Dans ce travail on s'intéresse aux sédiments transportés en suspension par le cours d'eau Oued Tafna à Beni-Bahdel. Une analyse de la variabilité de la charge en suspension de l'eau quant aux fluctuations des régimes hydro-climatiques, pour la période allant de septembre 1988 à août 1993, montre que la production des sédiments est très forte en automne mais de faible ampleur au printemps. En été, des crues éphémères fortement chargées sont parfois observées. En automne, le flux des matières solides en suspension représente 44% du flux annuel estimé à 286000 tonnes, alors que les apports en eaux ne représentent que 17% de l'apport annuel moyen évalué à 30,9 millions de m3. Au printemps, malgré l'augmentation des débits (60% de l'apport total annuel), la charge en suspension diminue de façon significative.
Cependant, les sédiments ramenés par le cours d'eau sont produits soit par ruissellement superficiel sur les versants soit arrachés du lit et des berges du cours d'eau. On montre alors que la contribution de ces deux mécanismes d'érosion reste remarquablement constante d'une année à une autre. La production du ruissellement superficiel en matières solides représente alors 62% du flux annuel alors que l'érosion du cours d'eau contribue par 38%.
Production and transport of suspended sediment transport in the Upper-Tafna river bassin (North West Algeria)
Knowledge of suspended sediment loads carried by rivers provides insight into the underlying erosion mechanisms. In Maghreb rivers there are few data available on suspended sediment transport (PROBST and SUCHET, 1992). In this context knowledge of suspended sediment loads in the Oued Sebdou River (Mediterranean Algeria), which is a main stream in the Upper-Tafna basin, is of interest. The purpose of this study was to determine periods of active erosion and estimate the contributions of slope erosion and channel erosion in this basin.
Situated in northwest Algeria, the Upper-Tafna basin covers an area of 256 km2 (Fig.1). Rising in Ouled Ouriach, the upper reaches of the river flow through Jurassic soils at altitudes up to 1400 m. These tributary streams join on the Sebdou plain (900 m), which is composed of Plio-quaternary alluviums. The soil is composed of calcareous-marnes, calcareous and Jurassic dolomites (BENEST, 1972; BENEST et al., 1999) until the Beni-Bahdel dam.
The study zone is characterized by a semi-arid climate. Analysis of the rainfall data at the Beni-Bahdel station (X=34°42'33"; Y=01°29'48"; Z=660 m) from 1939-1940 to 1997-1998 (Fig. 2) demonstrates that the rain supply has decreased greatly since 1975, with more than an 18% decrease in rainfall amounts.
Hydrological data and methodology
The study used instantaneous water discharge and suspended sediment load values (1257 values) measured and supplied by the National Agency of Hydrologic Resources [ANRH]. Five annual water years have been investigated during the period from September 1988 to August 1993. The discharge values (QL, in m3/s) were determined from the rating curve for the measured heights of the water at the gauging station. The suspended load (g/l) was determined on water samples taken from the stream channel banks. The number of samples taken was adapted to the hydrological regime (i.e. every other day or during flood periods) and samples were collected as frequently as increments of 10 cm.
The suspended sediments were carried from the slopes into the river by surface runoff or by stream channel erosion caused by flowing water. In order to express the contribution of these two mechanisms, we applied hydrograph separation methods used by ETCHANCHU (1988), ETCHANCHU and PROBST (1986), KATTAN et al. (1987) and PROBST and BAZERBACH (1986).
Balance sheet of solid and liquid contributions
The annual liquid contributions from 1988 to 1993 were evaluated to be 30.9 million m3. The abrupt slopes of the basin cause water to move quickly with a lot of force. In turn a large amount of suspended sediments (estimated at 286,000 tons) is generated, which corresponds to a soil erosion rate of 1120 tons/km2/year. This value is comparable to those estimated by WALLING (1984) where the erosion rates ranged between 1000 and 5000 tons/km2 /year.
As seen in Table 1, the annual suspended sediment load varies dramatically. In 1990/91, the soil erosion rate was estimated to be 4283 tons per km2 ; this amount was 180 times higher than the sediment load in 1992/93 (24 tons per km2). Furthermore, there is no relationship between annual suspended sediment load and water discharge or annual rainfall. The annual precipitation (355 mm) was recorded during the year 1988/1989, and generated a soil erosion rate evaluated at 1072 tons per km2. For a similar rainfall, 345 mm recorded during the year 1992/93, the erosion rate was only about 24 tons/km2 /year.
Autumn liquid contributions represent only about 17% of the mean annual supply (Fig. 4). Furthermore the heavy rains produce higher peak flows as the rain encounters dry soil with less vegetation cover and produces large quantities of solid by heavy rain splash erosion. The mean seasonal suspended sediment concentration was higher representing approximately 44% of total annual production. Spring corresponds to a dilution in sediment concentration caused by a large groundwater contribution. The volume discharged by spring flow is about 60% of total annual river discharge. The solids transported during this period are mainly eroded from the stream banks and represent 36% of total annual sediment load.
Calculated values of monthly suspended sediment load and water supply are reported in Table 3. The monthly suspended sediment load carried by the river was very irregular. The majority (94%) of the total sediment load was transported in only three months, 45% in September, 30% in March and 19% in July. The monthly amounts varied from year to year, and this can be explained by the occurrence of floods (Table 4).
The river suspended sediments were carried from the slopes into the river by surface runoff water or by stream channel erosion. Using a hydrograph separation method, the contributions of these two mechanisms are summarized in Table 4. According to these values we observed that :
- The contribution of slope erosion processes represents an average of 62% of the total river-suspended sediment transport.
- The second contribution from channel erosion represents at least 38% of the total river transport. This percentage is comparable with those obtained by ETCHANCHU and PROBST (1986) for the Garonne River basin (30%). This contribution is higher than those estimated by KATTAN et al. (1987) (22% for the Senegal River basin) and by ROBINSON (1977) for some American rivers.
- This contribution was far less than those estimated by DUYSING (1985) for a forest stream in Luxemburg, where sediment produced by surface runoff was about 62% of annual suspended sediment loads.
- This estimate is comparable to those estimated by where the erosion rates range between 1000 and 5000 tons/km2/year.
- Although the annual amount of suspended sediment was variable, the relative contributions of the slope and channel erosion were less variable from year to year.
- Liquid discharge,
- suspended sediment load,
- period of active erosion,
- slope erosion,
- channel erosion,
- Oued Sebdou River basin,
Veuillez télécharger l’article en PDF pour le lire.