Visual Analytics on Spatial Time Series for Environmental Data
Mithileysh Sathiyanarayanan1, Vijayakumar Varadarajan2, Pradeep .K .V3

1Mithileysh Sathiyanarayanan, City University of London, U.K.
2Vijayakumar Varadarajan, VIT University, Chennai (Tamil Nadu), India.
3Pradeep .K .V, VIT University, Chennai (Tamil Nadu), India.
Manuscript received on 20 June 2019 | Revised Manuscript received on 11 July 2019 | Manuscript Published on 17 July 2019 | PP: 1173-1181 | Volume-8 Issue-1C2 May 2019 | Retrieval Number: A12080581C219/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (

Abstract: Environemental science is a field of growing and vital importance utlizing cutting-edge research methods and techniques. In this paper we present a review of the visualisation techniques for spatial time series used in environmental research. Specifically, we focus on the following sub-domains in the field: resource depletion monitoring, climate change, environmental pollution and weather forecasting. These domains present their own idiosyncrasies in relation to data integrity and availability, answers to analytical questions sought, and therefore analytical approaches to visualisation when dealing with datasets referenced both spatially and temporally. We show that when data quality is poor (as in deforestation studies) most of the analytical processes target the data pre-processing stage of the visual analytics framework while the choice of visualization format aims for simplicity and visual impact via choropleth maps. When data quantity and quality are less of an issue (climate change, pollution) more sophisticated techniques are applied, by slicing the multi-dimensional dataset into more manageable parts. Typically, with time and space as references and multiple attributes considered, the visual analytical approach broadly splits two ways: either (1) considering the spatial variation of time series (via diagram maps and interactively linked displays) or (2) considering the temporal variation of spatial distributions (via the use of small multiples” and map animation). These visualisation techniques are an unvaluable tool to study the domain of interest. They help analyse attributes covariance, detect trends and anomalies, or more generally describe spatial, temporal and thematic relationships in the attributes-space. Finally, when data complexity becomes itself a challenge as in weather forecasting, in addition to traditional methods, more novel and experimental visualisation techniques are deployed to try and capture both spatial and temporal behaviour of given attributes in a single framework (3). 3D visualisations, space-time cubes, and more sophisticated approaches are here called upon to analyse model uncertainty and perform complex model comparison. Visual analytics therefore offers an extensive an evolving framework to support environmental research studies from beginning to end.
Keywords: Spatial Time Series, Visualisation Methods, Visual Analytics, Weather, Pollution, Climate Change, Deforestation.
Scope of the Article: Visual Analytics