Différentiation hydrogéochimique entre les nappes superficielles des alterites et profondes du socle fissur dans le bassin versant de Kolondièba (sud du Mali): approche statistique par la methode SOM des reseaux de neurones

摘要

The Kolondieba watershed (3010 km2, sub-watershed of Bani, the main tributary of the Niger River in Mali, in the Soudanian area), was designed under the International Program RIPIECSA (interdisciplinary and participatory research on Interactions between Ecosystems, Climate and Societies in West Africa) to understand the impact of climate variability on the balance of water flow. The objective of this work is to characterize the hydrogeochemical characteristics of groundwater (alterite and bedrock fractured aquifer) to better understand their contribution in the flow production. The statistical method of ordination SOM (Self Organizing Maps) of neural networks was used to discriminate individuals by identifying nonlinear relationships impossible to obtain with conventional methods of ordination. No significant chemical peculiarity can be found due to high chemical variability internal to each aquifer considered, reflecting on a disconnection between the different alterite groundwater on one hand and, on the other, between the fractures of the crystalline basement.