an integrated simulation system of meteorological and ... · 1) rainfall was observed only a few...
TRANSCRIPT
The 3rd International Conference on Water Resources and Arid Environments (2008)
and the 1st Arab Water Forum
An Integrated Simulation System of Meteorological and Hydrological Models for Predicting Dam Water Availability
Ryohji Ohba1, Adil Bushnak2, Toshiharu Kojiri3, Tomonori Matsuura4
and Haruyasu Nagai5 1. Mitsubishi Heavy Industries, Japan
2. Bushnak Water Group, Saudi Arabia 3. Kyoto University, Japan
4. National Research Institute for Earth Science and Disaster Prevention, Japan
5. Japan Atomic Energy Research Agency, Japan
Abstract
A Japanese national project team has developed an integrated simulation system to evaluate regional water cycles. The system consists of numerical models of rainfall, water discharge and ground water. These models numerically simulate the time-dependent relevant phenomena based on the conservation of momentum, heat and water.
The simulated results of rainfall amount, discharged water and ground water depth near Jeddah and Abha, Saudi Arabia are discussed in the present paper. Keyword: Water Cycle, Rainfall, Meteorology, Hydrology, Climate Change
Introduction
The Japanese Ministry of Education, Culture, Sports, Science, and Technology (MEXT) initiated a project in 2002 titled “Refinement of Numerical Modeling and Technology for Prediction of Global and Regional Water Cycles” (Global Water Cycle Project, http://www.kakushin21.jp/kyousei/) as a part of the Project for the Sustainable Coexistence of Humans, Nature and the Earth. The project was conducted between 2002 and 2006. The purpose of the project was
to create a master plan for the improvement of desert environments. Specifically, the project aimed for the creation of sustainable water cycles among the sea, land and atmosphere in coastal deserts such as the Asir region near the Red Sea. As the main contractor of the project, Mitsubishi Heavy Industries (MHI) organized an interdisciplinary project team consisting of Frontier Research System for Global Change (FRSGC), Kyoto University, Tottori University and Sophia University (see Fig. 1-1).
One of the findings of the project is that the annual variability of the sea surface temperature distribution of the Indian and Pacific Oceans, the Indian Ocean Dipole (IOD), and the El Niño-Southern Oscillation (ENSO) have the dominant influence on rainfall patterns through global-scale atmospheric teleconnection (Behera et al., 2005). In addition, the interaction between disturbances propagating from the Mediterranean Sea and moisture originating from the Red Sea was found to affect the severe regional rainfall events in Makkah and Asir regions. We have named this interaction and the resulting rainfall events the “Arabian Cyclone” (Chakraborty et al., 2006).
Wind turbines
Moisture collector
Meteorological simulation
Greening by water reuse
Photo-Voltaic system Steam
duct
Solar heat collector
Solar Desalination
system
Rain radar
Water pipe line
Satellite data
Cloud seeding
Humid convective flow
Global Climate Analysis(FRSGC)
Regional water cycle Analysis(Kyoto University)
Water Resource Development(MHI)
Oasis Creation(Tottori University)
Environmental Assessment(Sophia University)
Wind turbines
Moisture collector
Meteorological simulation
Greening by water reuse
Photo-Voltaic system Steam
duct
Solar heat collector
Solar Desalination
system
Rain radar
Water pipe line
Satellite data
Cloud seeding
Humid convective flow
Global Climate Analysis(FRSGC)
Regional water cycle Analysis(Kyoto University)
Water Resource Development(MHI)
Oasis Creation(Tottori University)
Environmental Assessment(Sophia University)
Fig. 1-1 Schematic of Global Water Cycle Project
In the Kingdom of Saudi Arabia, rainfall events are very rare and no river discharge data are available. Therefore, siting for a dam is a difficult task in this country.
The summary of the daily rainfall amount observed in Khamis Mushait (Fig. 1-2) shows:
1) Rainfall was observed only a few tens of days in spring, summer and winter. 2) Heavy rainfall occurs occasionally in spring and winter.
Khamis Mushait Rainfall during 2000
05
101520
1 30 59 88 117
146
175
204
233
262
291
320
349
Cale n d ar Day
Rai
nfal
l mm
Fig. 1-2 Daily rainfall amount in Khamis Mushait in 2000
Comparisons of monthly rainfall between Abha and Khamis Mushait (Fig. 1-3) imply a heterogeneous spatial distribution of rainfall amount in the Asir region. The two cities are separated by approximately 30 km. a) Abha b) Khamis Mushait
Fig. 1-3 Monthly rainfall averaged over approximately 30 years
There exist few meteorological observation sites in Saudi Arabia, and those that exist are situated mainly near big towns. Thus, meteorological
MEAN MONTHLY RAINFALL AT ABHA FROM 1978 TO 200
0
10
20
30
40
50
60
70
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
MONTH
MIL
LIM
ETER
MONTHLY MEAN RAINFALL AT KHAMIS FROM 1967 TO 2001
0
10
20
30
40
50
60
70
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
MONTH
MIL
LIM
ETER
observations are sparse in the country. Accordingly, the use of meteorological data is limited for estimating the volume of water that can be stored in dams. Total annual rainfall can be as much as six times greater in high-rainfall years than in low-rainfall years (Fig. 1-4). This variability is attributable mainly to changes in the global-scale sea surface temperature distribution.
YEARLY RAINFALL AT ABHA FROM 1979 TO 2001
0
100
200
300
400
500
600
700
Y79 Y80 Y81 Y82 Y83 Y84 Y85 Y86 Y87 Y88 Y89 Y90 Y91 Y92 Y93 Y94 Y95 Y96 Y97 Y98 Y99 Y00 Y01
YEAR
MIL
LIM
ETER
Fig. 1-4 Annual rainfall at Abha for 1979 – 2001
Based on simulation data, the IPCC (Intergovernmental Panel on Climate Change) predicted that year-to-year variation in rainfall amount will increase in the near future due to global warming (Fig. 1-5).
Fig. 1-5 Prediction of annual rainfall variation for 50 years with the influence of global warming (Ref.: the Canadian model, IPCC report)
2. Integrated simulation system (from rainfall to availability of dam water) 2.1 System components
The simulation system consists of simulation models for rainfall, water discharge and ground water. These models can numerically simulate the relevant phenomena continuously based on the conservation of momentum, heat and water (Fig. 2.1-1). Fig. 2.1-1 Flow chart for the simulation of rainfall, discharged water and ground water distributions 2.2 Meteorological model
Meteorological models such as MM5 and RAMS can simulate meteorological variables including temperature, wind velocity, humidity and rainfall both in time and space. The mesh size of these models is as fine as approximately 1km (Ter Maat et al., 2006). As an example, Fig. 2.2-1 shows comparisons of satellite-observed and numerically simulated horizontal distributions of clouds.
Input data Simulation models Output data
Geographical data
Global meteorological data from 1st
Jan. to 31st Dec.
Local meteorological
model
Horizontal rainfall
distribution
Hydrological model
Discharged water
distribution
Geological data
Ground watermodel
Ground water distribution
a) Satellite Image by EUMETSAT b) Numerical simulation by RAMS
Fig. 2.2-1 Comparison of satellite-observed and simulated cloud distributions
The three-dimensional distributions of clouds and rain were simulated for observed events from 9 to 12 November 2000 (Fig. 2.2-2). This result shows increased rainfall on mountaintops in Ethiopia and the Asir region due to the topographical effect. Simulation of meteorological phenomena over a one-year period by the high-speed parallel computer, “Earth Simulator” of Japan requires computational time of approximately one month.
Fig. 2.2-2 Simulated results of cloud (white)
and rainfall (blue) around the Red Sea in November 2000
Prediction accuracy of the Earth Simulator was evaluated using
meteorological data for the Asir region (Fig. 2.2-3).
16 March 2000 1200 UTC
a) Temperature b) Wind velocity
Fig. 2.2-3 Comparisons of hourly observations and simulation results by RAMS for Gizan between 21 February 2000 and 30 March 2000
Because of the significant spatial and temporal variation of rainfall,
comparison of the rainfall amount simulated by a numerical model to that observed at one point in space is inherently limited. With the awareness of this limitation, the monthly values of rainfall observed in-situ are compared to those simulated by MM5 (Fig. 2.2-4). Fig. 2.2-4 also includes the monthly values of rainfall that were evaluated by TRMM satellite data over a square of a few kilometers (Ueda et al., 2005).
Fig.2.2-4 Comparisons of the monthly surface precipitation (blue: simulation by MM5, pink: in-situ observation, yellow: satellite observation by TRMM)
y = 0.6881x + 0.3691R2 = 0.6402
0
2
4
6
8
10
12
0 2 4 6 8 10 12ECMWF observed [m/s]
RA
MS
win
dspe
ed a
t 10m
[m/s
]
y = 2.2793x - 387.36R2 = 0.6695
290
294
298
302
306
310
314
290 294 298 302 306 310 314ECMWF observed [K]
RA
MS
tem
p at
2m
[K]
2.3 Hydrological model
Hydrological models can simulate the spatial and temporal patterns of water discharge due to rainfall. The simulation takes into consideration the effects of topography and ground condition within a mesh size as fine as 1km (Fig. 2.3-1). The simulated distribution pattern of rainfall is complex (Fig. 2.3-2) and was validated using satellite-based flood water data (Fig. 2.3-3).
Fig. 2.3-1 Classification of surface types Fig.2.3-2 Simulated distribution of used for
simulation rainfall rate a) Simulated results b) Satellite-based observation
Fig. 2.3-3 Comparison between simulated and satellite-observed flood water distribution for 24 January 2005 (red squares outline the same region in both figures).
The volume of discharged water was calculated from the rainfall amount
simulated by the meteorological model under three kinds of ground surface conditions: mix of sand and shrubs (Fig. 2.3-4), sand only (Fig. 2.3-5) and
shrubs only (Fig. 2.3-6). The volume of water available to a dam can be estimated by integrating the values of discharged water over a period of 1 year or longer. a) Near Taif b) Near Mecca
0. 0
0. 5
1. 0
1. 5
2. 0
2. 5
3. 0
3. 5
4. 0
0 6 12 18 24 30 36 42 48
Ti me ( Hour )
Ru
n-
of
f
Ra
te
,
mm
/h
r
0
5
10
15
20
25
Rai nf a l l
Runof f
Evapor at i on
Loca t i on Mecca2-2 ( 51km2)
0. 0
0. 5
1. 0
1. 5
2. 0
2. 5
3. 0
3. 5
4. 0
0 6 12 18 24 30 36 42 48
Ti me ( Hour )
Ru
n-
of
f
Ra
te
,
mm
/h
r
0
5
10
15
20
25
Rai nf a l l
Runof f
Evapor at i on
Loca t i on Mecca2-51 ( 3, 942km2)
Fig. 2.3-4 Simulated values of rainfall (top), discharged water (bottom, blue), and evaporation (bottom, red)
on 24 January 2005 for a surface covered by shrubs and sand.
a) Near Taif b) Near Mecca
0 . 0
0 . 5
1 . 0
1 . 5
2 . 0
2 . 5
3 . 0
3 . 5
4 . 0
0 6 12 18 24 30 36 42 48
Ti me ( Hou r )
Ru
n-o
ff
Ra
te
, m
m/
hr
0
5
10
15
20
25
Rai n f al l
Runof f
Evapor at i on
Loc at i on Me cc a2-2 ( 51km2)
0. 0
0. 5
1. 0
1. 5
2. 0
2. 5
3. 0
3. 5
4. 0
0 6 12 18 24 30 36 42 48
Ti me ( Hour )
Ru
n-
of
f
Ra
te,
m
m/
hr
0
5
10
15
20
25
Rai nf al l
Runof f
Evapor at i on
Loca t i on Mecc a2-51 ( 3, 942km2)
Fig. 2.3-5 Simulated values of rainfall (top), discharged water (bottom, blue), and evaporation (bottom, red) on 24 January 2005 for a surface covered entirely by sand
a) Near Taif b) Near Mecca
0. 0
0. 5
1. 0
1. 5
2. 0
2. 5
3. 0
3. 5
4. 0
0 6 12 18 24 30 36 42 48
Ru
n-
of
f
Ra
te
,
mm
/h
r
0
5
10
15
20
25
Rai nf al l
Runof f
Evaporat i on
Locat i on Mecca2-2 ( 51km2)
0. 0
0. 5
1. 0
1. 5
2. 0
2. 5
3. 0
3. 5
4. 0
0 6 12 18 24 30 36 42 48
Run
-of
f
Rat
e,
mm
/hr
0
5
10
15
20
25
Rai nf al l
Runof f
Evapor at i on
Loc a t i on Mecc a2-51 ( 3, 942km2)
Fig. 2.3-6 Simulated values of rainfall (top), discharged water (bottom, blue), and evaporation (bottom, red) on 24 January 2005 for a surface covered entirely by shrubs.
The present simulation led to the following findings:
1) The rate of water discharge was small for the shrub-covered surface, attributable to the high water retentivity of the shrubs. 2) The rate of water discharge was higher in Taif than in Mecca because of the larger rainfall in Taif than in Mecca.
2.4 Ground water model
Conventional lumped models are based on the rainfall and river water amounts observed at limited locations and times. The present ground water model can simulate the temporal and spatial variations of ground water depth, storage in dams, and recharge by rainfall. The mesh size of the model is
approximately 1 km (Nawahda et al., 2004). The schematic of this simulation is shown in Fig. 2.4-1.
Fig.2.4-1 Overview of the ground water model
The ground water depth was simulated using topographical and rainfall conditions similar to those of the Asir region. The simulation was performed with the following assumptions: 1) a filtration layer existed to a depth of 50 m below the ground surface along the slope of the Asir Mountains; 2) rainfall of 24 mm/day occurred within the area 10 km from the mountain tops (Fig. 2.4-2a) on the first day of January, March, October and December; 3) The initial value of the ground water depth was 5m from the bottom of the filtration layer.
The preliminary simulation results show that the ground water depth in the rainy area near the top of the mountains increases with the rainfall of each month (Fig. 2.4-2b). However, it takes more than one year for the ground water to penetrate into the lower altitudes along the surface layer of the mountain slopes. a) Topography and rainfall condition b) Ground water depth
100km
1km50m
10km100km
1km1km50m
10km
4.6
4.8
5
5.2
5.4
5.6
5.8
6
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Distance from mountain top(km)
Gro
und
wat
er
dept
h(m
)
Initial condition
end of Jan.
end of Feb.
end of March
end of April
end of May
end of June
end of July
end of Aug.
end of Sept.
end of Oct.
end of Nov.
end of Dec.
Fig. 2.4-2 Simulation of the ground water depth with the assumption of 24 mm/day on the first day of January, March, October and December near the tops of the Asir Mountain
Conclusions
The conclusions of the present study can be summarized as follows:
1) Meteorological models such as MM5 and RAMS can accurately simulate meteorological conditions including temperature, wind velocity, humidity, and rainfall in time and space. 2) Hydrological models can accurately simulate water discharge due to rainfall in time and space by taking into consideration the effects of topography and ground surface conditions. 3) The volume of water available at a dam can be estimated over a one-year period or longer, using the integrated simulation system of meteorological, hydrological and ground water models.
Acknowledgment
This work was conducted as a Japanese national project titled Global Water Cycle Project, sponsored by the Ministry of Education, Culture, Sports, Science, and Technology (MEXT).
References
1) Behera, S. K., J. J. Luo, S. Masson, S. A. Rao, H. Sakuma, and T. Yamagata, 2005: A CGCM study on the interaction between IOD and ENSO, J. Climate (accepted).
2) Chakraborty, A., S. K. Behera, M. Mujumdar, R. Ohba, and T. Yamagata,
2006: Diagnosis of tropospheric moisture over Saudi Arabia and influences of IOD and ENSO, Mon. Wea. Rev. (in press).
3) Nawahda, A., T. Kojiri, and S. Ikebuchi, 2004: Distributed runoff model linking
surface with groundwater processes, Annuals of the Disaster Prevention Research Institute, Kyoto University, No.47C
4) Ter Maat, H. W., R. W. A. Hutjes, R. Ohba, H. Ueda, B. Bisselink, and T. Bauer,
2006: Meteorological impact assessment of possible large scale irrigation in Southwest Saudi Arabia, Global and Planetary Change
5) Ueda, H., et al. 2005: Meteorological simulation for the improvement of environment
by greening of desert, Second Oman-Japan Symposium: Preservation of Environmental and Water Resources amid Economic Development