Matrices

The major use of matrices is to indicate cause and effect by listing activities along the horizontal axis and environmental parameters along the vertical axis. In this way the impacts of both individual components of projects as well as major alternatives can be compared. The simplest matrices use a single mark to show whether an impact is predicted or not. However it is easy to increase the information level by changing the size of the mark to indicate scale, or by using a variety of symbols to indicate different attributes of the impact. An example of a matrix is given as Table 2. The choice of symbols in this example enables the reader to see at a glance whether or not there was an impact and, if so, whether the impact was beneficial or detrimental, temporary or permanent. Figure 8 is another example of a matrix, in this case used to clearly indicate the importance of a range of wetland values.

ICOLD has prepared a large and comprehensive matrix for use in EIAs for dams. The system of symbols for each box shows: whether the impact is beneficial or detrimental; the scale of the impact; the probability of occurrence; the time-scale of occurrence; and, whether the design has taken the impact into account, (ICOLD, 1980). This comprehensive approach, however, makes the final output rather difficult to use and a maximum of three criteria is recommended per impact to maintain clarity. Ahmad and Sammy (1985) suggest that the most important criteria are: magnitude, or degree of change; geographical extent; significance; and, special sensitivity. "Significance" could be further sub-divided to indicate why an impact is significant. For example, it may be because of irreversibility, economic vulnerability, a threat to rare species etc. "Special sensitivity" refers to locally important issues. A series of matrices at all stages of the EIA process can be a particularly effective way of presenting information. Each matrix may be used to compare options rated against a few criteria at a time.

The greatest drawback of matrices are that they can only effectively illustrate primary impacts. Network diagrams, described below, are a useful and complementary form of illustration to matrices as their main purpose is to illustrate higher order impacts and to indicate how impacts are inter-related.

Matrices help to choose between alternatives by consensus. One method is to make pair-wise comparisons. It provides a simple way for a group of people to compare a large number of options and reduce them to a few choices. First a matrix is drawn with all options listed both horizontally and vertically. Each option is then compared with every other one and a score of 1 assigned to the preferred option or 0.5 to both options if no preference is agreed.

TABLE 1 Results sheet for assessing the ICID check-list

 Project name/location: Assessment: 1st/2nd/ Assessor's name/posit/on: Date:

 For each environmental effect place a cross (X) in one of the columns Positive impact very likely Positive impact possible No impact Negative impact possible Negative impact very likely No judgement possible at present Comments A B C D E F Hydrology 1-1 Low flow regime 1-2 Flood regime 1-3 Operation of dams 1-4 Fall of water table 1-5 Rise of water table Pollution 2-1 Solute dispersion 2-2 Toxic substances 2-3 Organic pollution 2-4 Anaerobic effects 2-5 Gas emissions Soils 3-1 Soil salinity 3-2 Soil properties 3-3 Saline groundwater 3-4 Saline drainage 3-5 Saline intrusion Sediments 4-1 Local erosion 4-2 Hinterland effect 4-3 River morpholoqy 4-4 Channel regime 4-5 Sedimentation 4-6 Estuary erosion Ecology 5-1 Project lands 5-2 Water bodies 5-3 Surrounding area 5-4 Valleys & shores 5-5 Wetlands & plains 5-6 Rare species 5-7 Animal migration 5-8 Natural industry Socio-economic 6-1 Population change 6-2 Income amenity 6-3 Human migration 6-4 Resettlement 6-5 Women's role 6-6 Minority groups 6-7 Sites of value 6-8 Regional effects 6-9 User involvement 6-10 Recreation Health 7-1 Water & sanitation 7-2 Habitation 7-3 Health services 7-4 Nutrition 7-5 Relocation effect 7-6 Disease ecology 7-7 Disease hosts 7-8 Disease control 7-9 Other hazards Imbalances 8-1 Pests & weeds 8-2 Animal diseases 8-3 Aquatic weeds 8-4 Structural damage 8-5 Animal imbalances Number of crosses (Total = 53)

TABLE 2 Ultimate net environmental impact assessment at a glance, Feitsui reservoir

 Features likely to be affected Roads and trails Colony construction Blasting operation Borrowing of materials Importing of labour Dam construction Canal construction Evacuation and rehabilitation Soil conservation and landscaping Reservoir filling Irrigation Hydro-power generation Forestry/Vegetation -1P +2P -1T -1P -1P -1P +4P -3P +3P + 1P Birds -2T -2T -1T +3P +4P +2P Fisheries -1T +4P +2P Other wildlife/land animals -1P -1T -1T -1T -1T -1T -1P +2P +3P +2P Sedimentation/erosion -1T -1T -2T +2P +2P +3P -1P -1P Floods -1P -1P +1P +3P Historical/cultural monuments +2P -2P Communications +3P +2P +1P +2P -1P +2P Land/area development -2P +2P +2P +2P -2P +2P +2P +4P +3P Agriculture +2P +1P -1P -1P -1P +2P -1P +4P +3P Food production +2P +1P . -2P -1P -1P +2P -1P +4P +3P Public revenue/income +2P +2P +3T +2P -2P +2P -2P +4P +3P Drinking water +1P -1T -1T +4P +3P +2P Water quality -1T -1T -2T -2T -1P +1P Air quality -1T -1T -1T -1T +1P +2P +1P Climate +1P +2P +1P Groundwater table +2P +2P Industrialization +2P +1P +3T +2T +2P +3P +3P Housing +2P +1P +1T +2P -2P +1P +1P Employment/training + 1T + 1T +4T +2T +2P +2P +2P Health and safety -1T -1T -2T -1T -1T -2T +2p +2P +2P Scenic views and vistas +1P +2P -1P -2T +2P +2P +3P +4P +2P +2P Tourism +2P +2P +3P +3P +1P +2P

Notes: Likely effect is symbolized as follows:

 Mild Considerable High Very high Beneficial +1 +2 +3 +4 Detrimental -1 -2 -3 -4

T = temporary effect; P = permanent effect

An example of such a matrix is given as Table 3. As can be seen, Z is the preferred option.

A number of methods have been developed to compare impacts by applying values to them. The relative importance of impacts, eg wetlands loss versus rare species loss, or the relative importance of criteria, e.g. economic vulnerability versus probability of occurrence, will depend on the local environment and priorities. Ranking, and therefore implicitly value, can be determined by using the pair-wise comparison technique described above, except that, rather than comparing options, criteria are compared instead. This can enable a series of weightings to be developed which will be entirely site-specific and dependent upon the subjective choices of those participating in the group which develops the weightings.

TABLE 3: Example of pair-wise comparison

 Compare alternative With alternative Sum W X Y Z W - 0 0 0.5 0.5 X 1 - 1 0 2 Y 1 0 - 0 1 Z 0.5 1 1 - 2.5

A simple example would be to develop weightings for environmental versus economic acceptability. Thus, in the example illustrated in Figure 2, weightings would have to be developed to determine the preference for either option B or option C. Is more weight to be given to environmental or economic criteria?

Reducing information about impacts to a single number should be avoided as it obscures understanding and disguises the subjective nature of the analysis. However, it can be useful to compare, for example, the degree to which different mitigating options are effective in managing water quality.