The analysis of multivariate time series is a fundamental task that arises in numerous ecological and environmental applications. Scientists often need to understand how a variable changes, and how its relationships with other variables evolve over time. Such associations exhibit diurnal and seasonal patterns that need to be discovered and characterized. Line graphs are widely used to visualize phenomena that change over time, by plotting one or more variables as trend lines. However, line graphs can suffer from over-plotting when visualizing long time-series, or when depicting several variables in the same space. Furthermore, it is often difficult for the viewer to perceive the relationship between multiple trend lines when their association is not linear.

In this talk, we introduce Tempo, and interactive visualization tool for the analysis of time series. Tempo employs a combination of scatterplots and line graphs to enable the analysis of temporal data. It provides faceted views within an interactive interface to facilitate the discovery of patterns in paired, multidimensional time series. Tempo is web-based, scalable, and does not require any software packages beyond a web browser. We demonstrate the tool with case study involving the analysis of ecological time series to uncover relationships between spectral reflectance measurement and biosphere-atmosphere fluxes.