A Recession-Forecast Model for the Blue Nile River

Abstract

An analytical model for the forecasting of the Blue Nile River low flow is developed and compared with the current modified similar year method. The model is based on the general non-linear reservoir equation and the historical flow of the river for the calibration period 1912-1961. The comparison is made after the year 1961 for 6 different years in terms of the temporal 10-day flows and for the period 1962-1996 in terms of the standard error of estimates (SEE). The model and the modified similar year forecasted 10-day flows were also compared with the actual temporal 10-day flows during the driest and the wettest year of the whole record between 1912-1998. Results showed that the model temporal distribution of flows is more close to the actual ones compared with the modified similar year method. The model has less SEE in 31 years out of 38 compared with the modified similar year method.

Models for recession flows in the upper Blue Nile River

“…b D 0Ð6), and m was determined by multivariate regression analysis (R 2 D 0Ð645 . Mishra et al (2003) showed that their method provided generally better results than the results obtained by Abdelhadi et al (2000). Similarly in this paper the results obtained by where the residuals ε tC1 are assumed to be independent random errors.…”

“…Results indicate that the non-linear method, in which the recession parameter a is related to the preceding flow magnitudes, provides a substantially better fit to the observed data than that of the three other methods. The Abdelhadi et al (2000) non-linear Results indicate that the non-linear storage model provides a substantially better fit to the observed data than the linear and multiple linear storage models. Semi-logarithmic plots of flow recession performed in previous Blue Nile studies (Hamad, 1993;Abdelhadi et al, 2000) also emphasized that the Blue Nile does not follow a linear reservoir concept.…”

“…The recession forecast is needed immediately after the flood has tailed off for better planning of the winter crop area in Sudan. The total dry season plantation area depends merely on the expected low flow of the Blue Nile River (Abdelhadi et al, 2000). The decision on the winter plantation has to be taken as early as by first half of October.…”

“…Similarly 22 years of validation period, which also excludes years with abnormal recession periods, are selected from the rest of the data period . Mishra et al (2003) describe in detail the selection criteria, whereas Abdelhadi et al (2000) and Sutcliffe and Parks (1999) highlight the importance of forecasting during the recession period. …”

“…Non-linear recession flow model by Abdelhadi et al (2000) Abdelhadi et al (2000) developed an analytical model for forecasting the Blue Nile recession flow based on the non-linear reservoir model Equation (2). For parameter calibration they used a slope coefficient m, where m D n/a 1 n , with n D 1 b.…”