AbstractUsing new mathematical and data-driven techniques, we propose new indices to measure and predict the strength of different El Niño events and how they affect regions like the Nile River Basin (NRB). Empirical Mode Decomposition (EMD), when applied to Southern Oscillation Index (SOI), yields three Intrinsic Mode Functions (IMF) tracking recognizable and physically significant non-stationary processes. The aim is to characterize underlying signals driving ENSO as reflected in SOI, and show that those signals also meaningfully affect other physical processes with scientific and predictive utility. In the end, signals are identified which have a strong statistical relationship with various physical factors driving ENSO variation. IMF 6 is argued to track El Niño and La Niña events occurrence, while IMFs 7 and 8 represent another signal, which reflects on variations in El Niño strength and variability between events. These we represent an underlying inter-annual variation between different El Niño events. Due to the importance of the latter, IMFs 7 and 8, are defined as Interannual ENSO Variability Indices (IEVI) and referred to as IEVI α and IEVI β. EMD when applied to the NRB precipitation, affecting the Blue Nile yield, identifying the IEVI-driven IMFs, with high correlations of up to ρ = 0.864, suggesting a decadal variability within NRB that is principally driven by interannual decadal-scale variability highlighting known geographical relationships. Significant hydrological processes, driving the Blue Nile yield, are accurately identified using the IEVI as a predictor. The IEVI-based model performed significantly at p = 0.038 with Blue Nile yield observations.
Improving SWAT model performance in the Upper Blue Nile River Basin using meteorological data integration and catchment scaling