Monitoring water and sediment delivery to Lake Abaya and Lake Chamo in the Southern Rift Valley Basin, Ethiopia

Human activities like deforestation intensify runoff and soil erosion, leading not only to local land degradation but also to off-site impacts such as excessive lake sediment deposition. Climate change may further exacerbate the erosion rate, particularly where human activity is highest. Hence, the sustainability of natural lakes is threatened by soil erosion and climate change. This is certainly the case for the tropical lakes of Abaya and Chamo in the southern Ethiopian Rift Valley. High rainfall intensities, steep slopes, and increasing population levels make these lakes susceptible to high rates of sedimentation, and rising water levels due to changing sediment and water budgets lead to coastal flooding of agricultural land and infrastructure. However, in order to properly assess the impact of human-induced land use change and climate on the lakes, quantitative assessments of the water and sediment delivery to the lakes are required. High-quality data are often missing for tropical lake environments, especially in developing countries. 

Here, we established seven streamflow and suspended sediment concentration (SSC) monitoring stations across four selected rivers draining to Lake Abaya and Lake Chamo: Bilate (5480 km²), Elgo (298 km²), Kulfo (467.2 km²) and Shafe (191 km²). Together, these stations monitor 35% of the total area contributing to both lakes. Streamflow was measured at 10-minute intervals using a transducer, with atmospheric pressure corrections from a Baro diver. In total, 3501 samples were collected to measure SSC. Observed SSC ranges from 0.08 g/l to 107.75 g/l, whereas discharge varies between 0.01 and 410.65 m³/s. Sediment rating curves were developed using SSC and streamflow data to enable the estimation of total suspended sediment yield (SY) using continuous streamflow records. SY from the four gauged catchments was calculated at 11.3 Mt/year, with area-specific SY varying between 1083.43 and 10117.5 t/km²/year. Overall, Bilate River contributes 67% of the total sediment load, making it the most significant river in terms of the total SY. However, when normalized by catchment area, the Gamo highland catchments have higher net erosion rates. Strong temporal variability in SSC and SY is observed, which can be explained by seasonal changes in vegetation cover and rainfall intensity. Time series of SSC for each river can be correlated with NDVI-data in the corresponding catchments and rainfall erosivity calculated from high-resolution meteorological data. Catchments draining the Gamo highlands have their highest sediment transport rates at the start of the rainy season (May to June) when vegetation cover is low, contributing 60% of SY. In contrast, Bilate, which drains the rift valley itself, experiences a peak sediment yield in August-September, representing 61% of its annual SY.

The integration of satellite-derived NDVI, high-resolution rainfall erosivity data, and timeseries of SSC and discharge enables to identify periods and areas of enhanced erosion and sediment delivery to the lakes. Such spatio-temporal information will be used to calibrate and validate erosion models, which in turn can simulate the impact of management scenarios on lake water and sediment budgets.