1. Literature Review
Reclamation has been associated with all mine-related activities since the law became effective in 1977. For those areas that were abandoned before 1977, reclamation is even more important. In large areas, with much abandoned land needing to be reclaimed, it is difficult to determine which sites should be reclaimed first. This is so because of the complexity of the associated environmental variables, and the lack of an efficient information collecting system.

Current mine reclamation models often use a number of different data formats from various sources. For example, soils data may come from a soil-mapping agency, vegetation maps from a natural resources agency, and pollution data from yet another source. In addition, the reclamation procedure is often carried out by work being directed toward a succession of "hot spots" that are addressed in a sort of hodge-podge, descending order of priority. The result is that the site is never studied in its entirety. Because of this, less obvious but potentially more damaging areas are often overlooked in the reclamation plan. Modern technologies in information sciences make it possible to fully analyze the problem, and to extract the relevant data necessary to a global view.

Geographic Information System (GIS)

"Environmental databases contain an enormous diversity of types of data, most of which are spatially located either explicitly or implicitly. Consequently the capture, analysis, management, and display of environmental data are all activities that can greatly benefit from the application of GIS" (Townshend, Jr. G. 1991). The use of a GIS opens up a variety of additional analytical avenues, along which a variety of physical models in a GIS framework can be extremely useful.

Land use planning is probably the most obvious and widely cited field to benefit from GIS. Spatial information and GIS are key variables in environmental analysis, where the primary goal is to provide a synoptic view of large spatial data sets. The spatial analytic functionality of GIS consists mainly of the ability to perform deterministic overlays, spatial queries, neighbor analyses, buffer operations, etc. Such functions are of limited use when multiple and conflicting mapped criteria are involved.

Decision-makers faced with a complex spatial problem have multiple, and often conflicting, objectives for its solution. To be acceptable, a solution must reconcile these conflicting goals. "A variety of analytical techniques have been developed to help decision makers solve problems with multiple criteria" (Starr and Zeleny 1977; Cohon 1978; Nijkamp 1979). Consequently, decision-makers have turned to analytical modeling techniques to enhance their decision-making capabilities.

The literature on traditional a GIS contains many definitions. However, when it is applied to the decision-making process, a GIS alone falls far short of providing adequate analytical techniques. The problems include:

  1. A GIS provides a lack of support for analytical modeling;
  2. Many GIS databases provide support for cartographic display only;
  3. Graphic and tabular reporting capabilities are inflexible; and;
  4. There is a lack of versatility in accommodating variations in either the context or the process of spatial decision making.
The shortcomings of a GIS in the decision-making process can be overcome by integrating a GIS with environmental models (Estes et al., 1987; Haddock and Jankowski, 1993; Mitasova, 1993; Vieux, 1991) and with image processing techniques (Ehlers et al., 1991; Estes, 1992; Faust et al., 1991). The integration of a GIS with analytical tools such as these environmental and decision-making models provides the basis for the development of a spatial decision support system (SDSS) (Birkin et al., 1990; Densham and Goodchild, 1989; Nyerges, 1993).


Environmental modeling has a considerable developmental history. Early approaches to the subject explicitly lack spatial dimensionalities because of the limitation of computing resources. With the advent of powerful digital computers and new technologies dealing with complex spatial data, numerical simulation models became feasible.

Many soil erosion and non-point source pollution models have been combined with a GIS to capitalize on the spatial analysis and display capabilities of these new software tools. These data provide regional soil erosion and non-point water quality assessments that were made during the past decade. The six popular GIS-based land surface/subsurface models are:

    1. Universal Soil Loss Equation (USLE) (Wischmeier and Smith 1978)/Revised Universal  Soil Loss Equation (RUSLE) (Renard et al. 1997)
    2. Areal Nonpoint Source Watershed Environment Response Simulation (ANSWERS) (Beasley and Huggins 1982)
    3. Agricultural Non-Point Sources (AGNPS) (Young et al. 1987)
    4. AGNPS is a distributed parameter model developed by Agricultural Research Service (ARS) scientists and engineers. It predicts soil erosion and nutrient transport/loadings from agricultural watersheds for real or hypothetical storms. The AGNPS model was developed for the analysis of nonpoint source pollution from agricultural fields. It estimates the quality of surface water runoff and compares it to the expected quality of other land management strategies. AGNPS is a single event based model, though continuous simulated versions are under development. AGNPS uses a set of modified USLE equations in its erosion component.

    5. Chemical Movement in Layered Soil (CMLS) (Nofziger and Hornsby 1986)
    6. Leaching Estimation and Chemistry Model (LEACHM) (Wagenet and Hutson 1989)
    7. LEACHM is a process based model of water and solute movement, transformations, plant uptake, and chemical reactions in the unsaturated zone.

    8. Water Erosion Prediction Project (WEPP) (Albert et al. 1987).
WEPP is a new generation of soil erosion prediction technology for use in soil and water conservation planning and assessment (Flanagan and Nearing 1995). The WEPP model is based on physical descriptions of rill and interrill erosion processes and sediment transport mechanics. It does not use principles, parameters, or logic from the USLE for predicting erosion. Because it is to a large degree process-based, the model is well suited for studying the effects of environmental system changes on hydrologic and erosion processes, including interactions between climate change, hydrologic response, and sediment generation. WEPP is a continuous simulation model and works primarily on a daily time step in terms of updating system parameters that define the surface conditions for each rainfall. Among these models, the Revised Universal Soil Loss Equation (RUSLE) is the most mathematically simple. Others require many more inputs and, therefore, demand both larger databases and more complex linkages between the GIS and the erosion model. In this study, the intent of the erosion model is to generate maps of impacted areas and to develop relative erosion rates over a longer period of time. It is not our intent to quantify the erosion rates for specific rainfall events. The RUSLE can accomplish our purpose here very well with fewer inputs.

The RUSLE computes the average annual erosion expected on field slopes as

A = R K L S C P [1]

where :

A = The computed spatial average soil loss and temporal average soil loss per unit of area, expressed in the units selected for K and for the period selected for R. In practice, these parameters are usually selected so that A is expressed in ton.acre-1.yr-1.

R = The rainfall-runoff erosivity factor - the rainfall erosion index plus a factor for any significant runoff from snowmelt.

K = The soil erodibility factor - the soil-loss rate per erosion index unit for a specified soil as measured on a standard plot.

L = The slope length factor - the ratio of soil loss from the field slope length to soil loss from a 72.6-ft length under identical conditions.

S = The slope steepness factor - the ratio of soil loss from the field slope gradient to soil loss from a 9% slope under identical conditions.

C = The cover-management factor - the ratio of soil loss from an area with specified cover and management to soil loss from an identical area in tilled continuous fallow.

P = The support practice factor - the ratio of soil loss with support practices such as contouring, strip-cropping, or terracing to soil loss with straight-row farming up and down the slope.

Mine soils developed from mine spoils commonly have a wide range of particle size. The slopes of old spoil piles usually are marked by gullies due to years of uncontrolled erosion. These characteristics raise questions about applicability of available theories and models for estimating runoff and erosion. Recent research addresses the application of RUSLE technology and other models to mine spoils and to reconstructed top-soil (Barfield et al. 1988, Wu et al., 1996). The effects of compaction on erosion are significant in such instances and are treated as an integral part of the subfactor for calculating C. Furthermore, slope steepness effects on soil loss from disturbed lands (McIsaac et al. 1987a) are treated specifically in calculating LS factors. An investigation was made by Wu et al. (1996) to determine whether available erosion models can work for mine soils and can account for gully erosion. An investigation at an abandoned surface mine consisted of measurements of soil and sediment properties, measurements of runoff and erosion, observations of armor by rock fragments on gully floor, and calculations using available theories of sediment transport and slope stability. The results at this site suggest that predictions with erosion model have about the same accuracy as those made for agricultural lands.

Several attempts have been made to combine RULSE modeling with a GIS and to generate regional soil loss assessments. Ventura et al. (1988) used a series of GIS polygon overlays and FORTRAN programs to estimate soil erosion in Dane County, Wisconsin. James and Hewitt (1992) used a series of ARC/INFO coverages and Arc Macro Language programs to build a decision support system for the Black River drainage in Montana.

Image Processing

Remote sensing technology has been applied tremendously for applications to the environment management. The purpose of processing remotely sensed images is to extract useful information from the vest amount of data. Image classification is a major part of any image processing method. Many techniques have been developed to make classification more accurate. Normally, multi-spectral data are used to perform the classification with adjunction of other supporting data. For some applications, such as change detection, multi-temporal data are used to monitor the land changes over the time period.

Several studies have illustrated the role of remote sensing associated with GIS in supplying data and information for assessing water related pollution attributes and for formulating natural resource planning and management strategies. Newell et al. (1992) generated a ranking of nonpoint source water pollution loads in Galveston Bay, Texas, using eight land use categories derived from Landsat Thematic Mapper (TM) imagery incorporated with soil runoff models, rainfall amounts, and water quality parameters. In southwestern Lousiana, Subra and Waters (1993) developed a prototype nonpoint source water pollution model using 15 land cover categories generated from TM data, watershed, hydrology, slope, and soil type data. Nelson and Arnold (1995) conducted their research in a Connecticut watershed using six categories of land use extracted from TM and weighted by their percent of impervious area to produce current and future runoff values. Henderson et al. (1998) used existing land cover data, soil types, and proximity to water to produce potential spatial models in Carmans River watershed in Long Island, New York. The land cover data were subsets of nine categories from National Aeronautics and Space Administration’s Coastal Change Analysis Program (C-CAP) Long Island land cover classification (Henderson et al., 1998), which was created from TM imagery.

Multi-Criteria Decision-Making (MCDM)

Land use management and multi-criteria analysis provide great opportunities for mutual reinforcement. Over the past decade, these two fields have developed largely independently, but a trend towards the exploration of their synergies is now emerging. This is clear from the recent literature on land-use management, spatial analysis, and spatial planning, which increasingly includes references to multi-criteria methodologies and decision analysis. At the same time, a growing share of multi-criteria applications now focus on environmental and land use issues.

Geographic Information System (GIS), Multi-Criteria Decision-Making (MCDM) and other related information technologies have given us a powerful capacity to process a wide variety of social, economic, environmental and physical data. The integration of GIS and Multi-Criteria Analysis (MCA) systems, in particular, has become a central focus for the development of general purpose land use/ environmental planning tools and methods that can assist decision makers (Wright and Buehler, 1993).

Multi-Criteria Analysis (MCA) techniques linked to or integrated within a GIS can provide the user a valuable addition to the standard functionality of GIS (Carver, 1991; Eastman et al. 1995).

The need for developing a platform on which different contributions can be integrated is one of the main reasons for the growing interest in Multi-Criteria Analysis (MCA) for land use management (Nijkamp et al., 1990; Beinat, 1997). MCA has evolved from a mechanism for the selection of the best alternative from a set of competing options, to a range of decision aid techniques. At present, MCA comprises a wide set of tools, but MCA is especially a way of approaching complex decision problems. The development of MCA for spatial problems (Scholten and Stilwell, 1990) and the integration of spatial concepts in Multi-Criteria Analysis imply that MCA now offers a substantial contribution to land use management.

The basic aim of MCDM techniques is " to investigate a number of choice possibilities in the light of multiple criteria and conflicting objectives " (Voogd, 1983). In doing so it is possible to generate compromise alternatives and rankings of alternatives according to their attractiveness (Janssen and Rietveld, 1990).

The basic starting point is the construction of an evaluation matrix, the elements of which reflect the characteristics of the given set of choice alternatives on the basis of a specific set of criteria. The evaluation matrix can be summarized as follows:

S11 …S1j… S1J S = SI1 … SIj … SIJ


Sij is the score of alternative i according to criteria j.

j represents criteria. j = 1 .. J

i represents alternatives. i = 1 .. I

The second element of a MCDM technique model, besides the criteria scores, is the Decision Maker’s (DM) preference. The preferences may be formulated in regard to criterion scores, taking the form of cut-off values ( minimum or maximum threshold) or the desired aspiration levels (Lotfi et al., 1992). They may also be formulated in regard to decision criteria and expressed in a cardinal vector of normalized criterion preference weights W, where

W = (w1, w2, …, wj, …, wJ ). 0 <= wj <= 1 and j = 1..J [3]

The criteria scores and the DM’s preferences are processed using single or multiple aggregation functions. The basic form of the weighted summation techniques can be depicted in the matrix notation.

C1 S11…..…..S1J W1

= * [4]


Ci : appraisal score for alternative i.

The general objective of MCDM is to assist the decision-maker in selecting the best alternative from a number of feasible choice alternatives in the presence of multiple-choice criteria and diverse criterion priorities (Jankowski, 1995). Within the framework of a GIS, multiple-criteria decision (MCD) models have been used increasingly during the past five years as a useful method for spatial decision support.

Spatial Decision Support System (SDSS)

The last twenty years have witnessed the development of various computer-based applications of information systems which have changed the activity patterns and decision modes of spatial actors.

A Decision Support System (DSS) is a system that organizes information intended for use in decision making. It does not simply provide output in the form of reports, nor does it make judgements or decisions for the user. It does, however, support decision-making by presenting information that is especially designed for decision-makers, and that provides new perspectives on the decision-making process that are attractive and understandable.

The definitions of decision support systems rangeX from "interactive computer-based systems that help decision makers utilize data and models to solve unstructured problems" (Gorry and Scott Morton 1971) to "any system that makes some contribution to decision making" (Sprague and Watson 1986). The objective of a decision support system is to improve the procedural rationality in the decision process rather than the substantive rationality of the final decision. The development of a decision support system benefits from a variety of disciplines such as computer science, management science, planing, operation research, and geography (Janssen 1992). The literatures from different discipline gave a slightly different description of decision support system. For example, in computer science, it was described in terms of information processing and according to the various functional elements (Ginzberg and Stohr 1982; Ariav and Ginzberg 1985; McLean and Sol 1986; Sprague and Watson 1986; Keeney et al. 1988; Holtgrefe 1989). In management science and planning, the literature described decision making as a process with various stages. Specially, the division of a decision process in intelligence design and choice (Simon 1960; Newell and Simon 1972), and the concept of a decision process as a cyclical proces (Faludi 1971) can be found in almost all literature on decision support. In the field of geography and management science, the theory on the use of graphics as decision aids can be found (Bertin 1981; Remus 1984; Thfte 1985, 1990; Dickson et al. 1986).

GIS tools have revolutionized the monitoring and management of natural resources. It is not surprising that in recent years considerable interest has been focused on the use of GIS as a decision support system (Eastman et al. 1993a, 1994; Eastman 1997; Carver 1991; Janssen and Rietveld 1990; Honea et al. 1991).

Spatial decision support systems are explicitly designed to provide the user with decision-making tools that will enable him to analyze geographical information in a flexible manner.

Densham(1991) suggested that a SDSS "provides a framework for integration of analytical modeling capabilities, database management systems and graphical display capabilities to improve decision-making processes.... To assist decision makers with complex spatial problems, geo-processing systems must support a decision research process, rather than a more narrowly defined decision-making process, by providing the decision maker with a flexible, problem-solving environment."

In SDSS, the mapping, query, and spatial modeling functions of a GIS provide a capability to display at different scales, and allow preprocessing and data input into specialized models.

The characteristics of a spatial decision support system (SDSS) facilitate a decision research process that can be characterized as iterative, integrative, and participative (Densham, 1991). It is iterative because a set of alternative solutions is generated which the decision-maker evaluates. Insights gained from this evaluation are input to, and used to define, further analyses. It is participative because the decision-maker plays an active role in defining the problem, carrying out the analyses, and evaluating the outcomes. The benefit of participation is integration: value judgments that materially affect the final outcome are made by decision makers who have expert knowledge that must be integrated with the quantitative data in the models and qualitative information.

The major components in a SDSS architecture include a database management system, a model base management system, a display, a report generator, and a user interface. A Database management system (DBMS) is the core of a SDSS. It must be able to store and manipulate locational, topographical, and thematic data types to support cartographic displays, spatial queries, and analytical models. Locational data identify the coordinates at locations, topographical data represent geographic objects, and thematic data provide the attributes of topographical objects. The DBMS must allow the system users to construct and exploit complex spatial relations between all three types of data at a variety of scales, degrees of resolution, and levels of aggregation. Some traditional relational database system found in GIS lack the ability of spatial analysis, therefore additional data models have to be developed. The Environment Science Research Institute (ESRI) has produced the Spatial Database Engine (SDE) technology to solve this problem.

Models can be embedded in a DBMS through macro or script coding. They make it easy to query and manipulate the database. However, they tend to restrict the functions and portability across platforms and operating systems. Instead of storing data, a Model Base Management System (MBMS) stores elements of models that solve a step in an algorithm. Since some elements are common to several algorithms, this approach helps to save large amounts of code. The advantage of using MBMS is a flexibility that facilitates the modification of existing elements and implementations of new algorithms.

Display and report generators provide capabilities to better depict the results derived from models in a SDSS.

The user interface is the environment in which decision-makers interact with the SDSS. Ideally the SDSS should be easy to use. The interface needs to provide two spaces (Densham 1991): objective space and map space. Objective space depicts parameters and solutions of an analytical model, while map space represents the study area and output of the model. These two spaces are linked internally. A change in one space will be reflected in the other space. The better part of this kind of interface is that it allows the user to visualize the processes that underlie the model and to intervene and manipulate the model during the solution process.

Different approaches to spatial decision support systems have been proposed that depend on whether emphasis is placed on the generation of information (analytic tools), on the visualization of data and presentation of documents (media), or on group process (collective cognition) (Shiffer, 1992). Collaborative Decision Support Systems are explicitly designed to facilitate group decision making. In other fields, the effectiveness of decision support may depend critically on the visualization of spatial data and the presentation of documents or other forms of media. In planning, analysis plays an important role, and therefore the analytic tools are appropriate for it.

It is important to notice the difference between a decision support system and decision making system. The former is primarily designed to provide a flexible environment in which users can make their decisions; it is the users who have full control on the evaluation of the variables. That means the users have to provide rules. On the other hand, decision-making systems are usually designed by experts in specific areas who generate rules or regulations that are stored in the system. The user only needs to input the original values of each variable. The system, then takes the inputs and process through the rules by using "if – then" logic, finally generating a result. Decision making systems are sometimes also called expert systems.

Open Development Environment (ODE) and GIS Componentware

Spatial data are usually handled by special software packages such as ArcInfo and ArcView by Environmental Systems Research Institute (ESRI). [Xiao Liu, I think you defined this earlier as Environmental Science Research Institute.]  These packages provide a large variety of functions to deal with spatial data. They are designed for users with GIS background. But unfortunately the user has to learn a lot about the internal languages, such as ArcInfo Macro Language (AML) and ArcView Avenue, built within the software to communicate with the functions or libraries. This situation makes it very hard for a programmer to build flexible custom applications. In Spring 1997, ESRI announced its new product, Open Development Environment (ODE), which provides an environment for UNIX and Windows NT developers to gain access to ArcInfo GIS functionalities without using AML.

"The ARC/INFO ODE is a collection of programmable geographic information system (GIS) objects that lets UNIX and Windows NT developers add GIS capabilities to their applications. The most important method within those objects is the execute method that exposes the ARC,[?] ARCEDIT[?], and ARCPLOT[?] command syntax. The ODE's purpose is to deliver the complete functionality of ARC/INFO through a C interface on both UNIX and Windows NT platforms. It allows developers to build custom applications with ARC/INFO without requiring ARC Macro Language (AML) or interapplication communication (IAC) to do so. The specific objects the ODE supports are the commands and functions available at the ARCEDIT, ARCPLOT, and ARC command lines. Although AML and IAC are not required for building an application with the ODE, the Windows NT version still supports these environments." (ESRI, 1997). ODE provides additional Windows NT functionality via OCXs[?] for use by development environments such as Visual Basic, Delphi and Visual C++.

GIS componentware technology was brought into public use about two of years ago. It is based on Microsoft OLE[?] or OCX[?] architecture and the Dynamic Library Link (DLL) foundation. By wrapping GIS capabilities within one or more components to form Active X or OCX controls, it is possible to embed in programs written in commercial software using common computer languages. The functions encapsulated in OLE or DLL are available to industry standard develop environments such as Visual Basic, Delphi, and Visual C++. This new technology relieves some users from spending huge amounts of money on commercial GIS software for uncomplicated applications. The other advantage is that the users don’t need to bind themselves to a particular GIS product. It is therefore, easy to transfer from one project to another that might adapt different GIS software. The portability of GIS components makes it easier to disseminate the products. GIS componentware has become a part of a powerful toolbox for developers. Produced by ESRI, MapObjects provides basic mapping capabilities such as displays, spatial and attribute queries, new datasets generation, and a printing capability.