Expert of the Week   for  21 - 27 Jul 2014

Expert photo

You too can be featured here. Share your expertise in DRR with the community.

Sign up now!

Experts Log in


Julio Cesar Serje

United Nations Office for Disaster Risk Reduction (UNISDR) Expertise:  Disaster loss and damage databases

Julio Serje is a UNISDR program manager and senior software engineer with over 20 years of experience in disaster and risk information systems, in both the public and private sectors, and supported by a long standing experience within the UN System. During the past years Julio has focused on the development and application of information technology to disaster and risk management, climate change, environment and related areas in the form of applied GIS work and disaster and risk information systems and methodologies. He has been working closely with governments in the implementation and usage of these systems in practical risk reduction processes and in its policy applications.

Disaster and Risk Information Systems and Methodologies: Measuring and recording disaster loss and damage

Read more on the context

QQuestion by Mr Kerem KUTERDEM

It seems that after 2015 new strategy on DRR will have a more quantitative review system where disaster loss databases will be very important. What would you recommend countries planning to establish or develop disaster archieves or databases? Thank

Mr Kerem KUTERDEM Geologist | (AFAD) Disaster and Emergency Management Presidenc

APosted on 25 Jul 2014

Yes, the ongoing consultations on a new, Post-2015 Framework for Action are being very solid pointing out that countries need to better track losses.  The proposed monitoring system will contain a series of indicators related to losses and also to the existence of these databases on each country.

Implementing a Disaster Loss database implies both the solving the technological and institutional challenges, being the institutional challenge the main point.  The main recommendation is that a governmental institution (a Hosting Agency) takes full responsibility for the daily maintenance of the dataset.

Emergency Management agencies (Civil Defense and Protection mechanisms, for example) are usually great candidates for hosting disaster loss databases, given the fact that they are informed almost immediately of all incidents, especially those requiring assistance.  The ideal situations is that the system is also used and a near-real time incident reporting system. 

In terms of technology there are several ways a country can establish its loss accounting system.  In order to implement a disaster loss database almost any tool (from excel to sophisticated databases) can be used.  

 However  we at UNISDR (and other UN agencies) are convinced that countries do not need to go through a painful and probably costly software development, or using a tool that will not provide the necessary 'intelligence' to their users. Our recommendations would be to first evaluate the FOSS (free, open soft software) DesInventar,  available for download at DesInventar has also been recommended by the European Commission for those countries wishing to build aggregated disaster loss databases.

DesInventar is a specialized, well tested, always growing set of tools for creating, maintaining and analyzing disaster loss databases.  It's analysis capabilities include aggregation/statistic functions, charting of results and production of loss maps.  Previous experience indicates that while creating a data entry system is relatively easy, implementing the large pool of analytical tools of DesInventar takes a much more effort.

If Countries so desire, the UN can provide free hosting on, removing the need for investments in site maintenance and hosting. See the About page of the site for more information. 

DesInventar is a web-based tool easily installable on a server but can be also used on laptops or desktops. It can be run by most OS (Windows, Linux, Unix) and supports most database platforms, (SQL Server, Postgres, MySQL, Oracle, Access, etc).

Seeing Mr. Kuterdem, who posted this question, is from Turkey, I would like to add that the DesInventar implementation in that country is an example of what can be done with Open Source software:  please visit  with a great disaster loss database for Turkey. An important feature of this site is the addition of new functionality (the front page map was developed in Turkey) to complement the already rich set of functions of the system.  

QQuestion by Ms Lucia Serje

Is the software for schools also available in the App store for desktop computers? As we live in the US Virgin Islands - hurricane path- developing knowledge and awareness in this regard in my students is key. What is the exact name of this app?

Ms Lucia Serje AMI Elementary Teacher | VI Montessori School
United States of America

APosted on 25 Jul 2014

Unfortunately the apps are only available for tablets. However, please keep checking, we may come up with something!.   Also, please explore the Risk Data Viewer ( which may have some interesting maps and data - it can be accessed by any machine with a web browser.

In addition to the tab apps you may also explore the educational materials in this website. Just search "Educational", top of this page and you will see plenty of materials for all ages.


QQuestion by Ms Frances Francis

Are you developing software for use in schools? Geography lessons would be far more relevant if matched with contemporary issues and current data gathering practices. Is there modified software designed for younger students?

Ms Frances Francis supply teacher | teaching
United Kingdom

APosted on 23 Jul 2014

Yes!  We have developed and will continue to improve the Gft  (pronounced 'gift'), the GAR for Tangible earth, now available in the App Store for iPad, which will be available for Android devices as well very soon.

The GfT features a 3D view of the earth on which a large number of layers of information about hazards and risks, as well as stories, videos and many other teaching resources are available, using the well known interactive capacities of tablets, in a simplified version of the more technical Risk Data Viewer.

The Global Assessment Report on Disaster Risk Reduction (GAR) 2013 has augmented reality (AR) icons sprinkled throughout it, which facilitate connection with GfT. Pointing the camera of the tablet at  these icons will start a variety of animations and  dynamic information functions designed to enrich the reading experience.

The Risk Data Viewer, which can be used by more mature students ( displays and organize in 'perspectives' the growing number of layers of hazard, exposure and risk information produced by UNISDR and its scientific partners in the development of a global model of risk. 

QQuestion by Ms Maria Corazon de la Paz

What are acceptable secondary sources of data from line agencies that may be used to intelligently establish damages and losses as basis for recovery and rehabilitation?

Ms Maria Corazon de la Paz Treasurer of the Board, Program Development Consul | Balay Rehabilitation Center, Inc.

APosted on 23 Jul 2014

A source of data is said to be primary if it gets its information by direct observation of damages and losses on the ground, and reports exactly what was observed. For example, Fire Fighters do report very well what they observe when they are activated and deployed to attend an emergency. 

 Primary data sources are of several classes, those related to Emergency/Humanitarian Services, those related to infrastructure, utilities and services  and those related to insurance. Emergency services range from fire brigades to civil protection mechanisms to NGO's (Non Governmental Organizations) both national and international, such as Red Cross/Red Crescent, the United Nations (OCHA, WFP, WHO, etc). Line ministries or service providers obtain information related to its particular purpose, for example roads, agriculture, electricity or communications.  Insurance companies also use personnel in the field to verify and evaluate damages to insured property - although in most cases this information is not made available.

Any other source which summarizes, consolidates, interprets and disseminates information obtained from primary sources is therefore secondary.

Official secondary data sources are usually the most reliable. Local, provincial and national governments and its associated services/infrastructure providers are definitely the best secondary sources, as they consolidate information coming from smaller units and different sectors in the information chain. Large humanitarian and reconstruction organizations are also usually very good, especially the Red Cross/Crescent system, the UN system and the World Bank, all of which work with carefully crafted and well proven methodologies (PDNA, DALA, etc).

It is important to understand the advantages but also the limitations of secondary data, and how to use it. Given its role of consolidators of data, very often  the information from secondary sources is considered reliable for recovery, rehabilitation and reconstruction only after a few days, even weeks. Also, many secondary data sources use sampling and statistical  techniques to extrapolate the overall loss and damage, information that has to be used also carefully taking into account its uncertainties, as opposed to exhaustive data collection exercises which provide accurate measures with low level of uncertainty.

Last but not least, media sources are an important secondary data provider. The reliability of media providers has to be determined case by case; some media sources exaggerate impacts in order to make headlines (please note that secondary official sources can also be biased). However, 'serious' media sources can be extremely useful and may prove to be very reliable - especially those who obtain, consolidate and reproduce officially released information.  Media sources  can be very useful when conducting historical research. 

QQuestion by Ms Jana Junghardt

In our projects we realize that quantifying the risks and assessing local loss and damage will be needed to promote investments in DRR. Can you recommend tools and approaches at local scale for collecting data and setting up such databasess?

Ms Jana Junghardt Advisor DRR | Caritas Switzerland

APosted on 21 Jul 2014

Yours is a very interesting question as it highlights the fact that approaches vary with the scale covered by a project.

It should be answered in two parts, the first one regarding the assessment (and tracking) of loss and damage and the second on quantifying the risks. The order is important as it should reflect the way capacities and actions should be raised and implemented.

Assessing losses and damages in post disaster situations is a general process that is well documented from the methodological point of view in three inter-related methodologies, the UN-ECLAC (Economic Commission for Latin America and the Caribbean), the DALA (originated in ECLAC but refined and somewhat simplified methodology by the World Bank) and the more complex PDNA, which includes the DALA/ECLAC on loss and damage while addressing Needs Assessments from a humanitarian and social perspective, this last part based on a number of sectorial methodologies developed by several UN agencies (WHO, WFP, ILO, etc.).

All these methodologies work well and are focused on the local level, but also provide means of aggregation and macroeconomic views in case of large scale disasters.
An important question is the management of data about loss and damage over long periods of time. The results of these assessments (and even of very simplified D&L assessments) can be systematized on Disaster Loss Databases.  A well known disaster loss database is EMDAT, but is focuses on a global level of observation and  has a country level resolution.

The UN system (UNDP, UNISDR mostly) has promoted and supported the creation of disaster loss databases at a more local level, usually at municipality or village level, using a methodology and software called "DesInventar", originally developed in Latin America, and later refined as an Open Source, free software.   For those interested, many of these localised disaster loss databases are hosted (along with the software and documentation) at

When these disaster loss databases contain sufficient information they can deliver extremely good proxies of Risk, something very handy for hazards that are hard to model of for which there is no much information, and can uncover patterns and trends to focus further work on risk assessments. Disaster loss databases are also a crucial input to risk assessments and are great tools to validate and calibrate them. In general, building these databases is much cheaper than conducting a multi-hazard rigorous risk assessment. 

Once the message of "How much has been lost" is delivered by past loss and damage data, a risk modelling exercise is needed to calculate how much can be lost due to catastrophic events, many of which will happen probably in the long term.

There is a wealth of literature on risk assessments, a very broad subject, so I will only mention a few tools that can be applied at local level.  There are many types of risk assessments, but is important to define them as the process by which future losses are measured in size and in the probability of happening.

At local level there have been many efforts in conducting Community Based risk assessments. There are relatively informal and still very much qualitative. The positive aspect is that they can integrate the local knowledge of hazards (which can be very detailed) and about the social fabric.  They require relatively less technical knowledge but in many cases lack the quantitative measurements of risk, something required to define, for example, a financial strategy for risk reduction. 

Obtaining a quantitative measurement of risk can be a complicated task, requiring vast and complex data and highly trained professionals. However, there are more and more tools to support these efforts, that must start by developing Hazard, Exposure and Vulnerability models and then 'crunch' all these data to convert it into measures of risk such as Annual Average Loss, Probable Maximum Loss or more probabilistic expressions of risk such as a Loss Exceedance Curve. 

For local purposes the CAPRA  (Comprehensive Approach to Probabilistic Risk Assessments) set of tools and methodologies can be used.  CAPRA was developed as a partnership between the World Bank, the UN and other organizations. It is also Open Source and Free software, which removes one of the main barriers to use these solutions, previously reserved to big insurance and consulting firms. CAPRA tools and methodologies can be downloaded from