Abstract
This paper describes the integration of ACSL’s temporal data with GIS spatial data. ACSL is a simulation language developed for models described by time-dependent, nonlinear differential equations. The application areas for ACSL are typically chemical process representation, control system design, microprocessor controllers, etc. Models can be derived from block diagrams, mathematical equations, or conventional programming statements. The values of the specified variables are recorded on the data file. These values can be printed or plotted after a run and can also be used as input to the database connected with a GIS. The GIS platform can pre-process the temporal data into the required form and display the result. ACSL’s capabilities can be exemplified by modelling the deposition of sedimentborne radionuclides in the Solway Firth, but can be used in many similar applications.
The process has five phases: writing the ACSL program; running the model with runtime commands; exporting the data to a database; creating the project in a GIS having connections to the database; and, displaying temporal data in the GIS environment.
The ACSL program contains a deterministic physical model describing the radionuclide deposition in a few locations in the study area by differential equations. The state variables are the radionuclide concentrations in parts of the area. The parameters set the interactions between areas - involving transport and redistribution of the contaminated sediments (Allan, 1993; MacKenzie et al., 1994). The program also contains other sections to set the initial conditions of state variables, program controls and to save the variables into the data file.
Runtime commands create the environment for running any associated program and can change the initial conditions and model parameters. These commands also produce output such as printouts and plots.
The export of the data output from ACSL to a database depends on the import tools of the relational database management system. The integration of ACSL generated data has been tested with MGE-PC (a GIS environment supplied by the company Intergraph), which supports, among others, Oracle and the MicroStation Xbase server for standalone configurations. This part also includes creating the external database, which contains the table with temporal data.
"Project creation" in the MGE-PC enviroment is conducted following Intergraph's recommended schema. The data sources involve maps in raster or vector form and atributes in the database tables. The user can generate spatial analyses and queries of topological files based on their relational database information.
Displaying temporal data in the GIS environment has been extended with utilities in MicroStation, which use as input temporal data in the connected database and directly generate coloured maps of the area. The utilities were written using the MicroStation Development Language MDL - a development environment using graphics facilities from MicroStation CAD.
Although the target object is specifically the development of a structure enabling the integration of time-dependent models with GIS and the example used was a model of sedimentborne radionuclide deposition, the goal is a GIS that includes an engine for modelling time-dependent systems and which can operate as a standalone tool.
This research has been sponsored by the Royal Society of London in the frame of the Czech Postdoctoral Fellowship Programme.
References
Allan, R.L. 1993. Distribution, Geochemistry and Geochronology of Sellafield Waste in Contaminated Solway Firth Flood Plain Deposits, Unpublished PhD thesis, University of Glasgow.
MacKenzie, A.B., Scott, R.D., Allan, R.L., Shaban, Y.A.B., Cook, G.T. and Pulford, I.D. 1994. "Sediment Radionuclide Profiles: Implications for Mechanisms of Sellafield Waste Dispersal in the Irish Sea", J. Environ. Radioactivity, 23, 39-69.
Bibliography
Wesseling, C.G., Karssenberg, D., Burrough, P.A. and Van Deursen, W.P.A. 1996. "Integrating Dynamic Environmental Models in GIS: The Development of a Dynamic Modelling Language", Transactions in GIS, l (1), 40-48.