Euro-Mediterranean Information System on know-how in the Water sector
International portal

HTML Document flash57_fr Item only translated in French

Released 07/04/2008

Comments on Item (15), FLASH No. 57 by Mamdouh Shahin

-----Original Message-----

From: Mamdouh Shahin []
Saturday, April 05, 2008 5:59 PM
To: EMWIS Flash (HTML)
Subject: Impact of Climate Change on the Flow of the
Nile River

Dear Dr el-Kharraz,J.,
Thank you for mailing me FLASH No. 57. Please find my comments on Item (15) concering the impact of climate change on the flow of the River Nile. I hope through FLASH that the comments below will reach our Egyptian colleagues who are resposible for the
Nile water.
While thanking you in anticipation, please accept my sincere regards.
Mamdouh Shahin 


Comments on Item (15), FLASH No. 57 by Mamdouh Shahin

shahin1932 at

 Hydrologic variables, such as streamflow, occur in nature as a result of interconnected physical processes. Among these processes climatic and physiographic factors play a dominant role. Hence, hydrologic variables are essentially the outcomes of complex time-varying processes, which can be measured by a finite number of observations. Analyses of these observations show that hydrologic variables are both nonlinear (Amorocho and Orlob, 1961) and stochastic in nature (Yevjevich, 1971). These properties can be attributed to:

- Time-varying geological processes such as erosion, deposition, weathering, … etc.

- Climate changes in time

- State uncertainty in time

- Energy transfer of the hydrologic cycle, which is nonlinear in behaviour pattern


The literature of hydrology and climate of the Nile and its Basin is full of accounts related to time variation of the river flow. The next paragraphs, therefore, aim at highlighting the annual fluctuation of the River Nile flow in the course of time since the beginning of the 20th century and some of the approaches used for its interpretation.

In 1935, the Transactions of the American Society of Civil Engineers (Trans. ASCE, 1935) published a paper by C.S. Jarvis on the Flood-stage records of the River Nile. That paper opened the door wide for several professionals to debate whether river stage, and subsequently river flow, varies in time according to a regular cyclic pattern. Cycles such as Wolf, sunspot and Brückner were dealt with in the paper and the discussions that followed. One of the discussers of the said paper, H.E. Hurst, a leading figure in the hydrology of the Nile and in charge of the Physical Department of the then Ministry of Public Works, Egypt, denied any significant role of cycles in the time-variation of the water level and flow of the Nile. The high correlation between the river level at Namasagali on Lake Victoria and the sunspot cycle (11.1 year) for the period 1896-1923, practically disappeared in the subsequent period 1923-34. Instead, Hurst pointed to the strong possibility that the variation might be attributed to the Southern Seas Oscillation (SSO). He added that the then level of understanding and measuring technique of such an event was not developed enough to support his thinking.


The question of periodicity in river stages and discharges did not come to an end. As an example, Ändel and Balek (1971) upon investigating the flow of the Niger River at Koulikoro station, Mali, for the period 1907-57 concluded that the existence of at least one period of 25.5 year (close to Brückner) was proved at 95% significance. Periodicities of 7.3 year (close to Wolf) and 3.0 year (close to Clough) were also traced in the flow sequence. When the three cycles were incorporated in the model they developed the result was quite satisfactory. Later, Shahin (2002) found that the agreement between the measured and model-obtained flows for the period 1958-79 was less than satisfactory. The relationship between the cycles in the solar phenomena, weather and climate, especially the 11.1- year sunspot cycle and the 22.2-year Brückner cycle, and the rainfall depth at several locations all over the world has been strongly emphasized by King (ESA, 1975).


Almost two decades later, el-Tahir (1996) presented the results of a study in which he used two extensive sets of data; one set giving the sea surface temperature (SST) of the Pacific Ocean, and the other the Nile flow at Aswan. The hypothesis tested in his study was that the natural variability of the river flow at Aswan is related to El Niňo Southern Oscillation (ENSO). The question that followed was whether such information could be used for improving the predictability of the Nile flood. The other objective of the study was to use the relation between ESNO and the Nile flow as an approach for explaining the Hurst Phenomena. It has been suggested then that the ENSO events could be accurately predicted 1 or 2 years ahead using the coupled model of the ocean-atmosphere system available at the Geology Department of the University of Columbia, U.S.A. The predicted SST values for the 17-y period, 1973-89 were used to predict the annual flow volume at Aswan using the expression:


,                                                                         (1) all volumes are expressed in 109 m3 y-1


El-Tahir classified the flow arriving at Aswan into normal ((80-100)* 109 m3 y-1), low flow (less than 80*109 m3 y-1) and high flow (more than 100*109 m3 y-1). Twelve out of the seventeen annual flows predicted by Eq. (1) were in agreement with the measured flows. As a matter of fact, Eq (1) is based on annual flow averaged over a small window of a 0.5 o C centered about –1.5 o C, -1.0 o C, -0.5 o C, 0.0 o C, 0.5 o C, 1.0 o C and 1.5 o C. It should be remembered that flood predictions are made six months ahead of the occurrence of flood in August.  Lastly, El-Tahir (1996) developed a hypothesis according to which the annual flow in the Nile varies with time following ENSO resulting in a non-stationary process and causing the Hurst phenomena. The non-stationary mean and the random fluctuation component explain 25% and 75% of the observed natural variation, respectively.


Plisnier (1998), upon studying the hydrometeorology of Lake Tanganyika, which is close to Basin of Lake Victoria, concluded that the average air temperature in the lake area increased in the period 1964-90 by 0.7 to 0.9o C while the speed of surface winds decreased, particularly since the 1970’s. Beside the general trend of increase in the air temperature, oscillations around the mean were observed. The correlation between air temperature at each of Bujumbura (northern end) and Mbala (southern end) and the sea surface temperature (SST) in el-Niño 4 area of the Pacific Ocean (150o W to 160o E and 5 o S to 5 o N) is remarkable.


Wigley (1992) stated “While considering the future process of global warming, it is generally accepted that the equator-to-pole temperature difference will decrease, leading to changes in the atmospheric pressure field and atmospheric general circulation. It might be suggested that the subtropical high-pressure belt in the northern hemisphere will extend as a consequence of a comparatively small amount of warming.”  The consequences of global warming may not hold everywhere on the micro scale where prevailing local factors can lead to inconsistencies.   


To illustrate the effect of climate change on river flow, let us review briefly some of the findings of McCabe and Hay (1995) regarding the effects of hypothetical climate change on the flow of the Gunnison River, a tributary of the Colorado River, U.S.A. The model used in that study, known as Precipitation-Runoff Modeling System (PRMS), has been developed by the United States Geological Survey (USGS) around 1988. The data used were the mean monthly temperature To C and precipitation P, mm for the period 1973-89 taken from the basin of the East River.       

- In general the change of precipitation has a more pronounced effect on streamflow than the temperature change has.

- The change of temperature affects in the first place the distribution of streamflow over the year.

- The temperature rise, on annual mean basis, produces a limited decline in streamflow. For example, a rise of 4o C in the annual mean temperature produces a fall in streamflow equivalent to the effect caused by a reduction in precipitation of say 5%.

-          The natural climate change during the period of study was so strong and possibly could have masked the long-term effects of climate change.


Shahin (1999) presented the results of statistical analyses and time-linear regression models fitted to a number of temperature and precipitation time series in some parts of the Arab Region and the neighbouring areas to include the Nile Basin. We shall briefly review here the results of precipitation series at four stations; namely; Gedaref, Khartoum and Wau, all in the Sudan, and Addis Ababa, Ethiopia.


Station                   Φ,            λ,             Z,            Period of                Trend Equation   

                                N             E              m             observation                                                                                          

Khartoum              15o 37’     23o 33’    385          1938-89                   P = 179.6-1.125 t*

Gedaref                  14o 02’    35o 24’    610          1900-82                   P =-334.4+0.471X**

Wau                       07o42’     28o 01’     433          1940-84                   P = 9893-4.474 X

Addis Ababa        09o 02’     38o 44’    2440        1900-90                   P = 1264.9-1.261t


Where Φ = latitude, λ = longitude and Z = altitude or height above a certain reference level. t is the year number measured from the beginning of the record, so t = 0 and 1for the years 1938 and 1939, respectively, both for Khartoum station and X**is the absolute number of the year, e.g., X = 1900 and 1901for the beginning year and the year that follows, respectively, both for the series at Gedaref. 

The trends corresponding to the above series, in their respective order, show that the annual precipitation undergoes a decrease with time for all stations except Gederaf where the rainfall seems to increase slightly with time. Will this falling trend continue in the years yet to come before any counter change occurs? If so, the unpleasant consequences decline in the Nile flow will be steadily and strongly felt. Assuming that we are prepared to accept the findings of McCabe and Hay (1995), one should expect a decline of between 10% and 20% in the natural flow of the Nile before 2050. The problem will be much aggravated if one considers the future share per capita, not to forget how the riparian countries will then consent to having a new treaty for allocating the reduced shares among these countries.


The writer of the above comments, being himself one of the oldest Egyptian Irrigation Engineers and Nile Hydrologists still alive, shares his present and junior colleagues their worries about the impact of climate change on the natural flow of the river. He strongly advises the decision makers in Egypt to set up a task team of top experts to deal with the problem in its proper dimensions. Extensive and intensive attention should be directed to having adequate hydrometrical data and to learning lessons from experiences with other rivers in the world. For example, the volume on “Climate Variability and Change-Hydrological Impacts” (Editors: S. Demuth, A. Gustard, E. Planos, F. Scatena &E. Servat) of the Fifth International FRIEND Conference contains almost 120 papers from over 30 countries. It is Publication. No. 308 (2006) of IAHS, 708+xii pp. Additionally, the publication entitled “Regional Hydrological Impacts of Climatic Change-Hydroclimatic Variability” (Editors: S. Franks, T. Wagener, E. Bøgh,….….) includes investigations from different climate zones and regions, including many in Africa and Asia. It is Publication No. 296 (2005) of IAHS, 300+x pp. There are many more publications including well performed investigations of impact of climate change on water resources and they all need to thoroughly reviewed and  understood.    




Amorocho, J. & G.T. Orlob 1961 Non-linear Analysis of Hydrologic Systems, Water Resources Center, Contribution No.40, 147 p.

Ändel, & J. Balek 1971 Analysis of periodicity in hydrological sequences, J. Hydrolo. 14: 66-82.

El-Tahir, A.B. 1996 El-Niňo and natural variability in the flow of the Nile River. Wat. Resour. Res. 32 (1): 131-137.

Jarvis, C.S. 1935 Flood-stage records of the River Nile. Trans. ASCE, Paper No. 1944 (with discussions by H.P. Gillete, R.W. Davenport, H.E. Hurst, T.H. Means, J.W. Breadsley, J.C. Stevens, J.W. Shuman, K.O. Ghaleb and C.S. Jarvis), pp 1012-1071.

King, J.W. 1975 Solar phenomena, weather and climate. European Space Agency (ESA) Bull. No. 3, Neuilly-sûr-Seine, France, 24, 100 p

McCabe, G.J. Jr. & L.E. Hay 1995 Hydrological effects of hypothetical climate change in the East River basin, Colorado, USA. J. Hydrol. Sci. No. 40 (3), pp 303-318.

Plisnier, P.D. 1996 Lake Tanganyika: Recent climate changes and tele-communications with ENSO. Proc. of Int. Conf. on Tropical Climate, Meteorology and Hydrology, Brussels, Belgium, pp 228-250.

Shahin, M.M. 1999 Resources hydriques et modification du climat au Moyen-Orient. Bull. de la Soc. Géogr. de Liège, No. 37 (2), pp 75-90.

Shahin, M.M. 2002 Hydrology  and Water Resources of Africa. Wat. Sci.& Tech. Lib., Vol 41, 663 p + CD Rom, Springer (formerly Kluwer Academic Publishers),  Dordrecht/ Boston/ London. 

Wigley, T.M. 1992 Future climate of the Mediterranean Basin with particular emphasis on changes in precipitation. In: Climatic Change and the Mediterrnean (Jeffic, L., J. Miliman. & G. Sastini Eds.), United Nations Environment Programme.

Yevjevich, V. 1971 Stochastic Processes in Hydrology. Wat. Res. Publ. Ft. Collins, Colorado, USA.

Posted by jauad at 07 Apr 2008 18:03:50