**4.2.2 Input data for a forecast of the soil erosion hazard to 2100**

The agricultural and climatic data represent the two types of data that show a temporal variability. Nowadays, works relating to climate change allow us to have pieces of information that make consensus in the scientific community of climatologists for a distant future (GIEC, 2007a). Unfortunately, we cannot say the same for agricultural practices as the agricultural evolution depends at the same time on its interaction with climate, political choices and the socio-economic situation. Yet, it is not possible to precise with certainty what will be local agriculture in a distant future because we know little if nothing about the agricultural consequences of these interactions, political choices and socio-economic characteristics by 2100. So, our projection of the erosion hazard in 2100 leans on data from the GIEC that reproduce by default the current agricultural situation.

To characterize the Norman climatic context by 2100, we used and adapted the GIEC's simulation data (GIEC, 2007a) concerning the A1B scenario (Cantat et al., 2009). These ones reveal a annual increase of temperatures in Normandy on the order of 2.8° C from now to 2100 (Fig. 12A) with a global warming being more intense during summer (+ 3.2° C). With regard to rainfalls, the annual accumulations would remain stable, which would nevertheless hide differentiated seasonal behaviors: +9% during winter and -21% during summer (Fig. 12B). All these data were used to determine the conditions of rainfall erosivity in 2100.

SCALES: An Original Model to Diagnose Soil Erosion Hazard

a current period's year whose occurrence periodicity is 4 years.

and Assess the Impact of Climate Change on Its Evolution 247

2100. Knowing the frequency of occurrence of the 1994 climatic conditions, it leads us to the conclusion that the hazard levels obtained for an average year by 2100 would correspond to

Fig. 13. Maps of soil erosion hazard in current climatic context at monthly scale.

warming in the order of 2 to 3°C by 2100 (Séguin, 2010).

The climate change which has been observed in Europe from the pre-industrial to the current period resulted in an increase of temperatures in the order of 1°C and in a modification of rainfall distribution: +10% to +40% in Northern Europe during the 20th century and -20% in the South (EEA, 2008). On a world scale, we may not evade a climate

Available data of the climate change effects on soils are very insufficient in European countries (EEA, 2008). However, according to some work, It seems that this leads to an

**4.4 Discussions** 


Fig. 12. Regional climatic projections at horizon 2100 starting from scenario A1B concerning temperatures (A) and rainfalls (B)

#### **4.3 Results**

#### **4.3.1 Soil erosion hazard for the current period**

In order to show the most interesting results, our comments will concern more specifically on the period that goes from September to February in which the temporal variability of the erosion hazard is the highest. The monthly mapping of the hazard according to the current climatic data shows that the cultivated parcels are characterized by a susceptibility to the soil erosion by water which is medium or high (Fig. 13).

In September, the hazard levels are quite low or medium. At this time, the available water contents in the soils are not refilled, which postpone the presence of hydrological surplus. In addition, the plan covering is assured by corn and wheat stubbles.

In October and November, the average level of the soil erosion hazard concerns nearly 90% of the cultivated parcels. The increase of the soil susceptibility to erosion results from the bare soil of areas ensilaged and sowed with wheat. This can also be explained by the appearance of the first poor hydrological surplus.

The soil erosion hazard reaches its highest level in December and January. It becomes important for half the cultivated parcels. If the rate of plant covering is not very different from the previous period, the erosivity has increased due to important hydrological surplus and to the increase of days in which rainfalls exceed 10 mm.

In February, the susceptibility to erosion comes back to medium on the catchment's uphill border and quite poor elsewhere. This is explained by a better plan covering and land use and by a significant decrease of the amount of days in which the rainfalls intensity is exceeding 10 mm.

#### **4.3.2 Forecasts for 2100**

The projection of the soil erosion hazard for September shows levels that are similar to those found for the current period (Fig. 14). In October, we notice that in 2100 the cultivated areas concerned by the low and very low hazard increase and that parcels characterized by a high hazard vanish. The reduction of the susceptibility of cultivated surfaces to soil erosion could be explained by a huge decrease of summer rainfalls (-21%). This would lead to a delay in the filling of available water contents and thus to differ until November for the beginning of the first hydrological positive balance.

For November, the projection highlights a reinforcement of medium and high levels of the erosion hazard. The tendency would be even more marked in December. The parcels characterized by a high hazard would represent 90% of the cultivated areas. The increase of the susceptibility fo the catchment to soil erosion would be discernable until February. The strong hazard would still be present and the medium hazard would concern more than 80% of the cultivated areas.

Comparing the foreseen climatic data for 2100 with the ones from our period of reference (1991-2004), we noticed that 1994 was a very comparable year to the climatic projection of

Fig. 12. Regional climatic projections at horizon 2100 starting from scenario A1B concerning

In order to show the most interesting results, our comments will concern more specifically on the period that goes from September to February in which the temporal variability of the erosion hazard is the highest. The monthly mapping of the hazard according to the current climatic data shows that the cultivated parcels are characterized by a susceptibility to the

In September, the hazard levels are quite low or medium. At this time, the available water contents in the soils are not refilled, which postpone the presence of hydrological surplus. In

In October and November, the average level of the soil erosion hazard concerns nearly 90% of the cultivated parcels. The increase of the soil susceptibility to erosion results from the bare soil of areas ensilaged and sowed with wheat. This can also be explained by the

The soil erosion hazard reaches its highest level in December and January. It becomes important for half the cultivated parcels. If the rate of plant covering is not very different from the previous period, the erosivity has increased due to important hydrological surplus

In February, the susceptibility to erosion comes back to medium on the catchment's uphill border and quite poor elsewhere. This is explained by a better plan covering and land use and by a significant decrease of the amount of days in which the rainfalls intensity is

The projection of the soil erosion hazard for September shows levels that are similar to those found for the current period (Fig. 14). In October, we notice that in 2100 the cultivated areas concerned by the low and very low hazard increase and that parcels characterized by a high hazard vanish. The reduction of the susceptibility of cultivated surfaces to soil erosion could be explained by a huge decrease of summer rainfalls (-21%). This would lead to a delay in the filling of available water contents and thus to differ until November for the beginning of

For November, the projection highlights a reinforcement of medium and high levels of the erosion hazard. The tendency would be even more marked in December. The parcels characterized by a high hazard would represent 90% of the cultivated areas. The increase of the susceptibility fo the catchment to soil erosion would be discernable until February. The strong hazard would still be present and the medium hazard would concern more than 80%

Comparing the foreseen climatic data for 2100 with the ones from our period of reference (1991-2004), we noticed that 1994 was a very comparable year to the climatic projection of

temperatures (A) and rainfalls (B)

**4.3.1 Soil erosion hazard for the current period** 

soil erosion by water which is medium or high (Fig. 13).

appearance of the first poor hydrological surplus.

and to the increase of days in which rainfalls exceed 10 mm.

addition, the plan covering is assured by corn and wheat stubbles.

**4.3 Results** 

exceeding 10 mm.

**4.3.2 Forecasts for 2100** 

of the cultivated areas.

the first hydrological positive balance.

2100. Knowing the frequency of occurrence of the 1994 climatic conditions, it leads us to the conclusion that the hazard levels obtained for an average year by 2100 would correspond to a current period's year whose occurrence periodicity is 4 years.

Fig. 13. Maps of soil erosion hazard in current climatic context at monthly scale.
