**2.2 The SCALES model characteristics**

SCALES is displayed as a regional scale applicable model keeping high precision and high quality of information at local scale. This tool allows us to produce in a short time a diagnostic of erosion hazard using high resolution data coming from accurate data. This diagnostic is specific to arable lands. It cannot be proposed in context of "natural" vegetation such as woods or forests or for urbanized areas. SCALES is also designed to be accessible to a wide range of companies dealing with questions of environmental relevance. Furthermore, this model also displays the possibility to aggregate hazard data with administrative or hydrological units like municipality and elementary catchment, in order to adapt the diagnostic to intervention scales of local land managers.

## **2.2.1 Basic statements**

228 Soil Erosion Studies

After having shown the operational capability of SCALES at the scale of the Calvados department (Le Gouée et al., 2008) which represents 5,500 km² (Fig. 1), we then focused our efforts on adapting this model to produce a diagnosis of the erosion hazard at seasonal and monthly scales (Stepkow, 2008). That approach enabled us most recently to offer a prospective insight into the effects of climate change in the distant future (scenario A1B of the IPCC for 2100) on the evolution of soil erosion hazard in Normandy (Goulet, 2010).

Fig. 1. The regional council of Basse-Normandie. A: Catchment of the Branche. B: Catchment

SCALES provides a mapping of soil erosion hazard which offer local land managers spatialized data at regional scale while having high accuracy on the local scale. As a result, hazard assessment is carried out on erosion source areas identified by elementary spatial

In many European regions, agricultural land is structured visually and physically by the juxtaposition of these parcels. Each of these units is an erosion system whose activation depends less on near environment than on the distinct features of each parcel. By mapping the hazard at the scale parcel, the aim is to provide land management organizations with data so as to rationalize the Fig.ht against erosion hazard not at catchment area scale, but

Taking into account anthropic activity in assessing hazard means resorting to the concept of agricultural practices as opposed to that of land cover, unlike all others erosion models at regional scale. The idea here is to re-contextualize hazard by looking at agricultural practices strongly structured along annual cycles (duration and management of intercrop) and

Initial, intermediate and final data from the model needs to be implemented as numeric georeferred informations, usable with GIS and presented at different spatial resolutions

Since diagnostic of erosion hazard is based on data susceptible to change on a short or medium time, the model needs to be designed in order to easily generate data and hazard level updates. This perspective will then offer the possibility to develop an exploratory approach aiming to measure the positive or negative effects of a climatic tendency or a

showing the main agricultural, administrative and hydrological divisions of the area.

of the Lingèvres.

**2.1 Background** 

**2. The SCALES model** 

units such as agricultural parcels.

directly at source area level.

multiannual cycles (crop rotation).

planned change in agricultural practices.

SCALES is a tree form model based on the use of GIS in order to map the potential sensibility of areas and soil erosion hazard. Potential sensibility of areas represents the first fundamental concept of the model. This concept aims to precise if the studied area is able to generate erosive runoff when we integrate both erodibility of soils, land-use and topography. The computational model is therefore a global indicator.

Erosion hazard defines the probability of appearance of soil erosion by water when potential sensibility and rainfall erosivity are put together. The rainfall erosivity depends on meteoric conditions. The latter will be higher during wet years and lower during dry years. We thus estimate a mean of rainfall erosivity based on rainfall data originating from pluri-annual period. Therefore, the erosion hazard has to be considered as a mean hazard.

The factors of evaluation of soil erosion hazard (soil erosion, agricultural practices, topography and rainfall erosivity) are displayed by the input data which can be of quantitative or qualitative relevance. Each factor is defined by one or several types of input data. All data types are converted in measurable data which in turn will express levels of pressure on the erosive runoff trigger. Some input data are combined in order to obtain combined data which generate also level pressure. Levels of pressure from combined data will lead to the estimation of level of hazard.
