Trend analysis of streamflow with different time scales: a case study of the upper Senegal River

This study investigates long-term trends of three different time scales including monthly, seasonally and annually at the upper Senegal River basin. Daily streamflows for the period 1961–2014 at Bafing Makana station were used and analyzed to conduct this research. The serial structural of the different time series (monthly, seasonal, and annual) were investigated in order to detect the presence of autocorrelation. Mann–Kendall test was applied to no autocorrelated series and the Modified Mann–Kendall test for the autocorrelated. Theil and Sen’s slope estimator test was used for finding the magnitude of change and Pettitt test was applied for detecting the most probable change year. Results exhibited a decreasing trend of the annual streamflow yet at the 5% significance level, streamflow series did not have any statistically significant trend for the whole period; however, integrating the different change years, decreasing trend is significant before the first breaking point (1976) and increasing trend is significant from first breaking point to the second change point (1993). For the monthly series, all months exhibit a non-significant decreasing trend except for the month of June. The seasonal series show a decreasing trend which a significant at MAMJ season. Change years were varying accordantly to the scale.