DesInventar as a Disaster Information Management System

The challenge of information sources

Disaggregating data is not the only challenge to be faced by data collectors. Another major challenge arises when identifying all possible data sources for a national effort, making the information accessible and in many cases, trying to conciliate multiple data sources report dissimilar figures when describing the effects of the same event.

Common sources of information for disaster data can be classified in several main groups, ordered here by the priority in which they should be included in a national level inventory:

• Information files created and maintained by OFFICIAL emergency management agencies. These sources of information often contain only information about events that required intervention from part of the organization, which usually are medium to large size disasters. Those agencies in many cases tend not to record disasters that the communities can cope with, and leave unregistered many small and medium disasters. However, information coming from these sources is normally the most reliable when available.
• Information files created and maintained by OFFICIAL sectorial institutions, such as Ministries of Agriculture, Public Works, Communications and Transport. Again, these sources of information often contain only information about events that required intervention from part of the organization, which usually are medium to large size disasters. Those agencies usualy only report the damage and losses of related assets (crops, roads, etc.)
• Archives of relief or aid organizations, such as the Red Cross or the Red Crescent. These sources of information have extremely good coverage of those events that required search and rescue, or the provision of emergency health services. In most cases the strength of these data sources is the good data about human and social losses (deceased, injured, evacuated, etc.). However, they rarely record damages and impact on economic assets.
• Academic and Scientific files, maintained by research institutions that are frequently interested in a specific kind of event, and more focused in the event itself than in the impact of events on communities. Examples of these are databases maintained by seismological and meteorological research centers.
• Media releases, specifically written media (newspapers). Despite the resistance that this source awakes in many ‘scientific’ researchers, LA RED’s experience building disaster databases in over 16 countries and many regional instances over almost a decade of work has shown the usefulness of this information source.

There are several facts that cannot be ignored and turn the media information a frequent requirement for disaster inventories:


1. Especially small disasters are NOT registered by any other source of information. The use of media releases becomes mandatory if a comprehensive database is to be built covering disasters at all scales.
2. Media is self-controlling in nature: whereas there may be under or overestimation in damages in press releases, the abundance of this type of sources permit the researcher to compare between multiple visions coming from different newspapers and even between editions or articles within the same source.
3. Media takes in most information produced by the preceding groups, or at least is one of the inputs used to create their news.
4. Most newspapers keep very organized and publicly accessible archives as opposed to other sources whose information may be restricted, difficult to access or disorganized and mixed with an overwhelming amount of operative data.
5. Information on newspapers can be obtained for many years backwards, even for periods in regions and countries where no other formal sources of information on disaster effects or even agencies in charge of emergencies were put in place.
6. Locals can easily qualify newspapers reliability. Reputation of a newspaper is a measure that enormously helps when making decisions about the information to be integrated in the inventory.
7. There is some continuity in the quality and comprehensiveness of each media source, especially in this considered as ‘serious’.

UNISDR has conducted formal research of the above subject, with very interesting results showing that inventories made entirely based on media sources can be extremely comprehensive, and usually equally reliable as inventories made from official sources.

However, it’s important to concede that information in the databases is only as reliable as its source. The International Federation of Red Cross and Red Crescent Societies acknowledges that “...most reporting sources have vested interests, and figures may be affected by socio-political considerations.” Petak and Atkisson state that “...media exaggerations of disaster impacts on a community apparently increase as a function of distance from the disaster site.”. Our own experience in many developing countries suggests that coverage of the media decrease as a function of the distance from the disaster site (9).