4.2 Scoring: Distance to target/reference and scoring transformations

An important step in turning parameters into indicators is describing the meaning of particular values or ranges of values from an educational or decision-making perspective. For example, surface water temperature is a parameter that can be reported daily or annually, but if reported on its own, may not be overly meaningful. When water temperatures are compared with temperatures important for the salmonid life cycle, then water temperature can be reported as an indicator of condition relative to the needs of fish, this provides a more meaningful context in which to interpret indicator status and trends. A creek with a temperature of 20oC may be fine for recreational use and may support certain fish and wildlife species; however, salmon eggs and fry will be stressed at this temperature, thus the indicator score relative to salmonids may be low for this temperature.

Each indicator status value (or trend) was compared to a reference or standard value (Figure 4.1), and the comparison was used to generate a score. Although it is important to pick a reference value that is meaningful for decision-making, it is just as important to make the choice transparent so that the reference value can be changed in the future if warranted by changes in knowledge, goals or assumptions.

Figure 4.1. Example of parameter comparison to goal/standard. This comparison is used to evaluate the indicator status and trend.

We chose reference or target conditions specific to the indicator using best available science, goals expressed by stakeholder organizations, and professional opinion. These are all mutable choices and can be regarded as proposals for how indicators can be evaluated.

Figure 4.2. Relationship between ecosystem or community condition and something affecting that condition (driving variable)

A very important benefit of taking this step is that scores can be combined across very different indicators (e.g., water temperature and fish tissue mercury concentrations), whereas otherwise this would not be possible. Because all indicator conditions were quantitatively compared to a target, they were all normalized to the same scale.  distance to target. Once the normalization takes place, the new values, ranging from 0 to 100, mean the same thing and can therefore be combined.

Because environmental and socio-economic processes and conditions rarely respond to influences in a linear fashion, evaluating indicators relative to reference conditions must take into account these non-linear responses (Figure 4.2). For example, evaluation of water temperature used a non-linear function (section 3.1.2).

Indicator metrics were quantified in their raw or native units (e.g., oC or tons C sequestered), and evaluated on the basis of their separation from the target condition. This target or reference condition is sometimes called the “ideal point” (Malczewski, 1999). The ideal point method was first introduced in the late 1950s and expanded by Milan Zeleny in the 1970s (Pomerol and Barba-Romero 2000). Zeleny (1982) operationalized the measurement of closeness with

di = fi* – fi (xji)

Where didi is the distance of attribute state xjixji to the ideal value fi*fi*, i indicates the attribute and j indicates the objective. For the Report Card, indicator distances from target were calculated in their native units and converted to a common scale (0-100) to be compared among disparate indicators, or to be aggregated into composite indices. The common scale conversion was relative to a threshold or objective specific to each indicator and was based on the appropriate linear or non-linear rate of change relationship. For example, there is a linear rate of increase in carbon sequestration with area of vegetative cover, but non-linear effects of temperature on salmonid species.