Seasonal and interannual validation of satellite-measured reservoir levels at the Kainji dam

Abstract

Although several studies have demonstrated the potential for inland water monitoring using satellite radar altimeters, the technology still carries with it certain limitations. Therefore, the direct application of altimetry data to reservoir management is often preceded by validation of satellite altimetry data with reliable ground-measured data. Kainji reservoir in Nigeria was selected for this study due to its remote geographical location, hydrological behavior, operational requirements, and morphometry. Validation of Kainji reservoir altimetric levels would highlight the applicability of the technology for monitoring one of Africa’s largest reservoirs with a surface area of 1270 km2. We performed and improved the validation exercise by separating each lake level dataset into two separate seasons: dry and wet, to investigate the effect of different seasons on the relationship between ground-measured and satellite-measured data sets. Lake level data from Topex/Poseidon and ERS/ENVISAT satellite altimeters were compared with gage data from 1992 to 2002. The results suggested a significantly higher relationship between gage and satellite data for the wet season (R2=0.93) than for the dry season (R2=0.77). For interannual validation, the trend was the same for both satellite altimeters but the relationship was generally higher for the T/P altimeter (R2=0.95). Root-mean-square errors in water levels ranged between 0.50 to 0.83 m for both altimeters and seasons, agreeing with values expected for lakes of identical physical characteristics. Key words: Satellite altimetry, remote sensing water, reservoir level variation, Kainji dam, African lakes.