Multi-annual variability of streamflow in La Plata Basin. Part II: Simulations for the 21st century

Abstract

Water resources availability presents a high degree of uncertainty in the context of climate change. In this study, the multi-year variability of simulated streamflow for three rivers (Paraná, Uruguay and Negro) in La Plata Basin (LPB) in the period 1991-2098 is analyzed. Simulated streamflow for that period were produced in a two-stage process involving a Regional Climate Model (RCM) and a distributed hydrology model (VIC). Outputs from two RCMs (PROMES and RCA) weare used in order to assess the sensitivity of the results to different models. Various spectral methods (Singular Spectrum Analysis, Maximum Entropy Method and Multi-Taper Method) weare used in order to detect low frequency variability modes and preferred quasi-periodicities for annual and seasonal simulated time series.Both simulations generate larger run-off for 21st century than those observed for 20th century for the three rivers. For annual time series, variability in time scales longer than 30 years is detected by one of the simulations for the three rivers, and only weakly for Negro River in the other. Seasonal variations of the preferred modes of multi-annual variability are apparent. Paraná River shows a persistent 10-year period during most of the year in the PROMES-VIC simulation. No LFV modes were found for any trimester in any river for PROMES-VIC, while for RCA-VIC simulations, these appeared in austral summer (Negro and Uruguay rivers) or early winter (Paraná River). A striking finding, both for annual and seasonal simulated time series, is the conspicuous presence of pseudo-periods in the 2.5-5 years band that had already been captured in the observed 20th streamflow time series (Maciel et al.Paper 1, 2013). This pattern is probably linked to El Niño-Southern Oscillation (ENSO) phenomenon. These results imply that the well-known relationships between ENSO and precipitation and streamflow anomalies in LPB are expected to prevail during the 21st century.