The development of interactive 3D spring visualization for periodic multidimensional direction time-series data sets

This paper introduces a new approach for visualizing multidimentional weather-direction-related time-series data sets called “3D Spring Model”. Spring Model is designed to visualize pattern behind large time-series weather data set and to clearify seasonal structure in the data. In addition, it supports visibility of seasonal shift and wind direction anomaly by direct comparison betweem successive spring cycles. The visualization contained three data types: (1) Weather parameter (such as windrun, temperature or rainfall etc.), (2) Wind directions and (3) time. We mapped the color to the model in such the way that it comply with human perception using color gradient. Level-Of-Detail scheme is applied and adjustable resulting different pattern time focus for users. Spring Model is highly self-contained for accumulative long term data. It is interactive, flexible and user-friendly. Spring Model is very well-suited to high computing power visualization environment. At the end of the paper, the observation of weather pattern in Nakhon Si Thammarat, Thailand using Spring Model was proposed as the case study to present the model vast applications.