1SPRS, Vol.34, Part 2W2, “Dynamic and Multi-Dimensional GIS", Bangkok, May 23-25, 2001 
186 
IFLP algorithm is used for converting uncertain multiobjective 
problems into their deterministic forms. Thus, coefficients in the 
objective functions and the constraints' left-hand sides are 
handled as inexact intervals, while linear membership functions 
are assigned to fuzzy goals of the system objectives and the 
right-hand side constraint values. 
The study area was divided into seven subareas with different 
environmental, economic and resource characteristics. The 
planning horizon was 15 years which were further divided into 
two periods (1995 to 2000 and 2001 to 2010). Based on detailed 
investigation of the study system and extensive interaction with 
local authorities and stakeholders, three groups of objectives 
were considered to have top priority, including (1) economic 
return, (2) forest coverage, and (3) water quality (minimization of 
nitrogen, phosphorous, and COD losses) and soil conservation. 
The conceptual expression of the IFMOP model is as follows: 
maximize: 
economic return, 
forest cover, 
minimize: 
soil loss, 
water quality objectives: 
nitrogen loss, 
phosphorous loss, 
COD discharge, 
subject to: 
land availability constraints, 
agricultural production constraints, 
forest-related activity constraints, 
industrial activity constraints, 
tourism-related activity constraints, 
net-cage fish culture constraints, 
lime/brick production constraints, 
water demand/supply constraints, 
soil loss constraints, 
water quality constraints, 
technical constraints. 
The detailed model and solution method was provided in Huang 
et al. (1996). 
Generally, this hybrid inexact-fuzzy optimization approach will 
allow effective identification, quantification, communication, and 
assessment of uncertainties in different forms and the associated 
V«i*6WNo. 
secron 
Acrmrr(i) 
IAfin. N mt 
Figure 3 EXCEL and ORSYS modules used in model-base management subsystem 
pollution risks. The modeling outputs can be directly used for 
generating decision alternatives and providing bases for further 
tradeoff analysis and risk assessment. The computation of 
IFMOP is executed by combining two software tools: MS EXCEL 
and ORSYS. The modeling matrices are established in EXCEL 
worksheets which can be conveniently modified by users in an 
interactive solution process. The ORSYS is capable of dealing 
with large-scale (mix-integer) linear programming problems. It 
can retrieve the information from the EXCEL worksheet and 
export the solutions back to the EXCEL. Figure 3 shows EXCEL 
and ORSYS modules used in the optimization process 
GIS-SUPPORTED WATER QUALITY MANAGEMENT 
SYSTEM 
Geographical information system (GIS) is used for graphically 
analyzing managing knowledge of the study watershed by 
capturing, manipulating, processing, and displaying spatial and 
geo-referenced data. The GIS technology is chosen as a basic 
tool throughout the modeling processes due to its ability to 
clearly expose complex environmental conditions within the 
watershed. GIS is used in three areas throughout the system 
development and implementation process: (i) managing spatial 
and non-spatial database; (ii) linking models; and (iii) providing 
interface between the models and their users. 
Database Management 
The GIS database management subsystem focuses on the 
attributes and data necessary to run the simulation models. The 
major elements include soil properties, climate, topography, 
hydrography, land use, water quality, pollution sources, and 
environmental management practices. GIS is used for entering 
data, comparing data from different sources and formats, 
assessing data availability and quality (e g. accuracy and scale), 
and identifying errors in data (Grossmann et al., 1993). All 
encoded digital data, coverages and model variables in the GIS 
were spatially organized with a consistent resolution and 
coordinate system. Spatial graphic data from different sources 
with different formats are entered into and analyzed by Arc/Info, 
and presented by ArcView. Nonspatial data are stored as 
attribute files. Figure 4 graphically shows an ArcView display for 
data and map. The developed GIS system assures the integraty 
of data and makes it possible for them to be used in the 
modeling processes, and result analysis and presentation. 
Figure 4 GIS system for the Lake Erhai Basin 
Model-base Management 
The GIS system, with its spatial analysis functions, is used to 
connect each individual model to form an integrated modeling 
framework. Especially, the simulation and optimization models 
can be connected to allow the simulated results be input into the 
optimization models and the optimized results be reflected in the 
simulation. These connections are achieved through ArcView