Data Inventory Draft Technical Report

1Role of the Database and the Data Issue

A main issue is the ability of current or prospective indicators to provide information being location-specific, timely, and cost-effective. Macro-level environmental and socio-economic data can be useful to get an overview of the overall situation of a country/region, but are unable to provide information on those individual communities most in need of relief. On the other hand, while local-level indicators are better suited as sources of information on deteriorating economic conditions and food supply/access at village or sub-district level, their availability is scarce and data collection is expensive.

1.1The system

It should however be noted that, taking into account the complexity and the variety of factors involved in the development of a crisis, the decision leading process related, for instance, to the planning of emergency interventions can benefit from indicators available or generated by the information system only as part/components of the required knowledge. This means that the system is able to produce bits of information to be used by the emergency officers in conjunction with other factors derived from other “systems” such as field knowledge, personal experience, reports, guidelines and other sources. In other words, a system of this type is not meant to provide “the solution” of a considered problem, it rather provides a set of tools able to generate data and indicators, which can assist the decisional process.

When seen from this point of view, the emergency information system is rather an analytical tool supporting decisions than a comprehensive “decision support system” in the way it is commonly intended. This implies that modeling and analytical tools can help interpreting existing data without attempting to model the examined a situation in its complexity. The peculiarity of the individual situations reduces, in fact, the effectiveness of generic modeling tools intended for simulations and assessments of stereotypes. The effort might not be focused on developing a system able to provide solutions to a problem, but on the identification of key factors directly derivable from the system knowledge-base which can play a role in decisions concerning interventions and other emergency strategies and actions.

The natural conceptual framework of a system with the characteristics described above is modular, the components of which are virtually independent and specialized software applications designed to respond to particular (and therefore partial) user’s needs. Although this approach might seem to reduce the overall capacity of the information system to deal with complex situations, it actually gives the user the freedom to use data in the way it is more appropriate for the specific problem without being subject to fixed schemas or procedures which, at times, might not apply to the examined context. The modules can be available to the emergency officers who can decide to selectively use them and eventually to recombine outputs in appropriate ways, which might depend on the reality to be analyzed, but also on the availability of data. In addition, the modular approach does not prevent from developing situation-specific applications (more complex models) using the system’s knowledge-base or its products.

From the software development point of view, the modularity increases the potentials of the system to update, replace, add and delete components, decreasing - at the same time - the overall complexity of the software development operations, which would deal with smaller applications, and reducing the implementation costs.

1.2The data

Early warning systems, for example, are oriented toward the prediction of famine or food shortage. These systems use indicators such as global climatic data that may help predicting drought, but drought may occur without famine and vice versa. According to the meaning and weight attributed to the various indicators (food storage, undernourished population, access to water, etc.) different combinations of these parameters might help evaluating the magnitude of an event, estimating its potential danger and predicting its effects on local populations.

If these types of analysis are carried out to assess the impact of a certain event on a large region, the spatial resolution of the single indicator would not necessarily be high. Indicators gathered/available at global and national levels provide an overview of the region/country which might turn useful to acquire a first and general knowledge of the local conditions. A typical example of these indicators is represented by climatic and vegetation data derived from satellite providing good information on rainfall patterns and vegetation changes/health. Because these indicators (like also soils and hydrology) are more related to food production potentials rather than food needs and supplies, they might be more flexible in terms of spatial resolution.

Different considerations should be made for those indicators more related to the characteristics of the local communities and their food needs/availability such as size and distribution of the population, health and nutritional indices, data on economic activities or crop/livestock/fishery production. For emergency operations, information needs to be more detailed and should refer to specific locations, take into account the native characteristics and the number of the vulnerable groups or those communities more seriously affected by the crisis. At this level, the resolution of information has a paramount importance. Since macro-level data cannot be scaled down to provide information at micro-level, there is a gap between information that might provide indicators at regional scale and the information needed for purposes such as provision of food relief at village or household levels. At national scale, crop-climate models, volumes of crop sales, and market prices for crops and livestock may be useful as general indicators of food availability. However, it is only in a disaggregated form (sub-national level) that these indicators have the potential to be used for targeting populations in need of relief efforts.

Since finer spatial resolution data (i.e. at the sub-district or village level) may not be available for the lack of existing structures and for high costs of data collection, a balance should be made between hypothetical data needs and actual data collection capacity, for example, by limiting data gathering activities to only essential information. Following this logic, efforts on data capturing might be focused on filling gaps in data that is absolutely required for the implementation of identified applications.

The decision on how to approach the “data problem”, and eventually determine which data needs to be collected, is linked to the knowledge on both data availability and information requirements. This implies that outputs to be generated by the database be identified together with the associated data required by the applications to be developed. The figure below illustrates this process.