Table 2. We define three categories of scenarios: single-scale, loosely linked multiscale, and tightly coupled cross-scale. Each of these has specific advantages and disadvantages, and the type most suited to a particular scenario exercise will depend on the objectives of the exercise.

Single scale Multiscale
Loosely linked Tightly coupled
Number of focal scales        1 At least 2 At least 2
Consistency of storylines across scales Not relevant Storylines usually differ and are inconsistent across scales Storylines have a high level of consistency across scales, and there is an explicit focus on downscaling and/or upscaling.
Consideration of drivers at other scales Exogenous drivers from other scales included to the extent that they are relevant to the focal scale Exogenous drivers and constraints from higher and lower scales are included in a similar way to single-scale scenarios. The set of scenarios is usually constructed within a common broad conceptual framework and will incorporate similar types of drivers at different scales. Exogenous drivers and constraints from higher and lower scales are included via downscaling and upscaling procedures.
Consideration of feedbacks between scales Not considered May or may not be considered Explicit linkages between scales and incorporation of feedbacks
Main advantages Simple; no distraction by concerns at other scales Allows stakeholders at each scale to frame the issues that are important to them from their specific perspective Allows for consideration of feedbacks between scales and evaluation of how an issue plays out at different scales
Main disadvantages Important feedbacks between scales may be missed or important externalities at other scales may be overlooked. Scenario outcomes at different scales or different places are not directly comparable. Very costly; may lose credibility because stakeholders at, especially, lower scales may not have much latitude to define the issues to be considered
Example Mont Fleur Scenarios (Kahane 1992) Southern African MA scenarios (Biggs et al. 2004) MedAction scenarios (Kok et al. 2006a,b)