Analysis of earth observation time series to investigate the relation between rainfall, vegetation dynamic and streamflow in the Uele' basin (Central African Republic)

Abstract

The hydrology of tropical forests play a key role in watershed processes such as soil erosion, streamflow and ground water recharge. However, tropical forests of Africa are least investigated due to the poor network for data acquisition. Earth Observations can fill this gap by providing consistent time series of data. We analyzed trends of rainfall, vegetation index and river water levels derived from satellite data for the Uele sub-basin and we pointed out that rainfall and river water levels are positively correlated only during the dry season when vegetation activities is low. The unexpected low correlation during the season of highest precipitations is due to the role of vegetation, which is characterized by a significant seasonality also in evergreen tropical forests. These results underline the importance of modeling the role of canopy in the interception and evapotranspiration of the available precipitation in order to provide reliable information on stream flow dynamics.