Representing Time and Space in GIS
Jon Goodall, CRWR
Table of Contents
Storing data describing the hydrologic environment requires temporal as well as geospatial referencing of data. In dynamic systems like a flood event or pollutant transport, it is impossible to understand the system without temporal context. Commercial GIS systems have traditionally focused on geospatial and not temporal referencing of data, limiting the usefulness of GIS for visualizing dynamic hydrologic events. Recent research at the Center for Research in Water Resources has focused on designing a set of spatiotemporal data types capable of providing the foundation for a temporal GIS. Just as the point, line, polygon, raster, and tin provide the foundation for geospatial data representation, our goal is to develop extend these spatial data types to describe spatiotemporal phenomena.
Much of the thought driving this project is related to storing information currently collected or modeled by the South Florida Water Management District. Our approach was to use South Florida as an case study, but to generalize the concepts necessary for South Florida into a set of spatiotemporal data structures useful for other projects. The result was three spatiotemporal data types called "Attribute Series", "Feature Series", and "Raster Series" presented with a generic aspatial data type called "Time Series" in Figure 1.
Figure 1 - Time Series and three spatio-temporal data representations.
Time Series are a collection of time, value pairs. Time Series are not directly georeferenced, but can be indirectly georeferenced through a many-to-many relationship (many time series records can be related to many spatial features). In Figure 1, Time Series is disconnected from the other four features to indicate that it is not a spatiotemporal data structure, but instead a link to the "outside world" of hydrologic time series modeling and analysis.
Attribute Series are a collection of time, value pairs that are related to one spatial feature. Each time, value pair is related to one spatial features. The name Attribute Series comes from the idea that we are describing a feature with a dynamic attribute value - a particular attribute of that feature, whether it be flow, dissolved oxygen concentration, or precipitation, varies with time. Attributes series is identical to the current Arc Hydro TimeSeries component where there are two tables: one to store the time, value records, and one to stores metadata describe the type of time series being measured (see the Chapter 7 of the Arc Hydro book for more detail).
Feature Series are a collection of shape, time pairs. A Feature Series is a collection of features indexed by time. Each feature in a feature series exists for only a period of time, making Feature Series an ideal structure for representing a series of flood inundation polygons. Feature Series could also be used to represent the movement of particles through the environment. In this case, the Feature Series would be a set of points, each valid for some instant in time.
Finally, Raster Series are a collection of rasters indexed by time. Each raster is a "snapshot" of the environment at some instant in time. Grouping a series of rasters can describe how the environment changes over time. Raster Series is useful for describing the dynamics of spatial continuous phenomena, like ponded depth in the Everglades, or rainfall measured by NEXRAD.
For more information about the conceptual framework for representing time and space in GIS, please see:
Poster illustrating new conceptual model for spatiotemporal data types in ArcGIS: time series.pdf
Goodall, J.L., D.R. Maidment, and J. Sorenson. 2004. Representation of Spatial and Temporal Data in ArcGIS, AWRA GIS and Water Resources III Conference, Nashville, TN.
Primary contact for representing space and time in ArcGIS:
Jon Goodall
University of Texas at Austin
Center for Research in Water Resources
Austin, Texas 78712
e-mail:
goodalljl@mail.utexas.edu
Phone: (512) 471-0110
Fax: (512) 471-0072
These materials may be used for study, research, and education, but please credit the authors and the Center for Research in Water Resources, The University of Texas at Austin. All commercial rights reserved. Copyright 2004 Center for Research in Water Resources.