Modeling Water Futures Using Food Security and
Environmental Sustainability Approaches
Shahbaz Khan1,2, Jianxin Mu2,3, M. A. Rahimi Jamnani2,4, Hafeez M2 and Zhanyi Gao3
1
Charles Sturt University, Locked Mail Bag 588, Wagga Wagga, NSW, 2650, 2CSIRO Land and Water
, 3China Institute of Water Resources and Hydropower Research (IWHR), 4AB-KHAK Tehran
Consulting Engineers, Iran E-Mail: skhan@csu.edu.au
Keywords: BHIWA; PODIUM; VenSim; environmental sustainability; water balance; food security
EXTENDED ABSTRACT
Global water scarcity is becoming a constraint for
human development. Therefore there is a need to
enhance water productivity of the irrigated and
rainfed agriculture to meet the increasing food
demand due to the population growth. This paper
describes modeling efforts to assess present and
future water needs for food, people and
environment. This exercise is being carried out by
the International Commission on Irrigation and
Drainage (ICID) in five countries, China, Egypt,
India, Mexico, and Pakistan, which together cover
51% of irrigated area and 43% population in the
world. This paper presents two modeling
frameworks including the food security approach
of the International Water Management Institute
(IWMI) through policy dialogue model
(PODIUM), an environmental approach i.e. the
Basin wide Holistic Integrated Water Assessment
(BHIWA) model and its system dynamic version
using VenSim environment.
The two modeling approaches i.e. the BHIWA and
PODIUM can be used to simulate and analyze the
impacts of land use and climate change impacts on
water resources and eventually optimize the water
allocations
among
agricultural,
industrial,
domestic, and environmental sectors within a basin
context.
Both approaches described in this paper have
shortcomings for exploring water futures of a
country. BHIWA is not a distributed hydrologic
model and cannot deal with each land use
geographically distributed throughout the basin.
All such parcels need to be conceptually lumped
into a single land use unit. BHIWA model does not
depict the spatial variations in rainfall, potential
evapo-transpiration, intensities of cropping or
irrigation as well as slow horizontal groundwater
movement, from under one area to another.
1963
The PODIUM model is based on accounting of
more or less fixed contemn of “utilizable” surface
and ground water resource. The changes in the
hydrology due to land use changes, and the
resultant changes in the utilizable water are
important when considering longer term scenarios.
This paper highlights the benefits and
disadvantages of using these approaches through
model applications in the Qiantang River basin in
China. The Qiantang River basin lies in the
southeast of China and comes under a subtropical
monsoon climate with four well-marked seasons.
The average annual precipitation is between 1200
mm and 2200 mm, with annual evaporation
ranging between 800 mm and 1000 mm.
On the basis of the model simulations presented in
this paper it is concluded that the PODIUM model
is strong at representing basin/country’s food
needs and associated use of “utilizable” surface
and ground water resources but it does not depict
the inter-relations between different hydrologic
components e.g. surface and ground water. The
BHIWA model developed by ICID aims to
quantify water use by three sectors (i.e. nature,
food and people), therefore allows better
understanding of impact of land use changes as
well as soil and water conservation policies and
programmes through the simulation of overall
hydrologic cycle. A system dynamics based
representation of BHIWA model allows users to
conceptualize, document, simulate, analyze, and
optimize water management. It can help to
understand the most sensitive parameters for
improving the water use efficiency at system scale.
1.
INTRODUCTION
The world’s demand for freshwater is one of the
most critical issues in the 21st century. The world’s
primary water supply will need to increase by 22%
to meet the population needs in 2025. At the 2nd
World Water Forum in 2000, it was decided to
formulate the World Water Vision 2025 on “Water
for Food and Rural Development” (WFFRD) to
assess water for Food, water for People, and water
for Nature. The WFFRD projected substantial
increase in the global water withdrawal, water
storage and irrigation expansion that by 2025.
However, the “Overview Vision” of the 2nd World
Water Forum considerably scaled down these
findings. The International Commission on
Irrigation and Drainage (ICID), therefore decided
to develop a common framework through an
initiative called Country Policy Support Program
(CPSP) in 2002 (ICID, 2002). The CPSP aims to
assess and integrate water needs for three sectors,
i.e. food, people, and nature, for the present and for
the year 2025, with a goal to evolve policy
interventions. The CPSP program is currently
being carried out by the ICID in five countries,
China, Egypt, India, Mexico, and Pakistan, which
together cover 51% of irrigated area and 43%
population in the world (ICID, 2002). As part of
the project, two modeling approaches i.e.
simulation
of
food
security
and
environmental/water cycle were developed and
tested. This paper discusses how these modeling
approaches were applied in China.
The Basin wide Holistic Integrated Water
Assessment
(BHIWA)
model
uses
an
environmental and water cycle simulation
approach. It was developed by ICID in 2004 (ICID,
2004 a and b; China Institute of Water Resources
and Hydropower Research 2004). The original
BHIWA model is a semi-lumped model with a
Microsoft Excel interface. It is able to account for
the whole land phase of the hydrologic cycle,
including the consideration of hydrologic changes
due to changes in the land use and agriculture use.
The model is capable of depicting surface and
groundwater balances separately and allowing
interaction between them as well as impacts of
storage and depletion through withdrawals.
Khan et al (2005) have provided an overview of
system approaches in water management. A
system dynamics version of the BHIWA model
was developed by the authors under the VenSim
environment which allows users to conceptualize,
simulate, analyze, and optimize models for the
complex systems.
1964
The PODIUM modeling approach to address food
security was developed by the International Water
Management Institute (International Water
Management Institute 2003). This model can
generate food security scenarios at catchment,
national and global levels (Seckler et al., 1998).
The model maps the complex relationships
between numerous factors that affect demand and
supply of the water and food. It enables users to set
goals, such as food production for an adequate
level of per capita consumption, and explore ways
of reaching these goals through expanding
irrigated area or rainfed area, increasing cropping
intensity, or importing more food. Likely scenarios
can also be developed in terms of population
growth, diets, and developments in agriculture and
water resources; which will help to determine the
necessary steps for ensuring food security and
sustainable water use.
The primary aim of this study is to provide a
capability analysis of the food security and the
watercycle/environmental approach to assess the
current (year 2000) and future (year 2025) water
use in Qiantang River Basin of China.
2.
MODELLING WATER FUTURES
2.1.
Environmental /Water Cycle Approach
The BHIWA model specifically address the future
water scenarios for food and rural development,
water for people as well as water for nature, in
order to achieve sustainable development and use
of the water resources. The model was designed to
be simple and flexible. The conceptual diagram of
the model is given in Figure 1. The model can be
calibrated for the present conditions and applied to
derive water fluxes for future scenarios at monthly
intervals. The basin can be divided into a number
of sub-basins to allow the segregation of areas
with similar hydrologic and water use attributes.
The BHIWA model was imported into a system
dynamics environment using the VenSim system
(Ventana Systems, Inc., 2004). This approach
allows users to conceptualize, simulate, analyze,
and optimize models for the complex systems.
Most importantly, it has the powerful functions for
the sensitivity testing compared to the above two
models and it gives the opportunity to users to
identify the most sensitive parameters affecting the
water cycle. By connecting words with arrows,
relationships among system variables are entered
and recorded as causal connections. This
information was used by the Equation Editor to
help users to form a complete simulation model.
The sensitivity analysis can be done based on the
change of constant values of different variables e.g.
irrigation efficiency, and proportion of quick
runoff from rainfall into sub-basins and deep
percolation rate of paddy rice, etc. The effects of
Figure 1
Schematic of Hydrologic Model
The optimization of the model is based on
combination of different variables effect to river
flow such as the irrigation efficiency, proportion of
quick runoff from rainfall into sub-basins,
proportion of return flow from surface irrigation to
river, and deep percolation rate of paddy rice. The
observed river flow and model forecasted river
flow was compared based on the optimization to
find the best coefficients for the above mentioned
variables.
2.2.
change of the values are checked on the ground
water recharge and river balance.
Food Security Approach
PODIUM calculates the area of the crops needed
to meet the food consumption needs. Different
scenarios can be developed in for the food
consumption and demand with an interval of five
years. The food demand scenarios are simulated in
the model on the basis of population and per capita
dietary consumption for the basin or country.
Similarly, the scenarios for food production are
simulated on the basis of rainfed and irrigated area
in the basin or the country. Policies can then be
developed to improve the water use efficiency of
the cultivated area, the possibility of expansion of
irrigated area or rainfed area, and increasing
cropping intensity or importing more food in order
to meet the future food security for the population
of the basin. Projections for future years are
determined in relation to base year by the expected
changes in the key variables over this period.
PODIUM does not represent complete water flows
in a hydrological system but it represents system
water demands (current or forecasted) and
1965
remaining water balance as environmental or the
return flow.
3.
EXAMPLE APPLICATION OF TWO
APPROACHES
3.1.
Description of the Case Study
The Qiantang River basin lies in the southeast of
China and comes under a subtropical monsoon
climate with four distinct seasons. The favorable
average hydrologic conditions endow this basin
with rich agricultural production and ensuing rapid
economic development. However, this basin is also
plagued by escalating flood disasters in wet
seasons and serious drought in dry seasons due to
uneven spatial and temporal distribution of
rainfall. The total area of the catchment is 35,500
Km2 and the total population of the Basin is 10.67
million. Location of the Qiangtang River Basin in
China is shown in Figure 2. The annual
precipitation varies between 1200 mm and 2200
mm, with annual evaporation ranging between 800
mm and 1000 mm. The total volume of water
resources of Qiantang River basin (upstream of
Hangzhou Gate) is 38.64 billion m3 (BCM) which
includes 7.71 BCM of unconfined groundwater
resources that accounts for 20% of the total
volume. The cultivatable land area is 0.424 million
hectares (Mha) which includes 0.360 Mha of
paddy field and 0.064 Mha of upland crops. The
per capita cultivated area in the river basin is 0.04
Mha and total horticultural area is 0.1309 Mha
which accounts for 3.7% of the total land area. In
addition, there is also a variety of cash crops such
as tea-tree oil, oranges, bayberries, grapes,
persimmons, young trees, and loquats.
Figure 2 Location of Qiantang River basin in
China
Environmental/Water Cycle Approach
In Qiantang river basin, the major source of water
for agricultural, domestic, and industrial use is the
surface water while groundwater has not been used
for irrigation so far, except in small quantities for
domestic and industrial use (Provincial Institute
for Water Conservancy and Hydropower
Reconnaissance and Design, 1998, Provincial
Irrigation
and
Drainage
Technological
Development Company 1999; Provincial Institute
of Water Conservancy and Hydropower
Reconnaissance and Survey 2003). Studies were
done at the sub basin level. The two sub-basins
studied are:
SB1 : Upstream of Fuchunjiang Reservoir (two
thirds of surface water storage and land area)
SB2 : Downstream of Fuchunjiang Reservoir
(one third of surface water storage and land area)
The soil moisture capacity varies according to each
land use type, and values consistent with the likely
root zone depths and field capacities were taken
into account. Paddy rice is the major crop in the
basin, making up for 85% of the total cultivated
area. The cash crops like fruit and rapeseed etc.,
are also very common. The cropping area of
rapeseed in 2000 was 733 Km2, amounting to 12%
of the total cropping area. The land use types in the
model are shown in Table 1.
The BHIWA model was calibrated and validated
by comparing the monthly outflow (surface runoff
plus base flow) and the total groundwater recharge
and withdrawal computed by the model and
estimations made by Qiantang Basin Management
Bureau as well as the percentage of rainfall (with
an annual rainfall of 1632 mm for the year 2000),
percolating into groundwater (the natural recharge
rate from the precipitation) with the generally
1966
M o n th ly a v e ra g e riv e r
3
flo w (m illio n m)
3.2.
adopted norms. In terms of monthly outflow to
sea, this model matched the present conditions
fairly accurately, where the difference between the
total outflows computed by the model and data
observed by local hydrological stations was only
around 0.5%. Figure 3 shows the computed and
observed average monthly values of river flow for
the year 2000. The total annual observed river flow
is 3.6 BCM which is little bit less than the annual
average computed river flow for 2000. The model
calibration is achieved by manually changing a
number of parameters such as runoff from rainfall,
irrigation efficiency, and deep percolation from
paddy. This is a tedious process and it does not
allow parameters optimization, sensitivity analysis
and dynamic economic integration.
7000
Computed
Observed
6000
5000
4000
3000
2000
1000
0
Jan. Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Figure 3 Comparison of Computed and
Observed Average River Flow
Table 1
Land Categories Used in the Model
Forest and miscellaneous trees
Permanent pastures
Land not available for cultivation, waste, & fallow
Land under reservoirs
Rain-fed soybean and wheat
Rain-fed fruit
Irrigated double cropping of rice
Irrigated early rice and autumn maize
Irrigated single cropping of rice and
rapeseed/vegetable
Irrigated sugarcane and barley
Irrigated cotton and wheat
Irrigated sweet potato and vegetable
Irrigated vegetable
Irrigated fruit
The original BHIWA is written as an EXCEL
spreadsheet which is not suitable for dynamic
analysis and optimization. The BHIWA model was
imported into the Vensim environment which is a
powerful systems visualization tool. It is capable
of developing the causal and use graphs, and
reviewing equations of the workbench variables in
its Structural Analysis Tools. It can perform
sensitivity analysis and compare the behavior of
variables.
Sensitivity Analysis with VenSim
Vensim allows mutli-variable sensitivity analysis
using the Monte Carlo Simulation Method by
simultaneous changing the desired parameters. The
important parameters, for determining water
savings in this basin, are deep percolation rate for
paddy rice and irrigation efficiency. The default
values of these parameters at the beginning of the
simulation were 90 mm and 45% respectively. In
this paper, two parameters i.e. the deep percolation
rate of paddy rice (50-150 mm) and surface
irrigation efficiency (20%-90%) were used in this
model to test the sensitivity of surface and ground
water balance with 50%, 75%, 95% and 100%
probabilities respectively for 2000 time series data.
3.3.
Food Security Approach
One of the main features of the PODIUM model is
that all main variables and assumptions are made
explicit and can be changed easily by the user.
This feature makes the model an excellent tool for
scenario testing. The PODIUM model includes the
following four modules: Crop consumption; Crop
production; Water Demand (Agriculture, Domestic,
Industrial, Environment); and Water Supply
(Surface, Groundwater).
Figures 4 shows the sensitivity analysis of the end
of system river balance with the change in surface
irrigation efficiency in a range of 20% to 90%
respectively. It can be that the end of system river
balance is very sensitive to the change of irrigation
efficiency particularly at the peak of main
irrigation seasons i.e. the months of May and
August.
Figure 5
Sensitivity testing result of
Indicator 1 (withdrawals/total
runoff) with the change of
percolation rate (50 – 150 mm)
The total grain demand depends on the amount of
population and per capita grain consumption,
which is affected by the income, urbanization level,
development level of agricultural production
market and grain consumption policies. It is
estimated that the total population in Qiantang
River Basin will increase from 10.67 million in
2000 to 11.4 million in 2025.
Figure 4
Figure 6 shows the crop production, consumption,
and surplus/deficit (in terms of economics) for
year 2000 and the simulated results for year 2025.
Sensitivity testing result of end
of system river flow with the
change of irrigation efficiency
(20%-90%)
Figure 5 shows the sensitivity testing of the
indicator 1 ratio (withdrawals/total runoff) with the
change of deep percolation rate of the paddy rice
for the whole basin. The sensitive periods for the
deep percolation rate parameter are July, August
and September.
1967
With per capita grain consumption at the national
level scaled down as the average value at the basin
level, the total grain demand in 2025 needs to be
increased to 4.37 million tons according to the
simulation results of PODIUM. In terms of
economics, the food production in 2025 for all the
crops will be 1.5 Billion US$ higher than the food
production in 2000 for the basin as shown in
Figure 6.
of irrigation technology and adoption of
comprehensive agricultural measures. Therefore,
even though the cultivated area of rice and wheat
is reduced, however the grain production in year
2025 will still reach around 6 Mt, increased by
1.92 Mt than that in 2000. Figure 7 shows the
lumped water balance of the whole basin for the
year 2000. In 2000, agricultural water diversions
are 4.50 BCM and return flow is 28.89 BCM.
Where as agricultural water use in 2025 will be
reduced to 3.34 BCM due to high water demand
for industrial and domestic sectors. While, the
return flow in 2025 will be 28.61 BCM which
shows a very little reduction as compared to return
flow of 2000. Water balance for year 2000 suggest
there are is potential for the development of
groundwater in the Qiantang basin. The total
natural recharge computed by the model for the
basin from rainfall is 5.143 BCM, which is about
8.9 percent of average annual rainfall of 57.958
BCM. With the absence of groundwater use for
irrigation in this basin so far, exploiting
groundwater for both agricultural and D & I uses
inevitably has been a priority for local Integrated
Water Resources Development and Management
(IWRDM) in order to achieve the sustainable
development and use of water resources.
Production, Consumption and Suplus or Deficit
Grain Crops
Other crops
All crops
(BUS$)
8
7
6
5
4
3
2
1
0
-1
2000
2025
Requirement
2000
2025
2000
Production
2025
Surplus/Deficit
Figure 6 Crop Production, Consumption and
Surplus or Deficit Simulated with
PODIUM
According to the simulation results, the cropped
area will be decreased from 0.771 Mha to 0.59
Mha in 2025 which can be explained to the
adjustment of cropping pattern and implementation
of the policy of returning cultivated land to forest
and pasture. Out of total area, the areas for the
paddy rice and wheat will be dramatically reduced
while the maize area will increase slightly. This is
due to assumed increase in the water productivity
of rice, wheat and maize due to the improvement
2000
WATER USE BALANCE
Evaporation &
Qiantang
Consumptive use
3.22
Natural utilizable surface water resources
26.18
32.17
0.65
0.77
0.00
Unutilized
natura flow
27.37
1.52
Unutilized
return flow
Outflow
28.89
Surface Storage
7.68
4.50
Total
Diversion
0.72
Return
Flow
0.55
0.17
0.75
Total
Total Return
Diversio
n
Flow
Diversion
0.30
Return
Flow
3.27
Agriculture
0.58
1.61
Domestic
0.04
0.26
0.30
0.26
Industrial
2.07
0.00
1.59
Total Return
Diversion
Flow
0.00
0.04
0.10
Total Return
Total
Diversion
Flow Diversion
0.04
Return
Flow
Environment
Natural groundwater recharge
0.88
Swamp 0.29
6.53
Flows
to sea
0.59
Swamps & Sea
Figure 7
4.
Lumped water balance of the whole basin with PODIUM (BCM)
COMPARATIVE ANALYSIS OF TWO
APPROACHES
It is very difficult to have absolute comparison of
modeling approaches presented in this paper. Each
model has its own limitations and advantage for
analysing sustainable water management of a basin
or a country. The PODIUM model is strong at
representing country’s food needs and associated
used of “utilizable” surface and ground water
resources, it does not depict the inter-relations
1968
between surface and ground water. The BHIWA
model developed by ICID aims to quantify water
use by three sectors (i.e. nature, food and people),
therefore allowing better understanding of impact
of land use changes as well as soil and water
conservation policies and programmes through the
simulation of overall hydrologic cycle. BHIWA
model in VenSim in system modeling environment
allows users to conceptualize, document, simulate,
analyze, and optimize dynamic models. It can help
to understand the most sensitive parameters for
improving the water use efficiency at system scale.
5.
CONCLUSIONS
Using a food security approach the PODIUM
model indicated total grain demand for the
Qiantang River Basin needs to be increased to
4.37 million tons by 2025. In terms of economics,
the food production in 2025 for all the crops will
be 1.5 Billion US$ higher than the food production
in 2000 for the Qiantang basin even though the
cropped area will be decreased from 0.771 Mha to
0.59 Mha in 2025 which can be explained to the
adjustment of cropping pattern and implementation
of the policy of returning cultivated land to forest
and pasture.
PODIUM is strong on food security but does not
address the water management problem and it is
best applicable for a country. BHIWA model is
useful for understanding the water cycle and
associated intervention but does not address the
food security. BHIWA is best applicable for a
basin. BHIWA divides area into sub units where as
PODIUM treats areas as weighted average. System
dynamic model can be helpful to understand the
most sensitive parameters for improving water use
efficiency and it can also be used to optimize
unknown parameter. The VENSIM model has
indicated that the end of system river balance very
sensitive to the changes in irrigation efficiency.
6.
POSSIBLE WAY FORWARD
The authors recommend that PODIUM and
BHIWA models may be integrated using a system
dynamics such as VENSIM . This should be linked
with a cost benefits analysis at the basin and
country level. Podium can not deal with the
externalities such as virtual water trading within
and outside the country to evaluate production and
environmental tradeoffs. The authors propose a
global food security, virtual water trades, and
environmental stress analysis using an integrated
PODIUM-BHIWA approach.
7.
8.
REFERENCES
China Institute of Water Resources and
Hydropower Research (2004), Report of
Country Policy Support Programme for
Chinese National Consultation, 2-3, August,
2004 (In English).
International Commission on Irrigation and
Drainage (2004a), Draft Report of Country
Policy Support Program (CPSP) for Chinese
National Consultation, Beijing, China
International Commission on Irrigation and
Drainage
(2004b),
Proceedings
of
International Water Management Institute
(IWMI)/ICID,
Scenario
Development
Orientation Workshop for India and China, 34 September 2004, Moscow, Russia.
International Water Management Institute (2003)
User’s Manual on PODIUMSim –Policy
Dialogue Model : Version II.
Khan S., Mu J., Hu Y., Rana T., and Gao Z.
(2005), System Approach to Achieve Real
Water Savings, Proceedings of the 19th
Congress on Irrigation and Drainage.
Provincial Institute of Water Conservancy and
Hydropower Reconnaissance and Survey,
(2003), The Report of Overall Water
Resources Planning in Qiantang Basin,
Zhejiang
Provincial Irrigation and Drainage Technological
Development Company, (1999), The Tenth
Five-Year Development Plan and Perspective
Planning for Year 2015 on Water Saving
Irrigation in Zhejiang Province, Zhejiang
Provincial Institute for Water Conservancy and
Hydropower Reconnaissance and Design
(1998), The Report of the Master Plan for
Qiantang River Basin, Zhejiang
Seckler D., Amarasinghe U, Molden D., Silva R.,
and Barker R, (1998), World Water Demand
and Supply, 1990 to 2025: Scenarios and
Issues. IWMI Research Report 19, Colombo,
Sri Lanka.
Ventana Systems, Inc. 2004. Vensim 5 User’s
Guide, Ventana Systems, Harvard, MA, USA.
ACKNOWLEDGEMENTS
The authors wish to acknowledge technical support
by IWMI and ICID for the use of PODIUM and
BHIWA models. Data collection and collation
efforts of colleagues at CSIRO, Australia and
IWHR, China are highly appreciated.
1969