3.3.3 Baseline Studies
Identification, estimation and comparison of likely impacts from the feasible alternatives are often accepted as the principal tasks associated with an EIA study. While screening and scoping components of EIA process may identify the feasible alternatives and the likely impacts of significance against given environmental settings, the pre-project environmental setting itself require to be properly described through repetitive sampling so that the spatial and temporal variability of measured attributes can be statistically defined. The environmental inventory so prepared may then be used as the reference data against which the potential impacts of a proposed action on the environment can be evaluated (Canter 1977). The results of scoping exercise should, inter-alia, reveal the data and information needs of the decision-maker involved in project planning. Baseline study programmes may be chalked out only after taking into consideration such data and information needs for prediction, assessment and mitigation of impacts. A baseline study may thus be defined as a study undertaken to give a satisfactory 'description of conditions existing at a point in time against which subsequent changes can be detected through monitoring' (Hirsch 1980). Baseline studies provide the before-project records whilst monitoring gives the after-project measurements from which changes over space and time can be assessed (Beanlands 1988). However, the environmental baseline should not be considered static and the natural variability of the environmental attributes must be respected. Glasson et al. (1994) proposed an extension of the commonly held view on baseline studies by stating that "Description of environmental baseline includes the establishment of both the present and future state of the environment, in the absence of the project, taking into account changes resulting from natural events and from other human activities". In order to be an effective component of 'EIA as a planning tool' the baseline study should be modified to include preparation of a pre-project environmental inventory so as to suitably describe the supportive and assimilative capacity of the environment. It is also important to understand the natural interaction between the various environmental elements and the functional relationships involved.

A preliminary list of significant environmental attributes can be prepared using the following approaches (Canter 1996):-
1. Use of pertinent agency guidelines or regulations;
2. Use of professional knowledge regarding the anticipated impacts of similar projects;
3. Review of other recent EIA reports on similar projects; and,
4. Use of list of factors in EIA methodologies.

Baseline conditions can be characterised only after proper identification of ecosystem components. But it should also be recognised that establishment of a comprehensive, long-term monitoring programmes for all ecosystem components is virtually impossible. In order to rationalise the EIA process, it is therefore important that preparation of a straightforward environmental inventory be complemented by focussing on valued ecosystem components (VECs) and by identifying key biological processes at the beginning of the EIA process (Kennedy and Ross 1992, Treweek 1995).

3.3.4 Impact Prediction and Evaluation
3.3.4.1 Prediction of Environmental Impacts
Many EIA scholars, including Dooley (1979) and Lee (1982), have emphasised that the term "impact" refers to the effects of a proposed human actively on both "ecosystem" and "human society". As has been mentioned before this research focusses on ecosystem impacts.

Once the likely significant impacts are identified and feasible alternatives are selected for consideration the next component of the EIA process involves prediction of impact on all the significant environmental attributes at the development, operation, closure and post-closure stages of the proposed project and its various alternatives. Depending on the nature and type of impacts and on the project characteristics, the 'error of estimate' of impact forecasting can be significantly reduced if the following conditions are fulfilled.

1. Existence of sufficient predictive models and site-specific data to support a quantitative assessment of environmental impacts.
2. Possibility of utilising a quantitative threshold (e.g., A standard or an unambiguous criteria) to distinguish between significant and insignificant impacts.
3. Availability of quantitative/statistical methodologies for objectively describing levels of impacts.
4. Availability of reports of EIA reports of similar projects located preferably in a comparable site.

The accuracy of impact prediction increases when the need for subjective scoring is minimal or absent (USEPA 1993). Usually prediction of impacts on the various environmental attributes is made by separate specialist groups within the EIA study team. Techniques for impact prediction on various environmental parameters are many and they are often presented in EIA texts (e.g., Canter 1977, Rau and Wooten 1980, Heer and Hagerty 1977, Jain et al. 1977, Morris and Therivel 1994, Canter 1996, Westman 1985, Jain et al. 1993 etc.). Methods used for impact prediction are based on engineering, natural science and social science methods (Ortolano and Shepherd 1995). Since the level of understanding and accuracy in prediction varies with the directness of effect, many project-induced effects cannot be predicted accurately in complex environments (Sadler 1988, Berkes 1988). Technical specialists, therefore, often rely heavily on professional judgement to predict environmental impacts leading to situations where forecasts become so vague that they cannot be validated (Leon 1993, Culhane 1987, Bisset 1984, Beanlands and Duinker 1982). The intrinsic problems associated with prediction of impacts within reasonable degree of accuracy make impact prediction 'technically most difficult and challenging activity' (Canter 1996). Impact prediction often requires technically demanding mathematical models. However models are often presented as 'black-boxes', containing little information about the model arguments. This makes the bases for impact prediction unclear and the inherent errors become untraceable.

3.3.4.2 Impact Evaluation
In order to facilitate decision making the predicted impacts on the various environmental attributes must be assessed in terms of significance and magnitude (of change in overall environmental quality). According to Westman (1985), 'assessment' refers to 'analysis' and 'evaluation' of impacts. While 'analysis', inter-alia, includes impact prediction, 'evaluation' refers to determination of significance of predicted impacts against the total quality of the affected environment. This significance determination invariably involves subjective or normative evaluation. The main challenge of environmental impact assessment resides in analysing and assessing the likely impacts within acceptable level of errors (Julien 1995).

Objective of impact evaluation is to assess the value (cost) of quantitative and qualitative changes in environmental attributes. In the conventional sense impact evaluation provides a tool for a trade-off between choice alternatives with different environmental impacts (Nijkamp 1980). This research has already established that the goal of project decision making should be 'impact compensation' and not 'impact minimisation' and therefore 'impact evaluation' should reveal the complete impact compensation possibilities for all feasible alternatives in order to be an acceptable aid in the overall decision making process.

Over the years many attempts have been made to rate industrial projects on the basis of an overall index. Such indexes are calculated by using algorithms which combine impact predictions through subjective or normative judgements. A pre-set procedure may be used to assess the important feasible alternatives and rank them on the basis of likely attainment of environmental objectives.

Impact evaluation attempts to present a 'synthesis' of the predicted significant impacts using a commensurate scale. Completeness, consistency and pertinence of environmental information may be ensured through development of appropriate protocols. Impact evaluation and impact prediction are essentially complementary in nature. Impact evaluation attempts to generate information through analysis of predicted impacts. The generated information is then evaluated by organising and comparing (Julien 1995). Before comparison the predicted significant impacts are summarised or aggregated into overall impacts. Care must be taken to ensure that all the significant impacts associated with a proposed project are fully and adequately taken into consideration. Numerous methodologies have been developed for impact summarisation and aggregation. However, the success of impact evaluation exercise often depends on the expertise and attitude of the evaluators.

Estimation of overall impacts through aggregation is not flawless. As pointed out by Westman (1985), while overall rating definitely facilitates comparison of various alternatives, it often hides the judgement used in the calculation. The decision-maker is, thus, denied access to the thought process.