DEALING WITH UNCERTAINTY IN LARGE-SCALE SPATIAL DATABASES

This experimental study addresses the issue of uncertainty inherent in large-scale spatial databases. A contemporary Spanish large-scale topographic database, compiled according to the current standards, is used to examine uncertainty patterns and propose practical solutions for their communication. A set of specific, inexpensive and explanatory visualisations is developed for each of the data quality categories within an uncertainty typology proposed by Thomson et al. (2005). The resultant visualisations offer a compact and direct view of uncertainty aimed at novice users and provide means to deal systematically with uncertainty inherent in spatial databases. The disadvantage of this approach includes the necessity of arbitrary decisions for metadata classification. A concept of global visual uncertainty indicator (G-VisUI) is introduced as an extension to visualising individual categories of data quality. The G-VisUI provides an immediate overview of the overall quality of the database with the ability of alerting users to local variations and anomalies. However, the strategy for visualising G-VisUI is still immature and requires further research and testing