Analysing methods for visualizing time-series datasets in open-source web mapping

Abstract: Geography is the study of knowledge in a given location, and history is the study of knowledge over time. The combination of changes over time has always been challenging to visualize in maps. The main aim of this thesis was to analyze differences between various methods used to visualize spatio-temporal data with open-source web mapping technology. The methodology of this thesis has been split into three main parts: (1) data collection and preparation; (2) application development; and (3) performance testing and statistical analysis. Through a comparative study between four technology solutions/methods to visualize spatio-temporal data, this work has tried to document if there are differences in loading times and efficiency results evaluated based on performance tests. Four time series datasets ranging in sizes were utilized as the test data for the four different techniques, GeoJSON, WMS, D3 and Cesium. The performance time results shows that the WMS technology was the fastest for data loading in terms of the display time and total loading time. Cesium, GeoJSON and D3/TopoJSON were also usable for small datasets but failed the large data tasks. The results regarding animation efficiency were that WMS was the only technology that was able to display all four datasets. Conclutions The overall conclusion is that only one of the four technologies handled all datasets and that was the WMS application. WMS was the fastest method for data loading and had the smallest display times and total loading times, for all four datasets. And for efficiency and animation the conclusion was that WMS was the only technology that was able to display all datasets. What of four open-source-based methods are most optimal for visualizing spatio-temporal data are the contribution of this work. Future work should use more simular data with different sizes and find a better way to take the time measurements.