection and use in the overall disaster management. 11. UNDP Asia-Pacific Regional Center (APRC) has supported the establishment of disaster loss databases in several countries in Asia. The databases are configured and adapted to the needs of the country or province implementing them, including having the user interface in local languages and customizing the data items to be collected. The databases facilitate capture and analysis of the occurrence and impact of disasters for 30 different types of hazards and by three levels of administrative regions (e.g. province, district and sub-district). 2 3 Compatible with GLIDE (GLobal IDEntifier) number, see www.glidenumber.net/ See chapter 1 of the report. Available online at www.unescap.org/idd/pubs/asia- Pacific-Disaster-Report%20-2010.pdf 3
12. Countries where databases with historical data of the past events have been set up using the DesInventar methodology include Indonesia, the Islamic Republic of Iran, Maldives, Nepal, and Sri Lanka (country-wide coverage) and India (states of Tamil Nadu and Orissa). Indonesia has also adapted the methodology to monitoring of poverty. Similar efforts are ongoing in Timor-Leste, Viet Nam and the Lao People s Democratic Republic. The databases are not country-wide but pertain to regions that have historically suffered from disasters or have otherwise become aware of the need for and benefits of systematic disaster data management 4. 13. The data from national disaster databases are used by policy makers and other stakeholders at national and sub-national levels in support of disaster risk reduction, mitigation, preparedness, response and recovery. As a result, they have better analysis and understanding of extensive and intensive disaster risks and the link between disaster and poverty. The databases allow the government to better understand the disasters and threats in order to effectively mitigate and prepare for them, which is particularly important when a nationwide risk assessment does not exist. IV. Lessons learned from establishing disaster databases 14. Compiling statistics on disasters consistently over time and across regions is very challenging for a number of reasons that are valid both for international and national databases: (a) Baseline information that could be used to estimate the initial impact is often non-existent or is outdated, and the creation of the baseline ex-post is problematic. (b) sources. Geographic areas used are not consistent across available data (c) Where the assessments are able to use satellite and aerial imagery, the baseline data may not be available and their interpretation may be based on different methodologies. (d) The access to disaster areas and affected people who could provide accurate information is invariably difficult. (e) The made estimates, including their comparability with other disasters, depend on the assessors experience and the used estimation methodology. (f) While the direct human casualties are often relatively simple to estimate, the estimation of the economic and environmental damage can be done with a number of methodologies, resulting in magnitude-scale differences between estimates. (g) While the estimation of the impact for immediate disaster relief and early recovery usually takes the priority, and rightly so, the 4 Some databases are online: Indonesia (http://dibi.bnpb.go.id, and http://dibi.jogjaprov.go.id ), Sri Lanka (http://www.desinventar.lk ), Tamil Nadu (http://indisdata-tn.gov.in). 4
assessment of the overall impact for longer-term recovery requires comprehensive statistics of all sectors of the economy and monitoring the recovery over a long period of time. (h) Estimates on disaster variables are likely to suffer from different kind of reporting biases at different points of a long-term time series. (i) Where the disaster area covers land or water of more than one country, the problems related to data availability and reporting consistency are compounded. 15. The UNDP Regional Programme on Capacity Building for Sustainable Recovery and Risk Reduction has documented 5 experiences in implementing disaster loss databases using the DesInventar methodology in Asia. The lessons learned can be summarised as follows 6 : (a) Disaster loss databases must be developed as an integral part of overall disaster risk reduction initiatives. Before the methodology and data collection process is put in place, there must be a good understanding among the stakeholders of why the data are collected and what the end use of the data will be. (b) Government ownership of the database from the outset is a critical success factor. Implementation outside of the government system should only be considered as a last option. Other determinants of success include advanced planning, assessment and appropriateness with regard to identifying counterpart nodal agencies and human resources at regional and country levels (c) The process of disaster loss database implementation needs to be participatory and inclusive, involving governments and other partners to promote government ownership of the system and its institutionalization. (d) Clear and complete documentation (data card and photocopy or digital copy of the source of data) are essential for data validation and quality assurance. (e) Data collection staff must understand disaster terminology and be properly trained to collect and enter the data and requisite metadata. The training and professional development should be continuous and involve counterparts. (f) The data collection process must be structured and have cross checks within it. Data records should be from agreed and accepted sources and the sources must be documented and easily accessible. 5 6 Risk Knowledge Fundamentals: Guidelines and Lessons for Establishing and Institutionalizing Disaster Loss Databases. UNDP. 2009. Available online at www.snap-undp.org/elibrary/publications/dldguidelines.pdf See Annex II for details of UNDP s approach to establishing national disaster loss databases. 5
(g) Analysis must be professional, clear, understandable and relevant to the target audience. Different levels of analyses should be prepared depending on the audience of the analysis. But in general, they should comprise quantitative and qualitative information that is userfriendly and supports the decision-making process. The database teams should have multi-skilled personnel in order to interpret the results correctly and comprehensively and prepare quality outputs. (h) A disaster database without customization could be a wasted investment. Technical support has proved particularly useful in customizing the database to meet the needs of the government, and to ensure that it complements existing government systems and requirements, so that at the end of the day it is actually used. Also, the manuals are needed to be customized with the target group in mind. (i) More efforts are needed by national and international development partners in advocating for the importance of disaster data collection and analysis. V. International development goals call for improvement in statistical capacity 16. The priorities for the United Nations disaster-related statistical activities are inscribed in the United Nations resolutions and World Conferences, especially the Hyogo Framework for Action 2005-2015. Among the Hyogo Framework for Action indicators 7 is a specific one on systems that should be in place to monitor, archive and disseminate data on key hazards and vulnerabilities. From an institutional point of view, some others are also relevant for statistics, as compiled in Table 1. Table 1 Indicators of the Hyogo Framework for Action that have relevance for official statistics Hyogo Framework for Action indicator Systems are in place to monitor, archive and disseminate data on key hazards and vulnerabilities. Make disaster risk reduction a priority. Improve risk information and early warning Possible types of action by statistical offices Invest in the improvement of technical capacity, systems development, and coordination. Mainstream disaster risk reduction in planning, define responsibilities and plans for statistical support for disaster risk reduction and assign a budget. Provide data and methodological support for risk assessment and data 7 The Hyogo Framework for Action indicators are consolidated in annex III of the United Nations ISDR publication Words into Action: A Guide for Implementing the Hyogo Framework for Action 2005-2015: Building the resilience of nations and communities to disasters, which is available online at www.unisdr.org/eng/hfa/docs/words-into-action/words-into-action.pdf. 6
Hyogo Framework for Action indicator - National risk assessments based on hazard data and vulnerability information are available and include risk assessments for key sectors. Possible types of action by statistical offices analyses on hazards and vulnerabilities and advise in data systems development. - Systems are in place to monitor, archive and disseminate data on key hazards and vulnerabilities. Strengthening preparedness for response - An independent assessment of disaster preparedness capacities and mechanisms has been undertaken and the responsibility for implementation of its recommendations have been assigned and resourced. Provide data and methodological support for national, regional and community-level preparedness planning - Disaster preparedness plans and contingency plans are in place at all administrative levels, and regular training drills and rehearsals are held to test and develop disaster response programmes - All organizations, personnel and volunteers responsible for maintaining preparedness are equipped and trained for effective disaster preparedness and response. - Financial reserves and contingency mechanisms are in place to support effective response and recovery when required. - Procedures are in place to document experience during hazard events and disasters and to undertake post-event reviews. 17. The recent experiences in the region in dealing with disasters have demonstrated that the preparedness for disasters is generally not at a desirable level. Statistical data systems and databases and the capacity of various national authorities to collaborate on data issues before, during and after disasters need and can be substantially improved. 7
VI. Capacity building needed for integrating statistics to national disaster management 18. Statistical data and methodologies are applicable for all stages of the disaster management cycle. Table 2 outlines statistical capacities that critical for different stages of the disaster management cycle, along with supportive actions that relevant statistical offices can provide. Much of the statistical capacity created for the prevention and preparedness stages is also relevant for the eventual impact assessments and recovery monitoring. Table 2 Statistical capacity at different stages of disaster cycle Stage of disaster management cycle Mitigation Critical statistical capacity needed - Availability of small-area statistics, including community baseline statistics, to estimate and map the probability of different kind of hazards and the vulnerability of population to various risks - Regular conduct of population and agricultural censuses - Quality of statistics based on administrative records - Use of spatial coding in datasets. - Other capacity to link datasets - Use of GIS - Capacity to support evaluations of previous disasters - Capacity to support socioeconomic and environmental impact assessments of previous disasters Support that statistical offices might provide - Analyse and record lessons learned for statistics from previous disasters. - Review of regular data collection programmes and instruments for disaster management. - (Ad hoc and sustainable) exchange of data and metadata within the national statistical system and with national disaster management authorities - Adopt GIS and other systems improvements. - Collect additional data in hazard-prone areas. - Improve coordination between the national statistical system and national disaster management authorities. Preparedness - All the above, plus capacity to collaborate and coordinate data support for contingencies - Capacity to support baseline data - Map available official data and assess data gaps within the national statistical system. - Use data to design and implement action plans and scenarios that would be 8
Stage of disaster management cycle Response/relief/ recovery Critical statistical capacity needed - Capacity to provide and make use of methodological advice in data collection - Capacity to provide baseline data - Analytical support for damage loss and need assessment, and monitoring Support that statistical offices might provide triggered by different type and size of disasters - Improve relevant information systems and their ability to exchange data. - Assess the readiness of information systems and protocols to promptly provide data, including baseline, for rapid assessments. - Send statisticians and sectoral experts to contribute to rapid impact assessment teams - Review assessment reports for methodology and data presentation Reconstruction - Capacity to provide methodological advice for monitoring and evaluation, and to support longer-term economic impact assessments - Participate in and review impact assessments and evaluation reports 19. The facilitation of national reporting and standards, and the development of access mechanisms for national information are among the realistic data support activities that ESCAP could get involved, in addition to the support that UNDP is already providing in establishing national disaster databases. The strengthening of institutional capacity of all stakeholders, including the national statistical system, to coordinate disaster preparedness and contingency planning is something that both agencies could facilitate, and could involve other development partners. 20. The statistical contingency plans for disasters could include the definition of each party s responsibilities and specific actions that a disaster should automatically trigger (who will supply, what data, in which format, by when) and simulation exercises of different scenarios. A supportive undertaking could be the creation of a central repository to store methodologies, good practices and metadata related to hazard risks and vulnerabilities, which would eventually facilitate assessing and monitoring the impact of disasters. 9
Annex I Recent disaster trends in Asia and the Pacific The Asia Pacific Disaster Report 2010 a provided an overview of disaster trends in Asia and the Pacific from the year 1980 onwards. It found that floods and storms were the two main causes of disasters in the region and affected the largest number of people. 120 100 Number of disaster events meeting the EM-DAT threshold Flood Earthquake (seismic activity) Drought Volcano Mass Movement Wet Storm Extreme temperature Wildfire Mass Movement Dry 120 100 80 80 60 60 40 40 20 20 0 1980 1985 1990 1995 2000 2005 0 a www.unescap.org/idd/pubs/asia-pacific-disaster-report%20-2010.pdf. 10
The charts below provide an update on selected disaster trends in different world regions on the basis of global data retrieved in March 2011 from the same source (EM-DAT Emergency Events Database, www.emdat.be). Please note that all charts refer to the aggregated impact of all types of disasters. Africa Asia-Pacific Europe Latin America and Caribbean North America Number of disaster events meeting EM-DAT threshold 250 200 150 100 50 0 1980 1985 1990 1995 2000 2005 2010 Africa Asia-Pacific Europe Latin America and Carib. North America 700 People affected by natural disasters (millions) 600 500 400 300 200 100 0 1980 1985 1990 1995 2000 2005 2010 11
500 Deaths caused by natural disasters (millions) 450 400 350 300 250 200 150 100 50 0 1980 1985 1990 1995 2000 2005 2010 180 Economic damage caused bynatural disasters (billion US dollars, 2005 prices) 160 140 120 100 80 60 40 20 0 1980 1985 1990 1995 2000 2005 2010 12
Annex II UNDP approach to establishing national disaster loss databases During the recent years, UNDP through its Asia-Pacific Regional Centre (APRC) based in Bangkok has provided technical support and helped establish a number of national disaster loss databases in Asia. The establishment of these databases have been guided by the following principles: Developing national capacities for establishing and maintaining disaster loss databases Establishing and sustaining nationally led processes in the countries to create ownership of the database and increase its usefulness and relevance to national and sub-national contexts Establishment of national disaster loss database should be guided by the overall institutional and legal context of disaster risk reduction in the country The disaster loss database should address the needs and priorities of the country and the analysis must provide inputs to the policy and decision-making processes by at all levels within the government The database and any analysis should be shared with all key stakeholders and the public for developing wider understanding of risks and warranting actions from all sections of society. The goal of the support has been to build national capacities for better understanding of disaster risks over time and space and to provide support to policy and decision-making and planning for disaster risk reduction, preparedness, mitigation, response and recovery at national and subnational levels. In addition to being useful at the country level, the Global Assessment Report (GAR) on Disaster Risk Reduction in 2009 and 2011 made use of these databases to identify extensive risk which in the past were not captured by other databases. Disaster loss databases were established in a variety of institutional context with varying levels of capacities. Indonesia, the Islamic Republic of Iran, Nepal and Sri Lanka have established disaster loss databases with countrywide coverage and historical disaster data going back up to 30 years. In India, Tamil Nadu and Orissa states have achieved the same, and in several provinces of Indonesia (Aceh, Yogyakarta, East Nusa Tenggara, Bengkulu, and Central Java) have established separate databases with long time series. Several other states in India (Uttranchal, Andhra Pradesh, Kerala and Pondicherry) have established disaster loss databases with shorter time series going back several years. Many other countries are in the process of establishing their national disaster loss databases, with varying stages of progress such as Maldives, Timor-Leste, Viet Nam, Thailand and Cambodia. UNDP APRC has captured the experiences of establishing disaster loss databases and has made it available in a publication titled Risk Knowledge Fundamentals: Guidelines and Lessons for Establishing and 13
Institutionalizing Disaster Loss Databases. The following factors are essential for a successful implementation of disaster loss databases: Promote a clear understanding that data and inventories of disasters are vital for identifying and tracking patterns of disaster risk, and that they are a fundamental requirement for implementing efficient and effective disaster risk reduction programmes, policies and planning. Recognize that the establishment and support of an enabling environment for disaster risk reduction to assist with the institutionalization and long-term sustainability of a disaster loss database must either be in place, or fully resourced or provided for in such an initiative. Aim to create a database initiative in conjunction with other disasterrelated capacity building activities and within government structures to ensure local ownership and management of the data. Customize the system to meet national conditions including language, definitions of disasters and the local situation. Ensure that data collection and validation are from agreed and accepted sources, and involve government counterparts in this process. Produce high quality analysis from the database that can assist governments in disaster planning and preparedness, and in making informed decisions. Ensure that technical support is in place for the system even after the disaster loss database has been successfully implemented. Follow five key steps to successfully implement a disaster loss database: 1. 2. 3. 4. 5. Create an enabling environment for disaster risk reduction Find an appropriate home for the database Establish the disaster loss database within disaster risk reduction framework Collect, enter and validate data Analysing, managing data and sustainability It is essential to have a clear understanding that the successful implementation is contingent on a number of factors beyond the physical database itself, and requires the allocation of substantial resources and time and above all, full support and ownership by the government. 14