Détermination des priorités à l'échelle du bassin, évaluation des risques à l'échelle locale : propositions d'approches

Prioritising sediment sites for management at the catchment scale involves the balancing of several kinds of risk and management objectives. At this spatial scale, management objectives are globally threefold: (i) meeting regulatory criteria, or political goals, (ii) maintaining (or restoring) economical viability, and (iii) ensuring environmental quality. Although the corresponding drivers differ for these three categories of objectives, risk indicators at this spatial scale tend to overlap, or to display logical connections. For this reason, and because these management objectives are indeed not independent when setting management actions, a general prioritisation approach can be proposed at the catchment scale, using a pre-defined set of criteria. The relevant criteria for setting priorities at catchment scale must account for the location of a sediment site within the catchment, its potential energy, its quality, both in absolute terms and relative to surrounding sites, and the benefits (and risks) of potential management actions. The criterion energy cannot simply be represented by the slope or the relative elevation within the catchment, because some infrastructure, such as dams, will lower the risk of transport downstream. Thus, basin-scale prioritisation will require the development of criteria that take into account gradients with a term of directionality, as well as erosive potential. Nevertheless, the latter criterion expected benefit could be viewed as result of the combination of other criteria, in particular location and sediment quality. As an initial test of these concepts, our working group has selected two classification approaches, namely score ordination and fuzzy logic; a test of both approaches on a common dataset will be presented and discussed in light of the selection of relevant parameters. It will as well provide an overview of the strengths and limitations of the two selected approaches according to robustness, methodology (related to the ease of use), sensitivity, appropriateness (referring to the range of conditions), and transparency.