Publication: Hydrologic Impacts of Climate Change on the Nile River

Document technical information

Format pdf
Size 716.6 kB
First found May 22, 2018

Document content analysis

Category Also themed
Language
English
Type
not defined
Concepts
no text concepts found

Persons

John Henry Poynting
John Henry Poynting

wikipedia, lookup

James Hunt
James Hunt

wikipedia, lookup

Abraham de Moivre
Abraham de Moivre

wikipedia, lookup

Organizations

Places

Transcript

Hydrologic Impacts of Climate Change on the Nile River
Basin: Implications of the 2007 IPCC Climate Scenarios
Tazebe Beyene1, Dennis P. Lettenmaier1 and Pavel Kabat2
Abstract
The potential impacts of climate change on the hydrology and water resources of the Nile
River basin are assessed using a macroscale hydrology model driven by 21st century simulations of
temperature and precipitation downscaled from runs of 11 General Circulation Models (GCMs) and
two global emissions scenarios (A2 and B1) archived for the 2007 IPCC report. The results show
that, averaged across the multimodel ensembles, the entire Nile basin will experience increases in
precipitation early in the century (period I, 2010-2039), followed by decreases later in the century
(periods II, 2040-2069 and III, 2070-2099) with the exception of the eastern-most Ethiopian
highlands which is expected to experience increases in summer precipitation by 2080-2100.
Summarized as spatial averages over the entire Nile basin, multimodel-average Nile basin
precipitation changes as percentages of the historical period 1950-99 are 115 (117), 98 (104) and 93
(96), and temperature changes (as differences in ºC from 1950-99) are 1.5 (1.3), 3.2 (2.8) and 4.4
(3.6) for the global A2 (B1) emissions scenario. These changes in precipitation and temperature
resulted in streamflows at High Aswan Dam (HAD) that are 111 (114), 92 (93), and 84 (87) percent
of historical simulated streamflow (1950-1999) for periods I to III, respectively, for the global A2
(B1) emissions scenario. Implications of climate change on the water resources of the Nile River
basin were analyzed by quantifying the annual hydropower production and irrigation water releases
at High Aswan Dam, which generally would follow changes in streamflow, increasing early in the
century to 112 (118) percent, but then decreasing to 92 (97) and 87 (91) percent in Periods I and III,
respectively, for the A2 (B1) emissions scenario.
1
Department of Civil and Environmental Engineering Box 352700, University of Washington,
Seattle WA 98195
2
ALTERRA Green World Research, Wageningen University and Research Centre, P.O. Box 47,
6700 AA Wageningen, The Netherlands
1.0 Introduction
The climate of Africa is both varied and varying. Varied, because climate ranges from
humid equatorial to seasonally arid and sub-tropical Mediterranean and varying because all these
climates exhibit differing degrees of temporal and spatial variability. At the sub-regional scale,
Africa is vulnerable to ENSO and related extreme events (drought, floods, and changes in
hydrologic patterns). That portion of sub-Saharan Africa that depends entirely on the Nile River for
its water supply is particularly susceptible to hydrologic changes that might be associated with a
warmer climate. Flooding and droughts will be increasingly difficult to cope with in the face of
increasing pressures on water supplies due to rapid population growth and dwindling resources.
The Nile River basin is home to 336 million of Africa’s 850 million people. It has
experienced high population growth rates and the population of the basin is expected to double
between 1995 and 2025. Virtually all population projections are for continuing growth, which in
turn will increase demand for natural resources among the 10 Nile River riparian countries. The
potential effects of climate change on the basin have been given less attention than population
growth, notwithstanding several earlier studies (e.g. Yates et al., 1998) showing that the water
resources of the basin are susceptible to climate change.
The IPCC Fourth Assessment Report (AR4) has resulted in a wealth of General Circulation
Model (GCM) runs that have been archived in a consistent manner at the Lawrence Livermore
National Laboratory Program for Climate Model Diagnosis and Intercomparison (PCMDI). These
model runs provide the basis for a much more coherent analysis of possible effects of climate
change using multimodel ensemble techniques (e.g., Krishnamurti et al, 2000) than has previously
been possible. For instance, Maurer et al (2006) evaluated implications of projected 21st century
climate for California’s water resources using PCMDI-archived output from 9 IPCC/AR4 GCMs,
and Christensen and Lettenmaier (2007) evaluated implications of IPCC/AR4 climate projections
for Colorado River water resources from 11 GCMs and two global emissions scenarios archived at
2
PCMDI. We follow the lead of these recent studies in using multimodel ensemble methods to
evaluate the implications of 21st century climate change for the Nile River basin.
Future changes and uncertainties in the allocation of Nile water resources may have
significant effects on local and regional economies, agricultural production, energy availability, and
environmental quality (NBI, 2001, Hulme et al., 2005, Conway et al., 1993, Yates et al., 1998).
Water resource planning based on the concept of a stationary climate is increasingly considered
inadequate for sustainable water resources management (Mohamed et al., 2005). In addition to
natural variability, which is incorporated in existing water planning methods, new water projects
will have to deal with uncertainty associated with population growth and trends in climate change.
Therefore, understanding the uncertainty in projected climate change over the next century, which
is attributable both to uncertainty in the future emissions pathway (related to policy decisions and
public response) and uncertainties in model projections (due to differing sensitivities of the GCMs
to perturbations in atmospheric composition), is essential to understanding how the economy of the
Nile basin will evolve, including social and environmental impacts.
Water resources planning studies which typically are conducted for time horizons of several
decades now require consideration of ongoing global climate change and uncertainties in the
signature of future climate change. The near certainty of increased future water demand in the Nile
basin (notwithstanding uncertainty as to magnitude of demand increases) contrasts with the
uncertainty of climatically-induced changes in the water supply of the Nile River basin (Conway et
al., 1996, Yates et al., 1998(a), Strzepek et al., 1995, Strzepek et al., 2000), both as to magnitude
and direction.
Although this study represents the first attempt to apply multimodel ensemble techniques to
the Nile River basin using the IPCC/AR4 global emission scenarios, the Nile River basin has been
the focus of several previous studies of climate change using different climate models and
techniques. Yates et al., (1998a; b) found that several climate change scenarios implied that
3
agriculture would be negatively affected by climate change despite increased water availability and
only moderate yield declines, as a result of climate change impacts associated with changes in local
and regional biophysical systems and shifts in national agricultural economies.
Conway and
Hulme (1996) used CO2 doubling climate scenarios which provided widely diverging pictures of
possible future Nile River flows, ranging from a 30% increase to a 77% decrease. Strzepek and
Yates [2000] used a dry and wet scenario analysis of possible future conditions and found that
under wet climate scenarios, surplus water beyond 75 BCM (109 cubic meter) remained unused
whereas for drier scenarios (below 75 BCM), water was a constraint to agricultural production into
the 21st century, the result of which was that resources were diverted to less water demanding
crops and the livestock and non-agricultural sectors. Tate et al., (2004) analyzed the sensitivity of
the water balance of Lake Victoria to climate change using HadCM3 A2a and B2 emission
scenarios, and indicated that changes in annual rainfall and evaporation derived from HadCM3
implied that declines in water levels would occur during the 2021–2050 time horizon. This
contrasted with projected increases in water levels later in the century (2070–2099). Since most of
these studies were based on a single GCM and or scenario, the uncertainty in these predictions was
not evaluated. All of these studies were limited by the coarse spatial resolution of the climate
models and the small number of climate models that could be evaluated.
In this paper, we assess the hydrologic implications of climate change on the water
resources of the Nile River basin through a four-step process as follows. The first step is extraction
of the key hydrologic drivers (precipitation and temperature) from the PCMDI archives. In our
case, we extracted output from 11 GCMs, all of which included at least one 100-year ensemble
member, and two SRES (Scientific Report on Emission Scenarios) emissions scenarios (A2,
corresponding roughly to unconstrained growth in emissions, and B1, corresponding to elimination
of global emissions increases by 2100) from PCMDI. The second step is removal of the inherent
bias in the climate model predictions and resolution of the scale mismatch between the GCMs
4
(most of which run at a spatial resolution of two to four degrees latitude-longitude) and the spatial
scale of our hydrologic model (which was applied at a 0.5 x 0.5 degree latitude-longitude spatial
resolution). The third step is forcing the hydrologic model using the downscaled and bias corrected
climate model output. The final step is driving a Lake Nasser reservoir operations model using the
simulated streamflows associated with the 11 GCMs and two global emissions scenarios to assess
the impact of the changing climate on hydropower production and irrigation water supply extracted
from the Nile River at High Aswan Dam (HAD).
2.0 Approach
2.1 Study Area
The Nile is the longest international river system in the world. It flows some 6700 km
through ten countries before reaching the Mediterranean Sea. Its headwaters are in Lake Victoria at
about 4º S latitude, and it flows mostly northward to its mouth at 32º N latitude (see Figure 2(a)). It
has a drainage area of about 3.35 million km2, which covers 10% of the African continent, roughly
equivalent to half the area of the continental United States. Egypt and the Sudan are the two major
users of this river, while Ethiopia is the primary contributor to the bulk of runoff. This imbalance
in the primary producers and consumers of the basin’s water resources has lead Swain (1977) to
conclude that the Nile is the international river system which currently has the greatest potential to
precipitate major armed conflict.
The major lakes in the basin (Lakes Victoria, Nasser and Tana) account for 81,500 km2 and
the area covered by swamps is an additional 69,700 km2 (Biswas, 1994). Precipitation is to a large
extent governed by the movement of the Inter-Tropical Convergence Zone (ITCZ) and its
interaction with topography. In general, precipitation increases from north to south, and with
5
elevation. Precipitation is virtually zero in the Sahara desert, and increases southward to about
1200–1600 mm/yr on the Ethiopian and Equatorial Lakes Plateaus (Mohamed et al., 2005)
The Nile is formed by three tributaries, the Blue Nile, the White Nile, and the Atbara. The
flow of the Blue Nile is strongly seasonal because its runoff is primarily driven by monsoon
precipitation. The Blue Nile contributes about sixty percent of the total flow of the Nile, whereas
the Baro-Akobo (Sobat), and Tekezze (Atbara) contribute slightly less than fifteen percent each.
The headwaters of all the tributaries of the Blue Nile are in the highlands of Ethiopia, and the bulk
of their runoff (70% on average) occurs between July and September. Among the tributaries of the
Blue Nile, the Upper Blue Nile (with drainage area 175,000 km2), which contributes about 50% of
the Nile’s flow at High Aswan Dam (HAD), is the most important. The White Nile and the Blue
Nile join north of Khartoum, where they are joined by the Atbara. The river then flows north
through Lake Nasser, the second largest man-made lake in the world, before splitting into two
major distributaries just north of Cairo, the Rosetta branch to the west and the Darneita to the east.
2.2 General Circulation Models (GCMs)
Table 1 summarizes the 11 GCMs for which model output was used to construct forcing
data for the hydrologic model. For IPCC/AR4, six global greenhouse gas emissions scenarios were
used following the IPCC 2001 scenarios. The global emission scenarios ranked from warmest to
coolest in terms of global average emissions at mid-century are A1FI, A2, A1B, B2, A1T and B1.
Global emission scenarios A2 and B1 were chosen for this study because they are the most widely
simulated global emission scenarios in all models. Emissions scenario A2,which is based on the
assumption that future economic and population growth will not be constrained and that there will
be no future limitations on global emissions. The A2 global emission scenario projects global
average CO2 concentrations will reach 850 ppm by 2100. Scenario B1, on the other hand, assumes
6
an increasing dependence on clean and resource-efficient technologies. CO2 concentrations under
this scenario initially increase at nearly the same rate as in the A2 scenario, but then level off
around mid-century and reach 550 ppm by 2100.
2.3 Bias Corrections and Spatial Downscaling (BCSD)
Despite continuing improvement in their physical representations of the climate system,
there remains a substantial scale mismatch between GCMs and most hydrologic models.
Furthermore, while GCMs have a computation time scale that is typically less than one hour, the
physical realism at such short time steps is questionable, and the data volumes (realizing that GCM
output is archived globally) are considerable. As a practical matter, most studies like this one use
monthly aggregates of the GCM predictions, and therefore a temporal disaggregation to the
hydrologic model time step (daily in this study) is required.
The problem goes beyond
computation, however; issues of bias are critical for hydrologic applications, and even at the native
spatial resolution of GCMs and monthly time intervals, climate models are inherently biased. A
number of downscaling and bias correction strategies have been suggested to solve this problem
(e.g., Wood et al., 2002, Wilby et al., 2000, Wetterhall et al., 2005a among others). As shown by
Wood et al (2004), none of the current generations of GCMs are immune to the bias problem, nor
are higher resolution regional climate models.
For this study we used the bias correction and spatial downscaling (BCSD) method
developed by Wood et al. (2002). The method was initially developed for streamflow forecasting
and later was modified for climate change studies by Wood et al. (2004). It is a statistical approach
which uses empirical percentile-percentile mapping as its main element; the reader is referred to
Wood et al. (2002) and Wood et al. (2004) for further details. The method has previously been used
in several climate change studies, including Payne et al. (2004), Christensen et al. (2004), Maurer et
7
al. (2007), Christensen and Lettenmaier (2007) and Hayhoe et al (2007). In brief, the method
downscales monthly temperature and precipitation at the GCM spatial scale (we regridded the
climate variables to a common 2 degrees latitude by longitude spatial resolution to the one-half
degree spatial resolution at which the VIC hydrology model was applied through use of a
probability mapping procedure which utilizes the climatology of both observations (gridded to a
one-half degree spatial resolution) and the model(s). For bias removal, empirical transformations
via percentile-percentile mapping were constructed from the GCM climatology to the observed
monthly climatology for both Tavg (monthly average temperature) and Ptot (monthly total
precipitation).
For Tavg, the linear trend is removed prior to bias correction and is re-imposed afterward,
so as to avoid distortion of the tails of the probability distribution of temperature as temperatures
rise. The observed climatology for the historical run (1950-1999) was derived from the global
gridded precipitation data set described by Adam and Lettenmaier (2003).Sub-grid spatial
variability was represented by spatially disaggregating bias-corrected, GCM-scale forcings to the
one-half degree spatial resolution, following which the monthly time series of the one-half degree
bias-corrected scenarios were temporally disaggregated to daily through use of a resampling
procedure. The bias corrected and spatially and temporally downscaled climate variables were then
used to force our VIC (Variable Infiltration Capacity) land surface hydrologic model.
2.4 VIC land surface hydrologic model
The VIC land surface hydrologic model of Liang et al. (1994; 1996) was implemented at ½
degree (~48km) spatial resolution and 3-hourly temporal resolution in energy balance mode , in
which the model runs at 3hourly time step and iterates for surface temperature rather than setting
surface temperature equal to air temperature. VIC is a semi-distributed grid-based land surface
8
hydrologic model which parameterizes the dominant hydrometeorological processes taking place at
the land surface-atmosphere interface. The model consists of two major components, vertical and
horizontal. The vertical component calculates the water and energy balance components for each
individual grid cell. The horizontal component is a convolution integral, which routes the runoff
generated at each grid cell to basin outlet (tributary or main stem) channels (see Figure 2(b) for
channel routing network and gauging stations).
A mosaic representation of land cover, and sub-grid parameterizations for infiltration and
the spatial variability of precipitation and temperature, account for sub-grid scale heterogeneities in
key hydrological processes. The model uses three soil layers and one vegetation layer with energy
and moisture fluxes exchanged between the layers. The model was calibrated for the entire Nile
basin at three gauging locations, the Blue Nile at Eldiem, the main stem at Dongola, and the main
stem at HAD. A calibration procedure similar to that described in Nijssen et al. (1997) was
followed to assure a match between model-simulated and observed flows for the period in which
historic streamflow observations were available. A validation result of Nile river flow for two
gauging stations is shown in Figure 1.
3.0 Results and discussions
In the following section we summarize our analysis of the results from downscaled and bias
corrected GCM climate models and the selected global emission scenarios. The bias corrected and
spatially-downscaled hydrologic drivers (temperature and precipitation) of future climate (20002100) are compared to 1950–1999 gridded historical observations. We further analyzed and
compared derived hydrologic variables (runoff, SWE, evaporation) simulated by VIC for the GCM
scenarios to VIC simulations driven by the 1950–1999 climate observations. These climate change
results derived from both global emission scenarios are segregated into three time horizons, period
I (2010–2039), period II (2040–2069), and period III (2070–2099). Hydrologic response of the
9
basin to the changing climate is analyzed by quantifying future streamflow changes at selected
gauging stations and implication of these streamflow changes are further evaluated using reservoir
operation model targeting future hydropower production and irrigation
water release for
agriculture at HAD (High Aswan Dam).
3.1 Temperature and precipitation changes
Figures3 (a–c)show the downscaled and bias-corrected long-term multimodel average
precipitation, temperature and simulated hydrologic parameters (evapotranspiration, runoff, and
soil moisture). The downscaled and bias-corrected temperature and precipitation (and other derived
model forcings, following methods outlined in Maurer [2007]) time series from each ensemble
climate model were used to force the VIC model for both A2 and B1 emission scenarios. The 1950
– 1999 historical data were used as a baseline reference for evaluating predicted changes. Because
the sub-basins of the Nile have different physical and hydroclimatological characteristics, we
evaluated results over the main stem Nile (Main Nile) and the two most important sub-basins, the
Blue Nile sub-basin and the Lake Victoria region (Equatorial region) separately. The results
presented in Figures 3(a-c) include long-term changes in simulated runoff, evapotranspiration, and
soil moisture
Projected temperature changes across climate models early in the century (period I) are
somewhat consistent and mostly are in the range of 1 to 1.5 ºC warmer than the historical (19501999) average. The temperature results tend to diverge through the century from period II to period
III, with an increasing trend in all models. As expected, Scenario A2 is generally warmer than
Scenario B1.For the entire Nile basin multimodel average temperature increased by 1.5 (1.3), 3.2
(2.8) and 4.4 (3.6) ºC relative to the historical (1950-1999) mean for the A2 (B1) global emission
scenarios (Table 4). The ensemble spread as measured by the inner quartile distance tend to diverge
from period I to period III with inner-quartile ranges (1st, 3rd) of (1.4, 1.7), (3.4, 3.8) and (4.4, 4.9)
10
for Scenario A2 and (1.1, 1.3), (2.6, 3.0) and (3.5, 3.9) for scenario B1 for periods I to III
respectively (Figure 4(a-c)). Signatures of temperature projections within the Blue Nile and Lake
Victoria regions are different, but both have consistent positive trends. Projected multimodel
temperature increases in the Lake Victoria region are 1.1 (1.0), 2.5 (1.9) and 3.4 (2.9) ºC relative to
the 1950-1999 historical average with inner-quartile ranges of (1, 1.2), (2.2, 2.8) and (3.4, 4.0) for
emissions scenario A2 and (0.9, 1.1), (1.7, 2.0) and (2.7, 3.2) for scenario B1 for the three periods
respectively. The corresponding predictions of mean temperature changes for the Blue Nile subbasin are 1.2 (1.2), 3.1 (2.6), and 4.1 (3.4) ºC for A2 (B1) emission scenarios. The pattern of
ensemble spread is similar to that for the Lake Victoria region (Figure 4). In general, the
downstream part of the Nile basin is expected to experience more warming than the headwater subbasins.
Changes in precipitation are generally reflected in the annual runoff volumes more than in the
seasonal shape of the hydrographs (Lettenmaier et al., 1999) and there is much less consistency
between climate model projections of precipitation changes than for temperature. In the multimodel
ensemble average, precipitation increases over the entire basin for the period 2010-2039, however,
by 2070-2099, precipitation decreases substantially. The IPCC 3rd assessment report (IPCC 2001)
concluded that Nile basin will experience decreases in precipitation ranging between zero and forty
percent by the end of the 21st century, a result which was based on nine GCMs. Similar conclusions
are also inferred from the IPCC 4th Assessment Report on African climate change (IPCC 2007)
which found that temperatures will increase by up to 5.8 ºC by the end of the century in arid or
semi-arid areas that are prevalent in Africa. The report also added that sometime between 2080 and
the end of the century, average annual precipitation is very likely to decrease along Africa’s
Mediterranean coast by a fifth with declines also expected in the northern Sahara and the northern
west African coast. Declines are also predicted for much of southern Africa with the extreme west
of the region likely to experience falls by as much as forty percent through June and August. In
11
contrast, tropical and eastern Africa may experience increased rainfall of seven percent (IPCC
2007). As we will show below, the range of precipitation predictions by the IPCC/AR4 climate
models is generally consistent with this earlier expectation (IPCC 2001). In the remainder of this
paper, we more formally analyze these results from multiple models through computation of
multimodel ensemble averages.
Despite the variations in individual climate model predictions, over the entire Nile basin 8
(9) and 3 (6) of the 11 GCMs for the A2 (B1) global emission scenario predicted increases in
precipitation for 2010-2039 and 2070-2099, respectively (Tables 2 and 3). Multimodel average
annual Nile basin precipitation changes in percentages of historical (1950-1999) precipitation are
115 (117), 98 (104) and 93 (96) for the A2 (B1) emissions scenario with observed spread among
ensemble member prediction expressed as inner-quartile range (1st,3rd) for A2 emission scenario of
(-2, 26),(-12, 12) and (-20, 3. The inner-quartile ranges for scenario B1 are (5, 30),(-5, 15) and (-10,
6). The multimodel ensemble average annual precipitation changes for the Blue Nile sub-basin
expressed as a percentage of 1950-99 precipitation are 115 (117), 98 (104) and 106 (96) for the A2
(B1) global emission scenario and periods I to III, respectively. The changes for the Lake Victoria
sub-basin are 117 (123), 98 (105) and 90 (97) relative to the historical average (1950-1999). Spatial
variations in these changes over the Blue Nile sub-basin are noticeable, with most of the increases
occurring in the Atbara and Upper Blue Nile (Figure 5 (a-c)). A relatively similar spread pattern is
depicted within the sub-basins as for the entire Nile basin; the inner-quartile ranges are shown in
Figure 4 (d–f) for precipitation changes.
Tables 2 (a-c) and 3 (a-c) summarize the annual average changes in precipitation and
temperature for the entire Nile basin and the two sub-basins. We evaluate the end points in
precipitation projections among the 11 GCMs used in this study. The GISS (Goddard Institute for
Space Studies) and GFDL (Geophysical Fluid Dynamics Laboratory) climate models represent the
end points of the range of precipitation predictions. GISS is the wettest (precipitation increases
12
ranging between 40-58%) whereas the GFDL is the driest with predicted decreases in precipitation
of up to 34% for the period 2070-2100. The corresponding projected temperature increases range
from 1.0 to 6.0 °C for GISS and 1.1 to 3.7 °C for GFDL for the entire Nile basin.
We also analyzed seasonal precipitation changes for the wet (JJA) and dry (DJF) seasons.
As shown in Figures 5 (a-d) and summarized in Table 5, strong increases in both winter and
summer precipitation are predicted in the ensemble mean across the Nile basin particularly for
period I, although the magnitude and spatial patterns vary by sub-basin mainly due to the spatial
heterogeneity and hydroclimatic variations associated with general circulation and teleconnections.
Dominant JJA precipitation increases are predicted for the Blue Nile basin, mainly the Upper Blue
Nile. This might be associated with increased monsoon intensity in the Ethiopian highlands, which
is particularly important for runoff generation given that under current climate the Nile River at El
Diem receives more than 70% of its annual flow during the summer season from the Ethiopian
highlands (Mishrah et al., 2005). This implies in turn that increases in summer precipitation in this
part of the basin will result in substantial increases in the Nile River flow at HAD.
3.2 Streamflow changes
Conway and Hulme (1993) and Calder et al. (1995) found that the combined effects of
precipitation and temperature changes would have profound effects on the streamflow regime of
the Nile River.
In this section, we analyze the inferred changes in streamflow that would
accompany the precipitation and temperature changes summarized in Section 3.1. Table 2 shows
the projected changes in Nile River flow at the main stem Nile gauging station at HAD and the
Blue Nile flow at El diem (located upstream from the Roseires Reservoir in Sudan; see Figures 1a
and b) predicted using the methods summarized in Section 2. Although there is a consistent decline
in streamflow in the ensemble mean and for most of the GCMs by the end of the 21st century, the
specific magnitudes of the changes differ substantially from GCM to GCM.
13
Figures 6(a-b) and 6 (c-d) show changes in predicted long-term mean monthly Nile flow at
HAD and Blue Nile flow at El diem, respectively, for periods I and III for global emissions
scenario A2. Using the multimodel approach, the ensemble mean predicted annual streamflow for
Periods I-III for the main stem Nile at HAD are 111 (114), 92 (93) and 84 (87) as percentages of
historical (1950-1999) streamflow with corresponding multimodel spread expressed as innerquartile (1st, 3rd) ranges of (-2, 23),(-16, 5) and (-28, -5) for the three periods, respectively. The
same numbers for Blue Nile at El diem are 118 (120), 97 (99) and 93 (101) (see also Tables 6 and
7) with associated spread of (6, 31), (-14, 6) and (-21, 0) for the A2 global emissions scenario and
Periods I-III, respectively. Similar to temperature and precipitation, we evaluated the end points.
GISS predictions for Blue Nile mean annual flow at El diem for the A2 (B1) global emission
scenarios are 148 (154), 133 (134), and 126 (125); whereas the GFDL predictions are 83 (86), 77
(75) and 71 (70). Similarly, Nile streamflow at HAD is predicted to change to 134 (136), 126 (127)
and 112 (115) percent of the 1950-99 mean for GISS, and 77 (92), 71 (85) and 62 (75) percent by
GFDL for the A2 (B1) global emission scenarios and periods I-III, respectively.
These scenarios suggest major changes in the current hydrologic characteristics of the Nile
basin. Early in the century, multimodel average annual flows exceed the historical (1950-1999)
long-term average of 88 BCM at HAD; with 7 (7) GCMs in period I exceeding the 1950-99
average, whereas only 3 (4) of the models suggest increases in period III for the A2(B1) global
emissions scenario. Similarly, 9 (8) models predict increases in Blue Nile flow at El diem early in
the century (period I ), and 6 (5) for the A2 (B1) global emissions scenario in period III. The inter
model variation substantially differs as we go from period I to III for both emissions scenarios with
consistent negative trends in the longterm average streamflow. tThese spreads are presented in
Figure 4 (g-h).
14
Previous studies of the impact of climate change on Nile River flow have been confounded
by inconsistencies in emissions scenarios and other aspects of the model simulations and partially
for that reason have produced widely divergent results. Although there is considerable divergence
among ensembles (as well as models) in our results, there is nonetheless some general consistency.
Furthermore, even though other recent studies differ in the emissions scenarios used and other
aspects of the model simulations, there is some qualitative agreement between the results of these
studies and ours. For instance, Arnell's (1999) study suggested that precipitation in the Nile basin
would increase by about 10% by 2050, but he suggested that this increased precipitation would be
offset by increased evapotranspiration, implying that the net effect on the main stem flow might be
insignificant. Conway and Hulme (1996) used hydrologic models of the Blue Nile and Lake
Victoria basins similar to the water balance model of Piper et al (1986). By combining changes in
Lake Victoria outflows with changes in runoff in the other Nile sub-basins, they obtained a range of
–9% to +12% in changes in mean annual Nile flows at HAD for 2025.
Strzepek et al. (1995) assessed the impact of climate change using three GCMs (UKMO,
GFDL and GISS) with doubled global atmospheric concentrations (2xCO2) to predict Nile flow
changes at HAD for 2060. In their results, GISS was the wettest model and had a 30% increase in
annual streamflow; whereas for UKMO, there was a 12% decrease; and for the driest model GFDL,
there was a 78% decrease. Yates et al. (1998a) found declines up to -9% in the annual flow at HAD
by 2060 for doubled CO2, but they found increases for the GISSA and UKMO models for the same
period which produced about 40% increase in annual flow at HAD. Yates et al. (1998b) reanalyzed
the results of their CO2 doubling scenarios, and found that five of six GCMs showed increased
streamflows at HAD for the 2060s (roughly the time for CO2 doubling) with increases as much as
137% for GISS. Only one GCM (GFDL) showed a decline in annual flow at Aswan (-15%) relative
to the long-term average Nile flow.
15
In summary, the multimodel approach we used showed general agreement with the previous
studies in that temperatures will rise throughout the Nile basin, but there is considerable uncertainty
and disparity in spatial and temporal predictions of changes in precipitation as to both magnitude
and direction. This makes the analysis of implications of these changes for streamflow more
complicated and uncertain. To provide a sense of the characteristics of the multimodel ensemble
streamflow predictions, we summarize our streamflow prediction results in Tables 7 and 8. Theses
results show that for a change in precipitation of +15 (+17)% associated with warming of +1.5
(+1.3) ºC for the period 2010-2039 for the A2 (B1) global emissions scenario, annual Nile flow at
HAD will change by +11 (14)%. However, for a precipitation change of -7 (-4)% and warming of
+4.4 (+3.6) ºC for the period 2070-2099, mean annual Nile flow will experience changes on the
order of -16 (-13)% for the A2 (B1) global emissions scenario. Clearly the underlying assumptions
of the global emissions scenarios are manifested in the hydrologic response of the river system; the
global emissions scenario A2 seems to be the most realistic given the fact that the Nile river basin
economic development and population growth will change greatly over the next century.
3.3
Implications of climate change to water resource management of the Nile River
Since water resources are inextricably linked with climate, the prospect of global climate
change raises serious concerns as to implications for water resources and regional development
(Riebsame et al., 1995). Efforts to provide adequate water resources in the Nile rive basin will
confront several challenges over the next century, including population pressure and resulting land
use changes, with potential hydroecological consequences. Climate change and climate variability
(which results in droughts and flooding) will make addressing the pressing issue of water resources
management in the Nile basin more complex. According to the IPCC (1998), the Nile River
experienced reductions in runoff of 20% between 1972 and 1987, which leads to significant
16
interruptions in hydropower generation. A 1995 study by Reibsame et al. (1995) also concluded
that among the several major rivers studied, the Nile was the most susceptible to climate change in
terms of its potential for hydropower production.
To identify the nature of the interactions between climate change and the managed water
resource system, we used a simple reservoir operations model which had the objective of meeting
target hydropower and irrigation releases at High Aswan Dam (HAD). We term this model the
Lake Nasser Reservoir Model (LNRM).
LNRM is a simulation model which includes and
represents the major physical water management structures, reservoir operation rule curves and
water use policies of the HAD system.
In constructing LNRM, we made two assumptions. First we based operation of the HAD
system on the provisions of the 1959 treaty between Sudan and Egypt, a treaty that was not agreed
to or recognized by other riparian countries (including Ethiopia, which is the headwater source for
more than 86% of the Nile’s water). The treaty awarded Sudan 25% (18.5 BCM) of the mean
annual naturalized Nile flow at HAD of 84 BCM and 75% (55.5 BCM) to Egypt with any flow in
excess of the mean annual flow of 84 BCM to be shared equally between Egypt and Sudan (Said
1993; Johnson and Curtis 1994). The calculation includes 10 BCM which is estimated to be the
annual evaporation loss from the Sudd Swamps and reservoirs within the system.
The second assumption is based on an assessment of flow and hydropower impacts of two
development scenarios in the Blue Nile Basin by the Nile Decision Support System (DSS) of
Georgakakos (2004). The first scenario in the DSS was the current and existing conditions of river
flow and energy production whereas the second scenario was based on development of four major
hydropower plants (Progressive hydropower development in Ethiopia scenario), thus we have not
included future upstream development in our modelling approach. Georgakakos’ analysis was
based on the assumption of climate stationarity.
LNRM includes data for: (1) Lake Nasser
reservoir and HAD characteristics, (2) simulated river flows (reservoir inflows), (3) existing
17
installed hydropower facility, (4) water use policies, and (5) reservoir operation rules. All the
sequences of data flow are linked using flow network nodes, and represent locations of local inflow
and/or water withdrawals and returns. LNRM is driven by naturalized streamflow simulated using
our VIC land surface hydrologic model. The system basically models Sudan’s complex water
management system upstream of Lake Nasser as a single, lumped withdrawal upstream of the
inflow to Lake Nasser. LNRM calculates evaporative loss from the lake surface based on reservoir
surface area which is derived as a nonlinear function of head and reservoir storage, and models
hydropower production and irrigation water release as a function of inflow, minimum and
maximum release constraints, reservoir elevation, tailwater elevation, turbine efficiency, and
reservoir storage.
The hundred-year time series of monthly inflow records associated with each GCM
represents the dynamic supply to the reservoir, in which the initial storage in Lake Nasser was set
to its actual 1999 value of 162.5 BCM (reservoir elevation 182.1 m). The reservoir storage is
linked to numerous converters used to describe the relationships resulting from the particular state
of the storage. They produce reservoir elevation, power generation, spill, and evaporation losses.
Required release from the reservoir is linked to the set of converters and together they describe the
water release requirements from the reservoir for irrigation water supply. To make the LNRM
simulation realistic, the irrigation water demand was assumed to be supply driven and dynamic in
nature, which allows future agricultural sector expansion, with mandated 5%, 10% and 15% annual
average increases in irrigation water demand for Periods I - III, respectively. The model also allows
for cutbacks in the demand by percentages which are designed to overcome water shortages.
The monthly streamflows associated with each GCM were translated first into a time series
that accounts for upstream monthly withdrawals. The monthly withdrawal associated with each
climate model’s streamflow sequence was subtracted as lumped single withdrawal prior to running
LNRM. Such an approach was also used by Strzepek et al. (1996). LNRM system objectives
18
include meeting water supply targets and avoiding water shortages (for irrigation), avoiding
excessive spills and generating as much firm and average energy as possible.
Figure 7 shows a wide spread in the multimodel distribution of annual hydropower
production at HAD for the A2 and B1 global emission scenarios. The multimodel average
hydropower production is summarized in Table 7.
Much of the average mandated power
requirement is satisfied in period I, however predicted hydropower generation fails to meet the
annual average hydropower production targets for Periods II and III, mainly due to inflow
reductions to HAD. This is a significant issue, since the historical annual average hydropower
generated at HAD represents 1000 GWH, or 20% of total power generated within Egypt. The
multimodel annual average power production at HAD generally follows changes in streamflow,
increasing early in the century to 112 (118) percent of historical production for the period I A2 (B1)
emissions scenario in the multimodel ensemble, but then decreasing to 92 (97) and 87 (91) percent
of the historical mean for Periods II and III, respectively. In summary, the results indicate that
Egypt would maintain its historical annual hydropower production through Period I, covering 23%
(11500 GWH/yr) of its total energy production within the country, but will be limited to generating
18% (9000 GWH/yr) of its annual energy production by period III.
The irrigation water release results (Table 9) indicate that, averaged annually, Egypt will
satisfy its irrigation water requirement during Periods I with well above the historical irrigation
water supply, and will be able to meet the 5% presumed irrigation water demand increase during
Period I. In addition to the increased additional irrigable land, spills account for 4 BCM/yr through
Toshika to the west, which might be a potential source of irrigation water if properly utilized. In
general, the multimodel ensemble mean annual irrigation water release for the three periods and
two global emission scenarios A2 (B1) are 59 (60), 48 (49) and 48 (47) BCM, compared to the
historical release of 55.4 BCM (Table 9).
19
Irrigation water releases are not as affected by future climate in Periods II and III as
compared with hydropower generation, but nonetheless in periods II and III Egypt will suffer from
a reduction of 7 BCM (roughly equivalent to 457,000 ha of irrigable land). The nonlinearity in the
irrigation water and hydropower production tradeoffs is mainly due to the fact that hydropower
production and irrigation water releases are partially competing at HAD, and lower inflows are
magnified in hydropower production, which is dependent on both water level and reservoir inflows.
4.0 Conclusions
Temperature predictions for all GCMs considered show increases throughout the 21st century,
but the signature varies substantially from sub-basin to sub-basin and from GCM to GCM. In the
multimodel average over the entire Nile basin, warming increases to more than 3.5 ºC relative to
the historical (1950-1999) average by the end of Period III. Inner quartile ranges for the A2
emission scenario are (1.0, 1.2), (2.2, 2.8) and (3.4, 4.0) and for the B1 emission scenario are (0.9,
1.1), (1.7, 2.0) and (2.7, 3.2) for the three periods respectively. The multimodel average
precipitation (as percentages of the 1950-1999 mean for the A2 (B1) emission scenario) are 115
(117), 98 (104) and 106 (96) with associated inner quartile ranges of (-2, 26), (-12, 12) and (-20, 3).
For scenario B1 the inner-quartile ranges are (5, 30), (-5, 15) and (-10, 6). Similar analysis is made
for the two main sub-basins and presented as boxplots in Figure 4. In general, the inner quartile
ranges become narrower from period I through period III, however the spatial variations from subbasin to sub-basin persist. In general, winter (DJF) precipitation increases are predicted (in the
multimodel mean) for equatorial Africa (Lake Victoria) early in the 21st century, and summer (JJA)
precipitation increases are predicted for eastern Africa (Blue Nile region) throughout the 21st
century.
The timing and magnitude of changes in temperature and precipitation are critical to the
hydrologic response of the Nile basin. The Nile River is expected to experience increases in
20
streamflow early in the century at both gauging stations studied, the Blue Nile at El diem and the
main stem Nile at HAD, mostly due to increased precipitation. Subsequently, streamflow is
expected to decline during periods II and III as a result of both precipitation declines and enhanced
evapotranspiration due to increased temperature. The predicted multimodel mean streamflow at
HAD (High Aswan Dam) are 111 (114), 92 (93) and 84 (87) as percentages of the historical (19501999) period for the A2 (B1) emissions scenario with corresponding inner-quartile ranges of (-2,
23), (-16, 5) and (-28, -5) for A2 emissions and (5, 26), (-13, 4) and (-28, -5) for B1 global
emissions scenario and periods I-III, respectively. The corresponding numbers for the Blue Nile at
El diem are 118 (120), 97 (99) and 93 (101) for the A2 (B1) global emission scenario with
associated inner-quartile ranges of (6, 31),(-14, 6) and (-21, 0) for A2 and (9, 27), (-14, 6) and (-7,
8.5) for the B1 global emission scenario and periods I-III, respectively.
The potential impacts of 21st century climate change on water resources of the Nile River
were assessed using the simulated 21st century streamflows to drive a reservoir operations model
for Lake Nasser. The simulations showed that in the multimodel average, Egypt’s annual energy
production at HAD will remain relatively unaffected by climate change early in the 21st century
(2010-2039). However, Egypt will suffer significant reductions in hydropower production by mid
century despite projected increases in energy demand. The agricultural sector will experience
shortfalls relative to historical average irrigation releases from Lake Nasser by mid century due
increased evapotranspiration and hence reduced reservoir inflows, and increased evaporation from
the Sudd swamps and Lake Nasser surface, as well as reductions in precipitation. Overall, our
analysis suggests that well there will be some near-term benefits from climate change, primarily
associated with increased precipitation, but by mid century, management of Nile water resources
will become more challenging and complex. Furthermore, the uncertainty in climate change
predictions (particularly rainfall patterns in the basin) and the complexity of water management
21
issues facing basin water users will place a premium on more integrative approaches to dealing
with water supply and demand changes in the basin (Conway, 2005; Stern, 2007).
22
List of figures
Figure 1: Simulated and observed Nile River flow at selected gauging stations for the period 19661999 a) Main Nile River flow at DONGOLA (a) and b) Upper Blue Nile at Rosiers Upstream of El
diem, Sudan for the period 1970-1999. Validation period shown excludes the period 1950-1965
used for model calibration. ............................................................................................................30
Figure 2: Head waters and sub-basins of the main stem Nile (panel a) and locations of the two
gauging stations and flow routing network derived from ½ degree (latitude-longitude) DEM (panel
b). .................................................................................................................................................31
Figure 3: Downscaled and bias-corrected long-term (30yr) mean monthly temperature,
precipitation, simulated runoff, evapotranspiration, and soil moisture for a) the entire Nile basin b)
Blue Nile basin, and c) Lake Victoria region. The thick line (black) denotes the base historical
(1950-1999) period. ......................................................................................................................32
Figure 4: Ranges of temperature change among ensemble members( panel a) for the period 20102039, panel b) the same except for the period 2040-2069 and similarly panel c) 2070-2099. Panels
d), e) and f) show ranges of precipitation change as percentages for the three periods respectively.
Letters V, B, and N in the panels represent Lake Victoria region, Blue Nile, and entire Nile basin
respectively, where as letters A2 and B1 represent the two global emission scenarios A2 and B1.
The lower two panels show ranges of streamflow change for main Nile flow at HAD and Blue Nile
flow at El diem for three periods and two global emission scenarios .............................................33
Figure 5: Seasonal changes in precipitation and runoff for wet season (JJA) and dry season (DJF).
Panel a) shows A2 emission scenario JJA precipitation changes relative to mean historical (19501999), upper left panel shows observed JJA 1950-99 precipitation, upper right panel shows changes
in period 1, lower left shows changes in period 2, and lower right shows changes in period 3. Panel
b): same as panel a), but for runoff; Panel c), same as panel a) except for DJF precipitation; Panel
d), same as panel a ) except for DJF runoff....................................................................................34
23
Figure 6: Simulated mean seasonal cycle of streamflow derived from all 11 GCMs for A2
emissions scenario at the main Nile stem gauging station HAD (a, b) for the beginning (2010-2039)
and end of the 21st century (2070-2099) and at Blue Nile gauging station (El diem –Sudan) for the
same time period (c ,d)..................................................................................................................35
Figure 7: Projected mult-imodel ensemble average hydropower production at HAD for the base
(historical) period and three future periods, period I (2010-2039), period II ( 2040-2069) and
period III ( 2070-2099) for A2 and B1 global emission scenarios. .................................................36
24
References
Adam, J.C. and D.P. Lettenmaier (2003) Adjustment of global gridded precipitation for systematic
bias. J. Geophys. Res 108:1-14.
Arnell, N.W (1999)Climate change and global water resources. Global Environmental Change 9:
S31–S50.
Swain, A (1997) The Nile River dispute: Ethiopia, the Sudan, and Egypt. Journal of Modern
African Studies 35: 675-694.
Camberlin, P (1996)Rainfall anomalies in the source region of the Nile and their connection with
the Indian summer monsoon. J. Climate 10:1380-1392.
Christensen, N. S., Wood, A. W., Lettenmaier, D. P., and Palmer, R. N (2004) Effects of climate
change on the hydrology and water resources of the Colorado River Basin.
Climatic
Change 62:337–363.
Christensen, N., and D.P. Lettenmaier (2007) A multimodel ensemble approach to assessment of
climate change impacts on the hydrology and water resources of the Colorado River basin.
Hydrology and Earth System Sciences (in press)
Conway, D (2005) From headwater tributaries to international river: observing and adapting
to climate variability and change in the Nile Basin. Global Environemenatal Change 15:99114
Conway D, and M. Hulme (1993) Recent fluctuations in precipitation and runoff over the Nile subbasins and their impact on main Nile discharge.Climatic Change 25:127 151.
Conway, D. and M. Hulme (1996) The impacts of climate variability and future climate change in
the Nile basin on water resources in Egypt. Water Resources Development 12:277-296.
Georgakakos, A (2004) Decision support systems for integrated water resources management with
an application to the Nile Basin: IFAC Modelling and Control for Participatory Planning
and Managing Water Systems. Venice, Italy: International Federation of Automatic Control
25
Hassan, F.A. (1998) Climatic change, Nile floods and civilization. Nature and Resources 34(2): 3440
Hayhoe, K., C.P. Wake, T.G. Huntington, L.-F. Luo, M.D. Schwartz, J. Sheffield, E.F Wood, B.
Anderson, J. Bradbury, A. DeGaetano, T.J. Troy, and D. Wolfe (2007) Past and future
changes in climate and hydrological indicators in the U.S. Northeast. Climate Dynamics 28:
381-407.
Hulme, M., R. Doherty, T. Ngara and M. New (2005) Global Warming and African Climate
change: Climate Change and Africa: A reassessment, Cambridge University Press, 338pp.
Hulme, M (1992) Recent and future precipitation changes over the Nile Basin. in A. Zeid and A.K.
Biswas (ed) Climatic fluctuations and water management. Butterworth Heinemann 1:187201
Hulme, M (1990) “Global climate change and the Nile Basin” The Nile: Resource Evaluation, P.P.
Howell and J.A. Allen (eds), Resource Management, Hydro-politics and Legal Issues,
Centre of Near and Middle-Eastern Studies, SOAS, London.. 238pp
IPCC (2001) Climate Change 2001: The Scientific Basis Contribution of Working Group I to
the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) J.
T. Houghton, Y. Ding, D.J. Griggs, M. Noguer, P. J. van der Linden and D. Xiaosu (Eds.)
Cambridge University Press, UK. pp 944
IPCC (2007) Climate Change 2007: Impacts, Adaptation and Vulnerability.Contribution of
Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on
Climate Change (IPCC), Parry M.L., O.F. Canziani, J.P. Palutikof, P.J. van der Linden and
C.E. Hanson,(Eds) Cambridge University Press,Cambridge, UK, 1000pp.
Johns, T.C., Gregory, J.M., Ingram, W.J., Johnson, C.E., Jones, A., Lowe, J.A., Mitchell, J.F.B.,
Roberts, D.L., Sexton, D.M.H., Stevenson, D.S., Tett, S.F.B., and Woodage, M.J (2003)
26
Anthropogenic climate change for 1860 to 2100 simulated with the HadCM3 model under
updated emissions scenarios. Climate Dynamics 20: 583-612.
Johnson, P. and P. Curtis (1994) Water balance of Blue Nile River Basin in Ethiopia. Journal of
Irrigation and Drainage Engineering 120(3): 573-590.
Krishnamurti, T.N., and coauthors (2000) Multimodel ensemble forecasts for weather and seasonal
climate, J. Climate 13: 4196-4216.
Lettenmaier, D.P., Wood, A.W., Palmer, R. N., Wood, E. F., Stakhiv, E. Z (1999) Water
esources implications of global warming : A U.S. regional perspective. Climatic Change
43:537-579
Liang, X., D.P. Lettenmaier, E.F. Wood, and S. J. Burges (1994) “A simple hydrologically based
model of land surface water and energy fluxes for general circulation models. J. Geophys.
Res 99 (D7) 14:415-14,428
Liang, X., Wood, E. F., and Lettenmaier, D. P (1996) ‘Surface Soil Moisture Parameterization of
the VIC-2L Model: Evaluation and Modifications’.Global Planetary Change 13:195–206.
Maurer, E. P (2007) Uncertainty in hydrologic impacts of climate change in the Sierra Nevada,
California under two emissions scenarios. Climatic Change 82:309-325.
Mohamed, E (2005) Egyptian electricity sector an overview .ESI AFRICA ISSUE 4:38-40.
Mohamed, Y. A., B.J.J.M. van den Hurk, H.H.G. Savenijeand, and W.G.M Bastiaanssen (2005)
Hydroclimatology of the Nile: results from a regional climate model. J. Hydrology and
Earth System Sciences 9:263-278.
Nijssen, B., D.P. Lettenmaier, X. Liang, S.W. Wetzel, E.F. Wood (1997) Multimodel ensemble
forecasts for weather and seasonal climate. Water Res. Research. 33 (4):711-724.
Nile Basin Initiative (NBI) Shared Vision Program 2001(Report) on Nile River Basin:
Transboundary Environmental Analysis, United Nations Development Programme World
Bank and Global Environment Facility
27
Payne, J.T., Wood, A.W., Hamlet, A.F., Palmer, R.N. and Lettenmaier, D.P (2004) Mitigating the
effect of climate change and water resources of Colombia River basin. Climatic Change,
62(1-3):233-256.
Riebsame, W.E., K.M. Strzepek, J.L. Wescoat Jr., R. Perrit, G.L. Graile, J. Jacobs, R. Leichenko,
C. Magadza, H. Phien, B.J. Urbiztondo, P. Restrepo, W.R. Rose, M. Saleh, L.H. Ti, C.
Tucci, and D. Yates (1995) Complex river basins. in: As Climate Changes, International
Impacts and Implications [Strzepek, K.M. and J.B. Smith (eds.)]. Cambridge University
Press, Cambridge, United Kingdom and New York, NY, USA, pp. 57–91
Said, R (1993) The River Nile: Geology, hydrology and utilization. Amsterdam Pergamon Press
Sene, K.J., Tate, E.L. and Farquharson, F.A.K, (2001) Sensitivity studies of the impacts of climate
change on White Nile flows. Climatic Change 50:177-208.
Stern, N (2007) The Economics of Climate Change: The Stern Review. Cambridge
University Press, Cambridge, 692 pp.
Strzepek, Kenneth M. and David N. Yates (2000) Responses and thresholds of the Egyptian
economy to climate change impacts on the water resources of the Nile River. Climatic
Change 46:339-356.
Strzepek, K.M., S.C. Onyeji, M. Saleh, and D.N. Yates (1995) An assessment of integrated climate
change impacts on Egypt. In: As Climate Changes: International Impacts and Implications.
[Strzepek, K.M. and J.B. Smith (eds.)]. Cambridge University Press, Cambridge, United
Kingdom, pp. 180-200.
Tate, E., Sutcliffe, J., Declan, C., and Farquharson, F (2004) Water Balance of Lake Victoria:
Update to 2000 and Climate change Modeling to 2100.Hydrology and Earth System
Sciences 49(4):563–574.
28
Wood, A. W., E. P. Maurer, A. Kumar, and D. P. Lettenmaier (2002) Long-range experimental
hydrologic forecasting for the eastern United States.J. Geophysical Research-Atmospheres
107(D20):4429.
Wood, A. W., L. R. Leung, V. Sridhar, and D. P. Lettenmaier (2004) Hydrologic implications of
dynamical and statistical approaches to downscaling climate model outputs. Climatic
Change 62:189-216.
Y. A. Mohamed, B. J. J. M. van den Hurk2, H. H. G. Savenije, and W. G. M. Bastiaanssen, (2005)
Hydroclimatology of the Nile: results from a regional climate model. J Hydrology and
Earth System Sciences 9:263–278.
Yates, David N. and Kenneth M. Strzepek (1998a) Assessment of integrated climate change
impacts on the agricultural economy of Egypt. Climatic Change 38:261-287.
Yates, D.N. and K.M. Strzepek(1998b) Modeling the Nile basin under climatic change. Journal of
Hydrologic Engineering 3(2):98-108.
29
Figure 1: Simulated and observed Nile River flow at selected gauging stations for the period 19661999 a) Main Nile River flow at DONGOLA (a) and b) Upper Blue Nile at Rosiers Upstream of El
diem, Sudan for the period 1970-1999. Validation period shown excludes the period 1950-1965
used for model calibration.
30
Figure 2: Head waters and sub-basins of the main stem Nile (panel a) and locations of the two
gauging stations and flow routing network derived from ½ degree (latitude-longitude) DEM (panel
b).
31
Figure 3: Downscaled and bias-corrected long-term (30yr) mean monthly temperature,
precipitation, simulated runoff, evapotranspiration, and soil moisture for a) the entire Nile basin b)
Blue Nile basin, and c) Lake Victoria region. The thick line (black) denotes the base historical
(1950-1999) period.
32
Figure 4: Ranges of temperature change among ensemble members( panel a) for the period 20102039, panel b) the same except for the period 2040-2069 and similarly panel c) 2070-2099. Panels
d), e) and f) show ranges of precipitation change as percentages for the three periods respectively.
Letters V, B, and N in the panels represent Lake Victoria region, Blue Nile, and entire Nile basin
respectively, where as letters A2 and B1 represent the two global emission scenarios A2 and B1.
The lower two panels show ranges of streamflow change for main Nile flow at HAD and Blue Nile
flow at El diem for three periods and two global emission scenarios .
33
Figure 5: Seasonal changes in precipitation and runoff for wet season (JJA) and dry season (DJF).
Panel a) shows A2 emission scenario JJA precipitation changes relative to mean historical (19501999), upper left panel shows observed JJA 1950-99 precipitation, upper right panel shows changes
in period 1, lower left shows changes in period 2, and lower right shows changes in period 3. Panel
b): same as panel a), but for runoff; Panel c), same as panel a) except for DJF precipitation; Panel
d), same as panel a ) except for DJF runoff.
34
Figure 6: Simulated mean seasonal cycle of streamflow derived from all 11 GCMs for A2
emissions scenario at the main Nile stem gauging station HAD (a, b) for the beginning (2010-2039)
and end of the 21st century (2070-2099) and at Blue Nile gauging station (El diem –Sudan) for the
same time period (c ,d).
35
Figure 7: Projected mult-imodel ensemble average hydropower production at HAD for the base
(historical) period and three future periods, period I (2010-2039), period II ( 2040-2069) and
period III ( 2070-2099) for A2 and B1 global emission scenarios.
36
Table 1.Description of GCMs (General Circulation Models)
GCM
Modeling Group, Country
IPCC Model I.D.
Reference
CNRM
Centre National de
CNRM-CM3
Salas-Mélia et
Recherches Météoroliques,
al.,2005
France
CSIRO
CSIRO Atmospheric
CSIRO-Mk3.0
Research, Australia
GFDL
Geophysical Fluid Dynamics
Gordon, H.B. et al.,
2002
GFDL-CM2.0
Delworth et al., 2005
GISS-ER
Russell et al., 1995,
Laboratory, USA
GISS
Goddard Institute for Space
Studies, USA
HADCM3
Hadley Center for Climate
2000
UKMO-HadCM3
and Prediction and Research,
Gordon, C. et al.,
2002
UK
INMCM
Institute for Numerical
INM-CM3.0
Mathematics, Russia
IPSL
Institut Pierre Simon Laplace,
Diansky and
Volodin, 2002
IPSL-CM4
IPSL, 2005
MIROC3.2
K-1 model
France
MIROC
Center for Climate Systems
Research, Japan
MPI
MRI
developers, 2004
Max Planck Institute for
ECHAM5 / MPI-
Jungclaus et al., 2005
Meteorology, Germany
OM
Meteorological Research
MRI-CGCM2.3.2
Yukimoto et al., 2001
PCM
Washington et al.,
Institute, Japan
PCM
National Center for
Atmospheric Research, USA
2000
37
Table 2(a): 2010-2039 multimodel annual average change in temperature (ºC) relative to 19501999 historical average for sub-basins of the Nile basin and the entire basin for each GCM and the
A2 (B1) global emissions scenario.
Region
GISS
GFDL CSIRO CNRM MPI
MRI
MIROC
IPSL
INMCM HadCM3 PCM
Lake
0.99
1.11
0.97
1.33
1.05
1.22
0.87
1.12
1.02
(0.85) (1.23)
(0.92)
(1.09)
(0.96) (0.89) (1.16)
1.51
1.57
1.5
1.34
1.09
(1.38) (1.25)
(1.44)
(1.1)
(1.3)
(0.78) (1.66)
Entire
1.64
1.77
1.63
1.27
1.41
Nile
(1.28) (1.35)
(1.52)
(1.23)
(1.17) (0.94) (1.37)
Victoria
Blue
Nile
1.42
1.55
1.76
1.95
1.28
0.72
(1.03) (1.15)
(1.08)
(0.65)
1.65
1.35
0.96
(1.49) (0.9)
(1.12)
(0.99)
1.43
1.35
0.91
(1.12)
(1.09)
1.13
1.38
(1.23) (1.01)
Table 2(b): 2040-2069 multimodel annual average change in temperature (ºC) relative to 19501999 historical average for sub-basins of the Nile basin and the entire basin for each GCM and the
A2 (B1) global emissions scenario.
Region
GISS
GFDL CSIRO CNRM MPI
MRI
MIROC IPSL
INMCM HadCM3 PCM
Lake
2.01
1.89
2.67
3.23
2.96
2.2
2.57
2.45
(1.55) (1.44)
(1.62)
(2.05)
(1.95) (1.94) (2.46)
2.72
3.15
3.13
2.71
(1.55) (2.87)
(2.62)
(2.33)
(2.65) (2.67) (2.79)
Entire
3.0
3.47
3.77
3.59
Nile
(1.97) (2.73)
(2.84)
(2.66)
(2.58) (2.69) (3.43)
Victoria
Blue
Nile
3.21
3.67
3.56
3.78
3.54
3.99
2.75
2.76
2.22
(2.56) (1.78)
(1.68)
(1.95)
3.05
3.26
2.54
(2.83) (2.76)
(2.83)
(1.99)
3.31
3.75
2.77
(3.22)
(2.25)
3.25
3.69
(2.97) (3.09)
38
Table 2(c): 2070-2099 multimodel annual average change in temperature (ºC) relative to 19501999 historical average for sub-basins of the Nile basin and the entire basin for each GCM and the
A2 (B1) global emissions scenario.
Region
GISS GFDL CSIRO CNRM MPI
MRI
MIROC IPSL INMCM HadCM3 PCM
Lake
3.95
3.47
3.93
3.44
Victoria (3.22) (2.94)
3.23
3.32
3.88
(3.11)
(2.81)
(3.17) (3.05) (3.38)
4.56
3.91
4.01
2.39
(2.66) (2.18)
(3.26)
(2.25)
3.85
4.66
2.84
Blue
4.72
3.21
3.89
3.83
4.71
Nile
(3.40)
(3.87)
(3.42)
(3.21)
(3.38) (3.41) (3.79)
(3.13) (3.76)
(3.83)
(2.87)
Entire
5.04
3.67
4.53
4.77
4.87
4.56
4.21
3.38
Nile
(3.74)
(3.37)
(3.79)
(3.64)
(3.81) (3.59) (4.13)
(3.99)
(3.01)
4.88
5.54
3.37
5.86
4.23
4.49
(3.45) (3.58)
Table 3(a): 2010-2039 multimodel annual average change in precipitation (%) relative to 19501999 historical average for sub-basins of the Nile basin and the entire basin for each GCM and the
A2 (B1) global emissions scenario.
Region
GISS GFDL CSIRO CNRM MPI MRI MIROC IPSL INMCM HadCM3 PCM
Lake
141
82
135
128
124
115
122
131
97
96
124
Victoria
(147)
(92)
(148)
(139)
(129)
(110)
(132)
(135)
(101)
(100)
(125)
Blue
153
87
118
109
119
133
128
107
93
102
123
Nile
(158)
(96)
(121)
(119)
(114)
(139)
(135)
(122)
(90)
(109)
(118)
Entire
140
84
131
107
127
101
105
125
92
95
120
Nile
(142)
(90)
(138)
(114)
(133)
(110)
(109)
(126)
(97)
(101)
(126)
39
Table 3(b): 2040-2069 multimodel annual average change in precipitation (%) relative to 19501999 historical average for sub-basins of the Nile basin and the entire basin for each GCM and the
A2 (B1) global emissions scenario.
Region
GISS GFDL CSIRO CNRM MPI
MRI
MIROC IPSL INMCM HadCM3 PCM
Lake
131
68
104
105
103
98
107
Victoria
(137)
(87)
(112)
(109)
(106) (100) (112)
Blue Nile
124
66
89
87
105
(128)
(72)
(101)
(111)
(102) (119) (121)
Entire
120
62
113
94
109
Nile
(124)
(77)
(114)
(103)
(117) (99)
100
89
78
100
(105) (92)
(81)
(111)
83
85
101
(107) (83)
(93)
(103)
93
115
88
110
(103)
(112) (79)
(91)
(115)
103
112
81
79
81
Table 3(c): 2070-2099 multimodel annual average change in precipitation (%) relative to 19501999 historical average for sub-basins of the Nile basin and the entire basin for each GCM and the
A2 (B1) global emissions scenario.
Region
GISS GFDL CSIRO CNRM MPI
MRI
MIROC IPSL
INMCM HadCM3 PCM
Lake
114
97
107
68
73
101
56
98
100
90
82
Victoria (121)
(68)
(113)
(104)
(94)
(101) (96)
(102)
(77)
(84)
(103)
Blue
137
73
99
126
89
121
93
96
97
114
Nile
(142)
(79)
(101)
(113)
(94)
(123) (111)
(103)
(91)
(116)
(98)
Entire
116
64
103
84
91
77
85
102
71
93
104
Nile
(113)
(72)
(108)
(93)
(100) (96)
(88)
(106)
(72)
(94)
(106)
113
40
Table 4: Multimodel average annual and seasonal (JJA and DJF) temperature changes as
differences in degrees (oC) of the historical (1950-1999) for two sub-basins of the Nile basin and
entire Nile basin for A2 (B1) global emission scenarios and three periods.
Scenario
Historical
Lake
Blue
Entire Nile basin
Victoria
Nile
Annual
DJF
JJA
19.8 oC
21.25
23.45 oC
22.11 oC
25.24 oC
o
C
A2 2010-39
1.06
1.25
1.48
1.38
1.16
A2 2040-69
2.51
3.1
3.22
2.89
2.45
A2 2070-99
3.45
4.18
4.36
3.90
3.60
B1 2010-39
1.0
1.22
1.31
1.02
1.15
B1 2040-69
1.9
2.55
2.76
2.78
2.50
B1 2070-99
2.91
3.37
3.65
3.54
3.58
41
Table 5:
Multimodel average annual and seasonal (JJA and DJF) precipitation changes as
percentages (mm) of the historical (1950-1999) for two sub-basins of the Nile basin and entire Nile
basin for the A2 (B1) global emissions scenario and three periods.
Scenario
Historical
Lake Victoria
Blue Nile
Entire Nile basin
Annual
DJF
JJA
15mm
55mm
116mm
98mm
70mm
A2 2010-39
117%(136mm)
116%(113mm)
112%(78mm) +25%(3.7mm)
+18%(10mm)
A2 2040-69
98.81%(115mm)
91%(89mm)
98%(68mm)
-8.2%(-1.2mm)
-5.4%(-3mm)
A2 2070-99
89.63%(104mm)
105%(103mm)
90%(63mm)
+4.7(0.69mm)
-2.3%(-0.12mm)
B1 2010-39
123.45%(143mm)
120%(117mm)
117%(82mm) +9.8%(1.44mm) +24%(13mm)
B1 2040-69
104.7%(121mm)
104%(102mm)
103%(72mm) -6.7%(1mm)
-2.6%(1.44mm)
B1 2070-99
96.63%(112mm)
106%(104mm)
95%(67mm)
+2.04 (1.14mm)
(1950-1999)
-1%(-0.15mm)
42
Table 6: Simulated streamflow by each climate model routed to the mainstem Nile gauging station
at High Aswan Dam (HAD) for periods I to III and global emission scenarios A2 (B1). Numbers
are expressed as percentages of the mean annual historical flow (1950-1999).
GCM
Period1
Period2
Period3
2010-2039
2040-2069
2070-2099
Gfdl
77(92)
71(85)
62(75)
Giss
134(136)
126(127)
112(115)
Csiro
108(127)
102(105)
107(112)
Inmcm
95(99)
82(83)
70(73)
Mpi
130(128)
112(105)
97(93)
Cnrm
112(123)
96(100)
83(85)
Ipsl
101(110)
87(90)
75(79)
Pcm
110(125)
94(103)
81(91)
Miroc
119(119)
102(98)
85(86)
Hadcm3
89(95)
77(78)
67(68)
Mri
126(118)
108(98)
93(95)
43
Table 7: Simulated streamflow for each climate model routed to the Blue Nile gauging station at El
diem (Sudan) for periods I to III and the global emission scenario A2 (B1). Numbers in parenthesis
are percentages of the mean annual historical flow (1950-1999).
GCM
Period1
Period2
Period3
2010-2039
2040-2069
2070-2099
Gfdl
83(87)
77(78)
71(72)
Giss
148(154)
133(136)
126(125)
Csiro
126(125)
102(102)
112(117)
Inmcm
100(106)
81(87)
74(81)
Mpi
113(112)
92(85)
83(91)
Cnrm
105(106)
86(91)
97(105)
Ipsl
131(130)
106(113)
96(105)
Pcm
127(125)
102(101)
103(112)
Miroc
130(129)
105(105)
95(104)
Hadcm3
106(115)
86(92)
76(95)
Mri
131(120)
106(99)
96(101)
44
Table 8: Multimodel High Aswan Dam hydropower production metrics
Scenario
Minimum 1st
Mean
Quartile
Historical
3rd
Max
Quartile
Percentage
Change Relative to
9.66 x 103
historical annual
GWH/yr
Hydropower production
(x103GWH/yr)
A2 2010-39
4.7
7.5
10.7
13.2
20.5
110.3%(1)
A2 2040-69
3.7
8.2
9.9
11.0
17.3
102.6%(0.25)
A2 2070-99
3.1
6.5
9.0
10.4
17.5
93.4%(-0.65)
B1 2010-39
4.8
8.1
11.4
13.2
21.6
117.6%(1.7)
B1 2040-69
4.3
8.8
10.5
11.7
18.8
108.2%(0.8)
B1 2070-99
2.3
6.8
88
10.1
16.6
91.6%(-0.84)
45
Table 9: Multimodel long term average monthly irrigation releases from High Aswan Dam (BCM)
for historical 1950-1999 and three future periods
A2_2010-
A2_2040-
A2_2070-
B1_2010-
B1_2040-
B1_2070-
2069
2099
Month
historical
2039
2069
2099
2039
Jan
3.50
3.71
3.06
3.01
3.82
3.12
2.94
Feb
4.00
4.24
3.49
3.44
4.36
3.56
3.36
Mar
4.20
4.45
3.67
3.61
4.58
3.74
3.53
Apr
4.00
4.24
3.49
3.44
4.36
3.56
3.36
May
5.30
5.62
4.63
4.56
5.78
4.72
4.45
June
6.50
6.89
5.67
5.59
7.09
5.79
5.46
July
7.00
7.42
6.11
6.02
7.63
6.23
5.88
Aug
6.30
6.68
5.50
5.42
6.87
5.61
5.29
Sep
4.30
4.56
3.75
3.70
4.69
3.83
3.61
Oct
3.70
3.92
3.23
3.18
4.03
3.29
3.11
Nov
3.60
3.82
3.14
3.10
3.92
3.20
3.02
Dec
3.00
3.18
2.62
2.58
3.27
2.67
2.52
Annual
55.40
58.72
48.36
47.64
60.39
49.31
46.54
46

Similar documents

×

Report this document