Sensor networks generate substantial amounts of frequently updated, highly dynamic data that are transmitted as packets in a data stream. The high frequency and continuous unbound nature of data streams leads to challenges when deriving knowledge from the underlying observations. This paper presents (1) a state of the art review into visual analytics of geospatial, spatio-temporal streaming data, and (2) proposes a framework based on the identified gaps from the review. The framework consists of (1) the data model that characterizes the sensor observation data, (2) the user model, which addresses the user queries and manages domain knowledge, (3) the design model, which handles the patterns that can be uncovered from the data and corresponding visualizations, and (4) the visualization model, which handles the rendering of the data. The conclusion from the visualization model is that streaming sensor observations require tools that can handle multivariate, multiscale, and time series displays. The design model reveals that the most useful patterns are those that show relationships, anomalies, and aggregations of the data. The user model highlights the need for handling missing data, dealing with high frequency changes, as well as the ability to review retrospective changes.

ABSTRACT:Geospatial data has very specific characteristics that need to be carefully captured in its visualisation, in order for the user and the viewer to gain knowledge from it. The science of visualisation has gained much traction over the last decade as a response to various visualisation challenges. During the development of an open source based, dynamic two-dimensional visualisation library, that caters for geospatial streaming data, it was found necessary to conduct a review of existing geospatial visualisation taxonomies. The review was done in order to inform the design phase of the library development, such that either an existing taxonomy can be adopted or extended to fit the needs at hand. The major challenge in this case is to develop dynamic two dimensional visualisations that enable human interaction in order to assist the user to understand the data streams that are continuously being updated. This paper reviews the existing geospatial data visualisation taxonomies that have been developed over the years. Based on the review, an adopted taxonomy for visualisation of geospatial streaming data is presented. Example applications of this taxonomy are also provided. The adopted taxonomy will then be used to develop the information model for the visualisation library in a further study.

Transient spatiotemporal events occur within a short interval of time, in a particular location. If such events occur unexpectedly with varying durations, frequencies, and intensities, they pose a challenge for near-real-time monitoring. Lightning strikes are examples of such events and they can have severe negative consequences, such as fires, or they precede sudden flash storms, which can result in damage to infrastructure, loss of Internet connectivity, interruption of electrical power supply, and loss of life or property. Furthermore, they are unexpected, momentary in occurrence, sometimes with high frequency and then again with long intervals between them, their intensity varies considerably, and they are difficult to trace once they have occurred. Despite their unpredictable and irregular nature, timely analysis of lightning events is crucial for understanding their patterns and behaviour so that any adverse effects can be mitigated. However, near-real-time monitoring of unexpected and irregular transient events presents technical challenges for their analysis and visualisation. This paper demonstrates an approach for overcoming some of the challenges by clustering and visualising data streams with information about lightning events during thunderstorms, in real time. The contribution is twofold. Firstly, we detect clusters in dynamic spatiotemporal lightning events based on space, time, and attributes, using graph theory, that is adaptive and does not prescribe number and size of clusters beforehand, and allows for use of multiple clustering criteria and thresholds, and formation of different cluster shapes. Secondly, we demonstrate how the space time cube can be used to visualise unexpected and irregular transient events. Along with the visualisation, we identify the interactive elements required to counter challenges related to visualising unexpected and irregular transient events through space time cubes.