El Hage Hassan, Hussein, Charbel, Laurence, Touchart, Laurent, Lacaze, Bernard, Centre d'Etudes pour le Développement des Territoires et l'Environnement (CEDETE), Université d'Orléans (UO), Faculté des lettres et des sciences humaines, université Libanaise, Dekwaneh, Liban, and Université Libanaise
Extended abstract.Soil erosion is a phenomenon that undermines the natural environment and the agricultural activities in the Bekaa Valley (Lebanon) and on the surrounding foothills. In the absence of some data, that the situation in Lebanon during recent decades has caused, it may be relevant to change or adapt some of the classical equations about soil loss. This research is done on an area located near the village of Rachaya, extending over 130 km2. The study area lies between latitudes 33o30’30’’ and 33o38’30’’ North and longitudes 35o49’30’’ and 35o59’30’’ East (plate 1). It comprises two types of geomorphic-dynamic units: the plain and the slopes of the western chain culminating at 1750 m. The median elevation is between 1350 and 1400 m, according to the hypsometric curve (plate 2), and the mean slope has an order of 9 degrees. The Mediterranean environment, the aggressiveness of the climate, the relief and geology (plate 3), the slopes locally rather strong, the deforestation and the extent of crops form a context favorable to erosion.The methodology (plate 4) is based on a modeling adapting the Wischmeier and Smith Universal Soil Loss Equation, USLE, (A = R * K * C * P* LS) (Wischmeier & Smith (1960, 1978) whereby the factors entered into the model are: rainfall erosivity (R), soil erodibility (K), soil cover (C), cultural practices (P), slope value and slope length (LS) (plate 4). This model uses the contributions of geographical information systems and satellite images (Ikonos, Landsat). -The R factor (plate 5) is calculated based on the formula elaborated by Renard & Freimund, (1994). -The K factor (plate 6) depends on soil texture, structure and organic material. The calculation of this factor relies on the formula of Wischmeier et Smith (1978). Soil samples were taken from the site to show the soil granulometric characteristics of the study area (table 1).-The C factor (plate 7) is determined based on tables established by Wischmeier & Smith (1978). It depends on the vegetation height, the recovery rate and the bibliographic knowledge of the Mediterranean region (Sadiki et al., 2009, El Garouani et al., 2008) (table 2).-With respect to the P factor, two codes were identified, one for the agriculture terraces and the other for areas of anti-erosion practices. -For the purpose of this study, the adopted model for the LS factor (plate 8) combines only the length and the inclination of the slope.The results (plate 9) show the role of the steeper slopes and of the less protective vegetation cover. On a third of the surface of the study area, soil loss is between 400 and 1800 t/ha/year and on one fifth of the surface between 50 to 400 t/ha/year. On the ledges, losses decrease to values between 50 and 5 t/ha/year (plate 9). However, only 6 % of the study area shows significant soil losses. Soil cover is generally the most discriminant factor. It strengthens soil resistance by breaking the kinetic energy of raindrops and by intercepting some of the precipitations. It is true that the Rachaya area has been affected by a strong degradation of the vegetation cover, so that the C factor varies between 0.7 and 1 for 77% of the study area (plate 7). Anti-erosion practices are not widespread and subsist only on a few hectares of agricultural terraces. These conditions reveal the insufficiency of protection against the aggressiveness and that is why the R factor may exceed 1600 (plate 5). Several parameters are decisive to explain the soil erodibility: the infiltration capacity, the retention texture and the tearing susceptibility. Most of the soil in Rachaya is fragile and prone to erosion. The force of the relief, especially where there are steep slopes and topographical irregularities, greatly helps erosion; Rachaya high values of LS in some places, contributed largely to the erosive risk (plate 8). The validation of our results was made on the one hand on fieldwork and the other hand using the spectral response of the eroded areas, based on the gloss value processed from the Ikonos satellite image with high spatial resolution of 1 m. The gloss index dissociates vegetated areas from bare mineral areas: it authorizes to assess the surface condition of soils in the study area. This index, which is generated from the red and near infrared channels, characterizes the albedo and allows the diagnosis of soil degradation from color changes reflecting changes in the soil. The soil spectral signature depends on the organic matter content, on the moisture content and on the mineralogical composition. These variables have a major impact on the color and brightness of soil. The erosion mechanism, which strips the surface layer, makes the soil clearer because of the destruction of organic matter. At the opposite, the soil moisture, the presence of organic matter or an increase of the vegetation cover give a dark color to the soil. Finally the overall accuracy of our validated results is approximately 80 %. Due to the importance of the erosion phenomenon which has been quantified, it seems necessary to increase the stabilization of the slopes and to implement protective farming techniques, L’érosion des sols fragilise le milieu rural de la plaine de la Békaa et ses bordures. La méthodologie utilisée repose sur une modélisation adaptant l’équation universelle des pertes en sol (USLE) de Wischmeier et Smith, en utilisant les systèmes d’informations géographiques et des images satellitaires. Les facteurs entrés sont : l’érosivité de la pluie, l’érodibilité du sol, la couverture du sol, les pratiques culturales, la déclivité et la longueur de la pente. Sur un tiers de la surface, les pertes en terre sont comprises entre 400 et 1800 t/ha/an, sur un cinquième entre 50 et 400 t/ha/an. Sur les replats, elles tombent entre 50 et 5 t/ha/an. 6% de la surface présentent des pertes négligeables. La validation de ces résultats est faite sur le terrain et en utilisant la réponse spectrale, grâce à l’indice de brillance de l’image Ikonos de haute résolution spatiale. Devant l’importance du phénomène, il semble nécessaire de conseiller des techniques agricoles protectrices.