Integration of a GIS and a complex three-dimensional hydrodynamic, sediment and heavy metal transport numerical model

Abstract

The paper presents an integration of a geographic information system (GIS) and a complex three-dimensional hydrodynamic sediment and heavy metal transport numerical model. The integration overcomes the two-dimensional constraint of conventional GIS by supplementing advance tools for three-dimensional, dynamic visualization. The application of a GIS-based interface module together with the three-dimensional, dynamic display tools enhances communication of relationships and trends of hydrodynamic and pollutant transport simulation in both spatial and temporal context, and thus promotes better coastal water quality planning and management. Model functionality includes input data viewing and editing, mesh grid configuration, and result interpretation. The functionality of the GIS-model integrated system is illustrated through a case study on the Pearl River Estuary (PRE).

Introduction

The use of numerical models is the most feasible approach to solving hydrodynamic and pollutant transport problems. However, such complex models generally involve vast amounts of geographically-referenced information, and thus require enormous efforts in preparation of input data as well as in handling of output results for interpretation. Numerous efforts have been made over the decade to integrate hydrodynamic and pollutant transport models with geographic information system (GIS) [1], [2], [3]. Such integrations not only make efficient the modelling pre-processing and post-processing procedures, but also provide the system with spatial data management, analysis, and visualization functionalities.

Although GIS is demonstrated to be an invaluable asset in hydrodynamic and pollutant transport modelling, there is one major constraint in its implementation. Marine environmental data exhibit variations not only in both location and depth, but also through time [4]. It is difficult to establish a rigorous model to represent the ever-changing, three-dimensional nature of marine system in a GIS environment. This explains why most of the current GIS-hydrodynamic/pollutant transport models are limited to two-dimensional capabilities. Since hydrodynamics is intrinsically three-dimensional and dynamic, such models oversimplify the real situations and generate results with limited accuracy. As the fundamental concept of GIS is mapping of information in two-dimensional ground, there is a considerable challenge in unfolding this two dimensionality for applications into the three and four-dimensional world.

This paper describes the integration of a GIS and a complex three-dimensional hydrodynamic sediment and heavy metal transport numerical model. The model was developed at The Hong Kong Polytechnic University [5] to tackle the deteriorating water quality conditions at the Pearl River Estuary (PRE). The model covers the prediction of hydrodynamic patterns and sediment and heavy metal transports for the entire PRE using the finite element method. Such complex computation requires the model to be run on an efficient parallel computing system of ten CPUs using Message Passing Interface (MPI) scripting.

The parallelized system is implemented using a LINUX operating system, in which its user-friendliness is incomparably lower than that of proprietary operating systems such as Windows or MacOS. Both model input and output involve a substantial quantity of spatial data, however, the model is not equipped with any tools to handle spatial analysis and visualization. This constraint marks the necessity to provide the essential graphical tools for preparations and displays of model inputs and outputs; and the integration of the model with a GIS is identified as a plausible approach.

The primary objective of the integration is to build a connection between the model and a GIS. Using input data created in the GIS, numerical simulations can be run and model results can be analyzed in the GIS. Furthermore, to overcome the two-dimensional limitation of conventional GIS, a set of data displaying tools are developed to offer interpretation of information in the third (i.e. depth) and fourth (i.e. time) dimensions.