The characteristics of environmental impacts vary. Typical parameters to be taken into account in impact prediction and decision-making include:
- nature (positive, negative, direct, indirect, cumulative);
- magnitude (severe, moderate, low);
- extent/location (area/volume covered, distribution);
- timing (during construction, operation, decommissioning, immediate, delayed, rate of change);
- duration (short term, long term, intermittent, continuous);
- likelihood (probability, uncertainty or confidence in the prediction); and
- significance (local, regional, global).
The most obvious impacts are those that are directly related to the proposal, and can be connected (in space and time) to the action that caused them. Typical examples of direct impacts are: loss of wetlands caused by agricultural drainage; destruction of habitat caused by forest clearance; relocation of households caused by reservoir impoundment; increased air particulate emissions caused by operation of a new power station, etc. Indirect or secondary impacts are changes that are usually less obvious, occurring later in time or further away from the impact source. Examples of these types of impacts are: the spread of malaria as a result of drainage schemes that increase standing water and thereby create new vector habitat; bio-accumulation and bio-magnification of contaminants in the food chain through take up of agricultural pesticides; and anxiety, stress and community disruption associated with increased traffic volumes and noise caused by road development.
Cumulative effects, typically, result from the incremental impact of an action when combined with impacts from projects and actions that have been undertaken recently or will be carried out in the near or foreseeable future. These impacts may be individually minor but collectively significant because of their spatial concentration or frequency in time. Cumulative effects can accumulate either incrementally (or additively) or interactively (synergistically), such that the overall effect is larger than the sum of the parts.
Estimating the magnitude of the impact is of primary importance. Typically, it is expressed in terms of relative severity, such as major, moderate or low. Severity, as opposed to size, also takes account of other aspects of impact magnitude, notably whether or not an impact is reversible and the likely rate of recovery.
The spatial extent or zone of impact influence can be predicted for site-specific versus regional occurrences. Depending on the type of impact, the variation in magnitude will need to be estimated; for example, alterations to range or pattern of species or dispersion of air and water pollution plumes. This is much easier for direct impacts but can be attempted for other types of impacts.
Impacts arising from all of the stages of the life cycle of the project should be considered (i.e. during construction, operation and decommissioning). Some impacts will occur immediately, while others may be delayed, sometimes by many years. These impact characteristics should be noted in the EIA report.
Some impacts may be short-term, such as the noise arising from the operation of equipment during construction. Others may be long-term, such as the inundation of land during the building of a reservoir. Certain impacts such as blasting may be intermittent, whereas others, such as electromagnetic fields caused by power lines, may be continuous. Impact magnitude and duration classifications can be cross-referenced; for example, major but short term (less than one year), low but persistent (more than 20 years).
The evaluation of significance at this stage of EIA will depend on the characteristics of the predicted impact and its potential importance for decision-making. Significance is usually attributed in terms of an existing standard or criteria of permissible change, for example as specified in a standard, policy objective or plan. This concept is discussed further later in this topic.
|Impact Type||Adapted from The World Bank (1995)|
Presentation of impact results
When preparing the EIA report a systematic format should be used to present information about impacts to decision-makers. The organisation and display of information is an integral part of certain predictive methodologies, such as the Battelle Environmental Evaluation System. In other cases, this framework may need to be designed separately. One possible format is an impact characteristic summary table above. Finally, reference should be made to the confidence limits in impact data, in probabilistic terms for quantitative judgements (e.g. 95 per cent) or in relative terms for qualitative judgements (reasonably high, best estimate, etc.).
Methods for predicting the characteristics of impacts include:
- â€˜best estimateâ€™ professional judgement;
- quantitative mathematical models;
- experiments and physical models; and
- case studies as analogues or points of reference.
These are discussed in general below. Specific examples of the use of predictive techniques can be found in the references listed at the end of this topic.
As noted earlier, all methods of analysis involve professional judgement and the use of advanced tools and models will require expert knowledge. Sole reliance on â€˜best estimateâ€™ professional judgement may be unavoidable when there is a lack of data to support more rigorous analyses or there is a lack of predictive methodology (as in the analysis of certain social impacts).
Examples include the prediction of the effect of a water supply proposal on:
- the activities of women or community interaction; and
- the loss of a communal place or sacred site.
Such predictions should be made by specialists, who are familiar with the type of proposal, the geographic region and/or similar cases that are analogous to the situation. Where professional judgement is used without also employing other methods, the judgement and values of the specialist concerned may be open to challenge. Peer review and the use of agreed concepts and frameworks can be useful to corroborate findings.
Quantitative mathematical models
Quantitative models express cause-effect relationships as mathematical functions, derived from deterministic or probabilistic relationships. A number of such models are used in EIA to predict certain types of impacts, for example, on air, water, soil and habitat. More complex computer-based simulations are data demanding and often their use in EIA requires certain simplifying assumptions to be made.
The choice and use of quantitative models for impact prediction should be suited to the particular cause-effect relationship being studied; for example, transport and fate of oil spills, sediment loadings and fish growth and pesticide pollution of groundwater aquifers. Attention also needs to be given to the consistency, reliability and adaptability of models. Usually operational changes are made to the input conditions for the model to see how the outputs are affected. For instance, differences in air pollution can be calculated by changing the height of a stack or the rate of output of emissions.
Examples of the use of quantitative models include:
- air dispersion models to predict emissions and pollution concentrations at various locations resulting from the operation of a coal-fired power plant;
- hydrological models to predict changes in the flow regime of rivers resulting from the construction of a reservoir; and
- ecological models to predict changes in aquatic biota (e.g. benthos, fish) resulting from discharge of toxic substances.
Although traditionally this type of analysis has been carried out for physical impacts, there is increasing use of mathematical models to analyse biological, social/demographic and economic impacts.
When interpreting the results of quantitative mathematical models it should be remembered that all models are simplifications of the real world. They require the specialist to make a number of assumptions in both their development and their use. If these assumptions are inappropriate then there can be significant implications for the accuracy and usefulness of the output data. EIA project managers should ask all specialists carrying out mathematical analyses to clearly state the assumptions inherent in the use of their models, together with any qualifications to be placed on the results.
Experiments and physical models
Experiments and scale models can be used to test and analyse the effects of project-related activities and the effectiveness of proposed mitigation techniques. These methods have not been used extensively in impact prediction. However, they can be appropriate, depending upon the nature of the impact and the resources available, and providing certain cautions are remembered. When using the results of experiments or models, note that unpredicted outcomes can occur when the data are â€˜scaled upâ€™ to life size.
Experiments can be undertaken directly in the field or under laboratory conditions. Examples of their use include:
- the exposure of fish in a laboratory to concentrations of pollutants to determine mortality levels; and
- field trials of the effectiveness of different methods of erosion control.
Physical models can be built to predict the behaviour and effect of the actual project on the environment. For example, a physical model could be used to simulate changes to patterns of sand or sediment deposition resulting from port and harbour works.
Reviewing case studies of projects in similar environments can inform and assist impact prediction and analysis. Comparisons will be especially helpful if impact monitoring and auditing data are available. Otherwise, the results obtained by a comparable use of EIA methodology should be consulted. Sometimes, relevant case material will not be readily accessible or available. In that event, there is a large body of general information on the impact â€˜footprintsâ€™ of major types of projects, such as dams, roads, airports and power stations. However, this should be read with care as to its source and provenance.
Uncertainty is a pervasive issue at all stages of the EIA process but is especially important for impact prediction. Put simply, uncertainty is a state of relative knowledge or ignorance. Where cause-effect relationships are â€˜knownâ€™ and understood, however imperfectly, impacts can be forecast (or at least described). Certain impacts are unknown until they occur; for example, ozone depletion caused by release of CFCs and inter-species transmission of the human variant of Bovine Spongiform Encephalopathy (BSE) or â€˜mad cowâ€™ disease.
Sources of uncertainty in impact prediction include:
- scientific uncertainty â€“ limited understanding of an ecosystem (or community) and the processes that govern change;
- data uncertainty â€“ restrictions introduced by incomplete or noncomparable information, or by insufficient measurement techniques; and
- policy uncertainty â€“ unclear or disputed objectives, standards or guidelines for managing potential hazards and effects.
There are a number of approaches that can be used to address uncertainty in impact prediction, including:
- â€˜bestâ€™ and â€˜worstâ€™ case prediction to illustrate the spread of uncertainty;
- attaching confidence limits to impact predictions; and
- sensitivity analysis to determine the effect of small changes in impact magnitude.
The relationship between impact, size and severity may not be linear. Small changes in impact magnitude may cause larger than expected increases or decreases in the severity of environmental change. Where necessary, an assessment should be made of the effect that small changes in the magnitude of the impact (say less than 10 per cent) have on the environment, particularly if significant or valued resources are potentially affected. This is referred to as a sensitivity analysis.
A broader range of impacts and interrelationships are now routinely integrated into EIA. These include the social, economic and health aspects of environmental change. In comparison to biophysical impacts, less experience has been gained in analysing these and other non-biophysical impacts.