VARIABILITY AND PERIODICITY IN PRECIPITATION AND DISCHARGE OF THE RED RIVER

A major obstacle in identifying physical mechanisms responsible for increases in river flooding is differentiating among anthropogenic, climate, and natural hydrologic fluctuation effects. This study investigated the effect of climate induced changes on the flooding of the Red River in North Dakota. Both time and frequency domain approaches were applied to the meteorological and hydrological time series. Mann-Kendall (MK) and Trend Free Pre-Whitening (TFPW) tests were used to analyze monotonic trends in precipitation and deseasonalized river discharge data, and practical significance level in precipitation. Cross-correlation among different time series in El Nino Southern Oscillation Index (SOI), precipitation, and river discharge were evaluated using both parametric Pearson’s correlation test and non-parametric Kendall’s rank correlation test. A fast Fourier transform based spectral analysis was applied to investigate periodicity and coherence of precipitation and discharge. Results from the MK and TFPW tests indicated that all river gauges experienced significant increases in trends of deseasonalized discharges and 46 out of 62 rain stations had significant upward precipitation trends. The practical significance analysis signified an overall increase in annual precipitation from 1931 to 2009, ranging from 5% at Lisbon to 29% at Drayton. No significant correlations between SOI and precipitation or discharges were found, thus excluding El Nino Southern Oscillation effects as contributing to increased flooding. The spectral analysis revealed that precipitation and discharge time series shared significant 12 and six month periodicities. Precipitation signal was stronger than discharge signal at all stations exclusively for the 12 month cycle. The consistently tight correspondence between frequency peaks for precipitation and discharge suggested that precipitation is the dominant factor in controlling periodicity (although not necessarily volume) of discharge. This close correspondence confirmed the contribution of rising precipitation to the Red River flooding and also revealed its control over flood timing. Studies are underway to evaluate the anthropogenic and natural hydrologic fluctuation effects.