CO7 - Application of Computer Models in Irrigation Management

APPLICATION OF
COMPUTER MODELS
IN IRRIGATION
MANAGEMENT
Monika Marković,
David Kranjac,
Milena Jančić-Tovjanin
Publisher
Josip Juraj Strossmayer University of Osijek
Faculty of Agrobiotechnical Sciences Osijek
For the publisher

Monika Marković, PhD, associate professor
Authors
Monika Marković, PhD, associate professor
David Kranjac, PhD, postdoctoral researcher
Milena Jančić-Tovjanin, PhD, postdoctoral researcher
Technical editor
Ras Lužaić, dipl. ing.
Reviewers
Jasna Šoštarić, PhD, full professor
Magdalena Zrakić Sušac, PhD, assistant professor
Proofreader
Monika Petrinec, mag. philol. angl.
December, 2021
© All rights reserved. No parts of this book may be reprinted or transferred

in any form without the approval of the authors.
ISBN 798-953-7871-99-4
Josip Juraj Strossmayer University of Osijek
The approval for publishing the university textbook was issued

by the Senate of the Josip Juraj Strossmayer University of Osijek
at the third session held on 20 December 2021 under the number 42/21.
TABLE OF CONTENTS
1. Introdcution 4
1.1. Computer models 6
1.2. Wireless sensors 6
1.3. GIS (Geographic Information System) 7
1.4. Remote sensing in irrigation management 8
2. CropWat MODEL 9
2.1. About CropWat model 10
2.2. Step by step procedure 11
2.2.1. Station and climate data 11
2.2.2. Effective rainfall 14
2.2.3. ETc calculation 16
2.2.4. General soil data 18
2.3. Analysis of Crop Water Requirements (CWR) 20
2.4. Irrigation scheduling 22
3. ETo Calculator 25
3.1 ETo calculator example 27
4. AquaCrop 34
4.1. AquaCrop model 35
4.2. AquaCrop application 37
4.2.1 Climate file 38
4.2.2 Crop file 39
4.2.3 Management file 43
4.2.4 Soil file 46
4.2.5 Simulation 47
5. Databases in water management 50

5.1. AQUASTAT 51
5.2. AQUAMAPS 51
5.3. CLIMWAT 53
6. Partial equilibrium models 55

6.1. Partial equilibrium models and their applications 56
6.1.1 AGMEMOD (PE) model 58
6.1.2 AGMEMOD (PE) model in the practical use of
simulating future soft wheat yield and production
volumes under the irrigation 70
7. References 77
4
APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT
1.
Introdcution
Several key factors such as popu- improve and expand intensive ir-
lation growth, climate change, ris- rigation facilities.
ing food prices, declining natural Irrigation is the application of
resources call for more efficient water to meet the crop water
management in crops as well demand at the proper time in
as in water management. Good the proper way (Ali, 2010). Crop
water management maximizes water demand in arid and humid
the water use efficiency and pre- regions are not the same, i.e. in
serves water resources. The im- arid regions irrigation water is the
plementation of ICT (information main source of water for crop pro-
and communication technology) duction while in humid areas the
in water management improves crops are irrigated only during the

efficiency by introducing different period of drought. Nevertheless,
tools, devices, software, remote irrigated agriculture is one of the
sensors, and applications into major consumers of freshwater
the management. It enables en- resources. Therefore, proper ir-
gineers, scientists, farmers, and rigation scheduling is required
policymakers to increase produc- for better and more efficient irri-
tivity with improved planning and gation systems. The purpose of
management of the production irrigation scheduling is to calcu-
system. Furthermore, the imple- late net irrigation, irrigation rate,
mentation of ICT in water man- crop water demands (CWD), and
agement has improved the data irrigation time. Different meth-
storage capacity, exchange of in- ods could be used for the deter-
formation, and ability to access mination of irrigation scheduling,
data remotely. ICT can help to i.e. calculation method, observa-
prevent and reduce water losses tion method, estimation method,
through well-planned irrigation and measuring of soil moisture
schedules which is crucial to in- content. Each of these methods
creased productivity. ICT is now has its advantages and limita-
being used in irrigation projects to tions, which indicate the fact that
improve water management in a irrigation scheduling is a demand-
variety of ways: planning, data col- ing and time-consuming process.
lecting, storage, and sharing. Fur- However, the introduction of
thermore, it is used for weather computer models, instrumenta-
forecasting, and to improve water tions, and techniques for com-
efficiency with the introduction of puterized irrigation has made it
hardware, software, models, and easier since it provides irrigation
remote sensors. Such innovative scheduling exactly according to
water management is helping to the water needs of the crops.
5
1.1. Computer models
Some of the computer mod- ter which computer model
els that are used in irrigation is used for irrigation sched-
scheduling as well as in crop uling, a basic understanding
production are APSIM, CROP- of soil – water - plant interac-
WAT, AQUACROP, and DSSAT. tions is obligatory for achiev-
Dynamic simulation mod- ing efficient crop production.
els of soil-plant systems are Also, it must be stated that
powerful tools which com- the model to be as efficient
bined with automated weath- as possible it is necessary
er station networks, soil, and to calibrate and validate the

plant measurement systems model for specific crops as
are increasing the efficiency well as for specific growing
of decision making. No mat- conditions.
1.2. Wireless sensors

One of the developing sectors end-user in a short period of
in irrigation management is the time, ending with efficient tem-
implementation of wireless sen- poral and spatial data network.
sor networks (WSN). Sensors Of course, no matter what type
can provide real-time weather, of sensor is used, experts in
soil, and crop data which con- soil, climate, or crop science are
siderably simplifies the deci- required so that the sensor’s
sion-making in irrigation sched- maximum efficiency would be
uling (Figure 1). In such a form, achieved.
irrigation scheduling becomes The introduction of sensors to
smart irrigation technology. irrigation management is one
Due to the latest technological of the latest technological inno-
advances, sensors have become vations for improving irrigation
more efficient, reliable, and less efficiency.
expensive. Sensors are set up in
a field, i.e. soil sensors, canopy
sensors, or compact weather
sensors to collect (usually peri-
odically) and send data to the
6
Figure 1. Wireless sensors in smart irrigation
1.3. GIS (Geographic Information System)

Some of the GIS functions tions of terrain, including ele-
that are related to crop pro- vation and property boundar-
duction can successfully be ies, which can help delineate
used to monitor and improve irrigation canals and drain-
water management in irrigat- age systems. Laser scanning
ed agriculture. Of course, the (LiDAR) is a technology for ob-
use of sensors for collecting taining a highly detailed digi-
temporal and spatial data tal terrain model that can be
that are mentioned in the used more broadly in water
previous section is obligatory. management. It works similar
Some of the GIS software that to radar but it uses the beam-
can be used in irrigation man- ing light pulses instead of the
agement are ESRI, ArcPAD, radio waves for identification
ArcGIS, etc. One of the GIS of areas with the slope and el-
features is Digital orthopho- evation, water flow and catch-
to quads (DOQs), digital maps ments maps as well as ero-
that provide spatial illustra- sion control.
7
1.4. Remote sensing in irrigation
management
A large scale of crop produc-
tion justifies the implementa-
tion of remote sensing, i.e. sat-
ellite data. Precision irrigation
mostly uses data that are part
of hydrology and of course me-
teorology and collected data
are used to improve water
management in irrigated agri-

culture. Satellite imagery has
wide-ranging applications in
agriculture, for example, early
identification of plant stress
caused by pests and diseases,
lack of water, or monitoring of
weeds.
8
9
CropWat MODEL
2.
2.1. About CropWat model
CropWat 8.0 is a computer CropWat model calculates Crop
model designed by experts water requirements (CWR) and
from the Land and water devel- irrigation water requirements
opment division of the Food and (IWR) from climate, soil, and
Agricultural Organization (FAO). crop data (Figure 2).
Figure 2. Climate, soil, and crop data used for calculation of irrigation
scheduling and crop water requirements
The CropWat 8.0 model pro- can be found: http://www.

vides calculation of irrigation fao.org/land-water/databas-
schedules for different irriga- es-and-software/CropWat/en/.
tion management conditions Procedures for calculation of
as well for different crops. For the CWR and IWR are mainly
example, if irrigation has a full based on methodologies pre-
or supplemental (regulated sented in FAO Irrigation and
deficit) character. Full irrigation Drainage Papers No. 24 “Crop
will provide the total amount water requirements”. It would
of water needed for crops to be the best if climate, soil, and
avoid water deficit. Opposite to crop data are collected in the
full irrigation, deficit irrigation field or at least in the local
aims to reduce water demands area. If not, CropWat 8.0 pro-
with optimization of econom- gram includes standard crop
ic output. CropWat model is and soil data that are integrat-
available on the following link, ed into the model.
where also installation guide
10
2.2. Step by step procedure
A step-by-step procedure for tion, and intensity. The amount
the CWR calculation and irri- of water that crop uses for the
gation scheduling is explained evapotranspiration process can
below. The chosen crop is maize be observed as “lost” water that
(Zea mays L.) grown and irrigat- should be compensated with ir-
ed in the Osijek region. Each rigation. The process of evapo-
step is briefly described as well transpiration depends on the
as the main definitions that are amount of energy available for
related to the model. water evaporation, while the

largest source of energy is solar
2.2.1. Station and climate radiation. Due to differences in
data the position of the Sun, the po-
tential amount of solar radia-
CropWat computer program al- tion that can reach the surface
lows calculation of the CWR and from which water evaporates
IWR for different agro-ecolog- depends on the geographical
ical zones with the assumption position and the season. On the
that at the very beginning of the other hand, the actual amount
analysis the exact geo-coordi- of solar radiation reaching the
nates and altitude of the select- surface from which water evap-
ed area are given. The accurate orates depends on the atmo-
geolocation and climate data sphere, i.e. on the presence of
are needed to calculate the crop clouds that absorb and reflect
water requirements that are en- the main parts of the solar ra-
vironment-dependent. diation. Also, when assessing
Crop yield is highly dependent the impact of solar radiation on
on weather conditions during evapotranspiration, it should be
the growing season, these are noted that all available energy
rainfall, air temperature, and is not only used to evaporate
humidity, wind speed, etc. The water but part of the solar en-
primary source of water, that ergy is spent on warming the
is available water for crop pro- atmosphere and soil. Solar ra-
duction for most of the world diation that is absorbed into the
is rainfall. In crop production, atmosphere increases the air
rainfall is observed through temperature which in turn af-
three aspects, quantity, distribu- fects the ETo process.
11
Solar radiation absorbed by the vapor compared to cold air. In
atmosphere and heat emitted arid areas, relative humidity is
by the Earth increases the air low and evapotranspiration is
temperature. The heat of the more intense, so crops in such
surrounding air transfers en- conditions have a greater need
ergy and thus affects the rate for water. On the other hand, in
of evapotranspiration. Thus, rainy, tropical conditions, high
during sunny, warm days, the relative humidity reduces the
loss of water from the soil sur- loss of water from the soil sur-
face and plants by the process face and plants by the process
of evapotranspiration is higher of evapotranspiration.
compared to the evaporation The process of removing vapor
of water from them during cold from the air largely depends
and cloudy weather. From this, on the wind and turbulence of
it can be concluded that the the air above the surface from
higher the air temperature, the which the water evaporates.
more water will evaporate into When water evaporates from
the atmosphere from the sur- plants and soil, the air above the
face of the soil and plants. The surface from which the water
consequence is a higher crop evaporates becomes saturated
for water requirement, which in with water vapor. If such air is
case of lack of rainfall must be not constantly replaced by dry
compensated by irrigation. air, the process of evapotranspi-
Relative air humidity is the ratio ration weakens. In other words,
of air vapor pressure to its satu- continuous airflow contributes
ration vapor pressure. It shows to increased evapotranspira-
how much vapor is in the air tion. Thus, the higher the wind
according to the maximum speed in hot and dry conditions,
amount of water vapor that the the more water will be released
air could contain at the same into the atmosphere from the
temperature. As the air tem- surface of plants and soil.
perature changes, so does the Two main weather-related vari-
relative humidity. If the amount ables that determine IWR are
of vapor in the air remains un- the amount and timing of rain-
changed and the air tempera- fall, and evapotranspiration.
ture rises, the relative humidity CropWat gives the opportunity
decreases as warm air may con- to choose, how different types
tain a larger amount of water of climatic data will be entered
12
into the model. The air tem- from the Croatian Meteorologi-
perature can be provided as Tmin cal and Hydrological Service.
and Tmax, or as an average air At the very beginning, it is neces-
temperature (Tavg). Air humidity sary to enter the basic location
can be given as air relative hu- data (Figure 3), such as Country
midity (%) or as a vapour pres- and station data (altitude, lon-
sure (kPa). Daily sunshine data gitude, and latitude), and after-
can be given as hours of sun- ward the climate data so that
shine, as a % of daylength, and the ETo could be calculated. The
as a fraction of daylength. Wind combination of two separate
speed data can be given as km/ processes whereby water is lost

day or m/sec. It would be best on the one hand from the soil
to collect data from a weather surface by evaporation, and on
station that is located on the the other hand from the crop by
field. If that is not the case, the transpiration is referred to as
weather data can be obtained evapotranspiration (ET).
Figure 3. Station and climate data
13
Simply said, ET is the amount part of the rainfall is not effective
of water that is lost and should for the plant. Part of the rainfall
be compensated with irrigation percolates below the plant root
water. In the CropWat model, zone (deep percolation) and
the ETo calculation is based other parts of the rainfall flow
upon the Penman-Monteith away over the soil surface (run-
equation (Allen et al., 1998) off). Deep percolation water and
which is the established meth- run-off water cannot be used by
od for determining the ET of the plants. So, part of the rainfall
major herbaceous crops with is not effective but the remaining
sufficient precision for man- part is stored in the root zone
agement purposes (Fereres and can be used by the plants

and Soriano, 2007). and that is so-called effective
As presented in figure 3., cli- rainfall. In general, the efficiency
mate data include monthly data of rainfall will decrease with in-
for minimum and maximum air creasing rainfall.
temperature (or average), air hu- For most rainfall events below
midity (%), sunshine hours, and 100 mm/month, the efficiency
wind speed (km/day). Mentioned will be approximately 80%. The
climate data are used for the factors which influence which
calculation of reference evapo- part is effective and which part
transpiration (ETo, mm/day). is not effective include the cli-
ETo is the evapotranspiration mate, the soil texture, the soil
rate from the reference sur- structure, and the depth of the
face (grass) that represents the root zone, and these factors
evaporative demand of the at- have been discussed in previ-
mosphere regardless of crop ous studies of Dastane (1978).
or agricultural practices. As can There are different methodolo-
be seen from figure 2, the only gies for the calculation of effec-
factors affecting ETo are climatic tive rainfall and a few of them
parameters. are integrated into the Crop-
Wat model (Figure 4), a fixed
2.2.2. Effective rainfall percentage (usually 80%), FAO/
Before the calculation of Crop AGWL formula, empirical for-
evapotranspiration under stan- mula, and USDA method. In our
dard conditions (ETc) the effec- example, the effective rainfall is
tive rainfall is to be calculated. calculated with the use of the
This is an important step since FAO/AGWL methodology.
14
Figure 4. Effective rainfall calculation methodologies
After the rainfall data are en- rainfall (Figure 5). This significant
tered, the model calculates ef- reduction of available water for
fective rainfall which in our ex- the plants is important for the ir-
ample is 59.9% lower than total rigation scheduling of CWR.
Figure 5. Effective rainfall calculation
15
2.2.3. ETc calculation ETc=Kc x ET
After the ETo and effective rain- The crop coefficient varies thru
fall calculation is done, the the growth period because
model moves to the calculation of changes in vegetation and
of ETc. The crop evapotranspira- ground cover. In the CropWat
tion under standard conditions, model, the trends in Kc during
denoted as ETc, is the evapo- the growing period are repre-
transpiration from disease-free, sented in the crop coefficient
well-fertilized crops, grown in curve. Single (time-averaged)
large fields, under optimum soil crop coefficients, Kc, and mean
water conditions, and achieving maximum plant heights for non-
full production under the given stressed, well-managed crops in

climatic conditions (Allen et al., sub-humid climates for use with
1998). Crop water requirements the FAO Penman-Monteith ETo
(ETc) over the growing season are are presented in FAO publica-
determined from ETo and esti- tion (FAO, 1998). Next, the fol-
mates of crop evaporation rates, lowing data must be entered to
expressed as crop coefficients calculate the ETc (Figure 6):
(Kc), based on well-established • planting date (the harvest
procedures (FAO, 1977), accord- date is calculated by the model
ing to the following equation: according to used crop data);
Figure 6. Crop data
16
• Kc values - Only three values during the initial stage (Kc
for Kc (Figure 7) are required ini), the mid-season stage (Kc
to describe and construct the mid), and at the end of the
crop coefficient curve: those late-season stage (Kc end).

Figure 7. Single crop coefficient (Kc) (Allen et al., 1998)
• Crop development stage • Critical depletion fraction - is

– refers to the duration the critical soil moisture level
(in days) of specific crop where the first drought stress
growth stage; initial, de- occurs affecting crop ETo and
velopment, mid-season, yield. It is a fraction of Total
and late season. Of course, Available Water (TAW) that
the duration of a specific usually ranges from 0.4 to
growth stage is climate and 0.6. Sensitive crops with lim-
crop-dependent. ited rooting systems under
• Rooting depth – refers to high evaporative conditions
the depth of the rooting sys- have lower values, while the
tem at the beginning and the higher values are for deep
end of the growing season. and densely rooting crops
Effective root depth is the and low evaporation rates.
upper soil layer from which • The yield response factor – is
the plants can take 80 % of the crop yield response to the
water. Rooting depth is soil water supply. It relates rela-
type and crop dependent. tive yield decrease to relative
Vegetables are shallow-root- evapotranspiration deficit.
ed (< 60 cm), while for exam- Critical depletion fraction and
ple, maize shas deeper root- Yield response data are avail-
ing depth > 120 cm. able on: http://www.fao.org/3/
x0490e/x0490e0e.htm.
17
2.2.4. General soil data after having been wetted with
water and after free drainage is
The next step is to enter soil negligible (Soil Science Glossary
data. It is first necessary to Terms Committee, 2008).
enter the data on Total avail- FWC as a soil moisture con-
able water (TAW). Simply said, tent comes after the drainage
TAW (Figure 7) is the difference has stopped and it represents
in soil water content between the upper limit of the available
Field water capacity (FWC) and water or the maximum amount
Wilting point (WP). FWC is de- of water that soil can hold. This
fined as the content of water, indicates, of course, that the
on a mass or volume basis, FWC is soil type dependent.
remaining in a soil 2 or 3 days
Figure 8. Soil water content
Total available water differs as the depth of water that has

from Readily available water been depleted in the root zone
(RAW) which is the SWC that (USDA, 1997). When the MAD is
could be easily taken by the reached, the lack of water in the
plants. RAW is the difference in soil should be compensated with
SWC from FC and Management irrigation water to prevent crop
allowable depletion (MAD), the stress. This means that the RAW
desired soil-water deficit at the is equal to MAD. As the SWC de-
time of irrigation, and It can be creases, it is more difficult for
expressed as the percentage of the plants to extract the water
available soil-water capacity or from the soil because water be-
18
comes more strongly bonded to the water movement into and
soil particles. Afterward, in the through the soil profile. It is a
case of drought and dry farm- measure of how fast the water
ing the Wilting point (WP) is enters the soil (mm/day, mm/
reached. At this point SWC is low hour) and it greatly affects irri-
and plants cannot extract water gation scheduling. In case of low
so at the end they are wilting. infiltration rate the ponding or
This process is mainly influenced waterlogging can occur while in
by soil texture (Rai et al., 2017). a case of high infiltration rate,
As presented in Figure 9. the an excessive amount of rainfall
maximum rain infiltration rate or irrigation water can cause
must be entered. Infiltration nitrogen losses and contamina-

refers to soil ability to enable tion of aquafers with pesticide.
Figure 9. Soil properties
Basic infiltration rates for vari- and could be used in the Crop-
ous soil types (Table 1) are pre- Wat model as reference data.
sented by Brouwer et al. (1985)
Table 1. Infiltration rates for various soil types

Soil type Infiltration rate (mm/hour)
Sand < 30
Sandy loam 20 - 30
Loam 10 - 20
Clay loam 5 - 10
Clay 1-5
19
Then the Maximum rooting 1989), and basically, it represents
depth (MRD) is to be entered the lack of soil water content at
in the model. MRD is one of the the beginning of irrigation.
factors that determine TAW. At the end of the soil data, the
Values of MAD can be found at: value of Initially available soil
http://www.fao.org/3/x0490e/ moisture is to be entered. This
x0490e0e.htm. value represents the SWC at the
The next step is to enter the initial beginning of irrigation (mm/m).
Soil moisture depletion (SMD) As it can be seen, the irrigation
which is expressed as % of TAW. practice is highly dependent on
SMD is a measure of soil moisture soil science. Therefore, the effi-
between field capacity and exist- ciency of irrigation scheduling
ing moisture content and multi- requires a thorough evaluation

plied by the root depth (Walker, of the soil properties.
2.3. Analysis of Crop Water Requirements

(CWR)
Crop water requirements sity (g cm-3), field water capacity
(CWR) refer to the amount of (FWC, %), and soil moisture (%))
water required to compensate and root depth (cm). Irrigation
for the evapotranspiration loss- time has the most significant
es from a cropped field during a impact on the yield and quality
specified period of time. Given of the crops because if too early
that it is important to know the or too late irrigation events can
difference between Net irriga- cause overirrigation, or it could
tion and Irrigation rate. Net ir- cause drought stress. That is
rigation is the amount of water why the growing competition
needed to meet the crop water for water, especially in the agri-
loss through evapotranspiration cultural sector demands the cor-
(ETo) during the growing season. rect time for the application of
This means that the net irriga- irrigation water. Furthermore, it
tion is mainly rainfall and ETo is important to maximize the ef-
dependent. Irrigation rate is the ficiency of crop water produc-
amount of water added in one tivity (CWP) and to find a strat-
irrigation event, and it depends egy for saving water and better
on soil properties (soil bulk den- use of irrigation water.
20
In the CropWat model decadal range from 1.15 to 0.44, while
CWR are presented according the ETc ranges from 0.97 at the
to growth stages, i.e. initial, de- end of the growing season (Sep-
velopment, middle, and late. As tember) to 6.21 mm/day during
it can be seen from Figure 10, the summer month July.
in this example the kc values

Figure 10. Crop water requirements (CWR)
Now the CWR is expressed as mm of effective rain during the

ETc, meaning that during the growing season (April-Septem-
tested growing season 679.1 ber) which means that the Net ir-
mm of water is required to fulfill rigation (irrigation requirements)
crop water demands. Yet, as pre- for maize crop is 610 mm.
viously calculated there was 68.9
21
2.4. Irrigation scheduling
A CropWat model is a useful tool (Figure 11), and the yield reduc-
that can be used for irrigation tion is set to be 0.0%. This means
scheduling, water savings, and that it is planned to achieve the
for achieving high crop yields. maximum yield potential. Here is
In general, irrigation scheduling the Net irrigation (mm) expressed
is based upon the monitoring as Irrigation rate (mm) so that
of SWC and the CWR. Of course, there is no lack of water (deficit)
it is also soil type, climate, and and also there are no water loss-
crop dependent. es. The difference between Net
In this example, the planting date irrigation water (469.2 mm) and
was set on April 21st, while the Gross Irrigation Requirement
harvest date is September 27th (GIR) is clearly visible (Figure 11).
Figure 11. Irrigation scheduling
22
GIR (670.2 mm) is net irrigation In this example, the Irrigation
water to be applied in reality, time is set on irrigation at crit-
considering water losses (deep ical depletion. This means that
percolation, leaching, runoff, the irrigation event will occur
water application). Here, NIR when the SWC is at a MAD level
is 70% (Field efficiency, Figure and that the irrigation water will
11) from GIR. The efficiency of refill soil water content to FWC.
the irrigation system can range It should be stated that the
from 0.55 (light soil) to 0.8 (drip model offers different criteria
and trickle irrigation system). for irrigation timing (Figure 12)
Since the model is set to calcu- that are used for non-rice crops:
late the irrigation scheduling • Irrigate at user-defined in-

without the yield reduction, the tervals,
Efficiency of irrigation schedul- • Irrigate below or above crit-
ing is 100%. ical depletion,
Actual water use by crop (678.1 • the irrigation event will
mm) is the ETc for maize crop occur when the SWC is
(April-September). Further- below or above MAD,
more, the model calculates the • Irrigate at fixed interval per
moist deficit at harvest time, stage,
which in this example is 74 • Irrigate at fixed depletion,
mm. Irrigation efficiency is set • Irrigate at given ETc reduc-
as 70 % which is a default value tion per stage,
for a well-managed sprinkler ir- • Irrigate at given yield reduc-
rigation system. tion.
23
Figure 12. Scheduling criteria for non-rice crops
In the irrigation scheduling for the entire growing season

graph (Figure 13), the readily are presented, and basically, it
available water, total available shows the crop irrigation sched-
water, and soil water depletion ule from Figure 11.
Figure 13. Irrigation scheduling graph
24
25
ETo Calculator
3.
ETo calculator is software de- resents the evaporative demand
signed by the experts of FAO of the atmosphere regardless of
(Land and Water Division), and its crop or agricultural practices. This
main task is to calculate reference means that ETo does not repre-
evapotranspiration. Reference sent the shortage of water for a
evapotranspiration (ETo) is the specific crop (maize, vegetables,
evapotranspiration rate from the or other crops) but for well-wa-
reference surface (grass) that rep- tered grass (Figure 14).
Figure 14. Reference grass ETo
Also, the soil is not considered as a reference method (Raes,

to have an impact on ETo since 2009). According to Allen et
water is abundantly available al. (1998), ETo expresses the
at the reference evapotranspi- evaporating power of the at-
ration surface. The only factor mosphere at a specific location
that has an impact on ETo is and time of the year and does
climate, i.e., air temperature, not consider the crop charac-
solar radiation, wind speed, teristics and soil factors. The
and humidity. ETo is calculated ETo calculator is available at:
from meteorological data with http://www.fao.org/land-wa-
the use of the Penman-Mon- ter/databases-and-software/
teith model which is selected eto-calculator/en/.
26
3.1 ETo calculator example
After the software is started an an existing data file or to create a
interface shows database man- new one (Figure 15). For this ex-
agement with the option to select ample, a new file will be created.
Figure 15. The interface of the ETo calculator
27
The given filename is Osijek Further, the interior coastal lo-
2018 since the ETo is assessed cation is chosen since Osijek is
for the Osijek area, the year in a continental area of Croatia,
2018. Following, the Station also a semi-humid or humid
name is Osijek, and the Country area with light to moderate
name is Croatia. Location data winds is marked. Monthly me-
are described with longitude teorological data for the period
and latitude data expressed in from January to December 2018
degrees and minutes, and al- are used for ETo assessment
titude (meter above sea level). (Figure 16).
Figure 16. Station description and location description and

meteorological data
28
The next step is to set the cli- ture could be expressed as
mate parameters (Figure 17). minimum and maximum or the
Of course, the oC is chosen as mean air temperature.
temperature unit. Air tempera-

Figure 17. Input data description
For this example, the mean air sured at 2 m height. Sunshine

temperature is chosen. Air hu- is expressed as hours of bright
midity could be expressed as sunshine (n). Besides that the
the mean of relative air humidi- sunshine and radiation could be
ty (%), minimum and maximum expressed as relative sunshine
air humidity (%), the mean of hours (n/N), solar radiation (MJ/
dew point temperature (°C), or m2 day), or net radiation (Rn).
the mean of actual vapor pres- It should be noted that if some
sure (kPa). For this example, the of the climate parameters are
mean relative air humidity is missing, the software offers cal-
chosen. As for wind speed, the culations for wind speed, solar
mean wind speed (m/s) is mea- radiation, and air humidity.
29
Figure 18 presents input climate sponds to the Osijek region. The
data limits. For air temperature, relative humidity ranges from 15
the lower limit is -15 °C, and the to 100 %, and wind speed from 0
upper limit is 45 °C which corre- to 8 m/s or 691.2 km/day.
Figure 18. Input climate data limits
Once given period can be cor- as daily, then in this step it is

rected by adding or deleting a possible to add or delete days
month or entire year (Figure 19). and to adjust the study period.
If the climate data are expressed
30
Figure 19. Time range determination
The next step is to enter climate (n) for each month. If the climate
data (Figure 20), meaning the data are expressed as daily data,
mean air temperatures (oC), the then in this step daily data should
mean of air humidity (%), wind be inserted for all previously
speed (m/s), and sunshine hours mention climate parameters.
31
Figure 20. The input of meteorological data
After climate data have been software calculates daily ETo.

entered the software calculates For better visual presentation,
ETo for each month. If daily cli- the ETo data can be presented
mate data are used, then the as a plot (Figure 21).
Figure 21. Reference evapotranspiration presented as a plot
32
Once the ETo is calculated, the climate file which could be used
software offers to save results with AquaCrop (Figure 22).
on PC as a report or to export a

Figure 22. Data saving
33
34
4.
AquaCrop
4.1. AquaCrop model
AquaCrop is a crop growth and main goal is to simulate the yield
productivity model created by response to water of herbaceous
the FAO’s Land and Water Divi- crops and is particularly suit-
sion experts. Crop productivity ed to address conditions where
refers to the yield response to water is a key limiting factor in
water describing the relationship crop production (FAO, 2016).
between crop yield and water The flowchart of the AquaCrop
stress because of insufficient model is presented in Figure 23,
supply of water by rainfall or irri- indicating the main components

gation during the growing period of the soil-plant-atmosphere
(Raes et al., 2018). The model’s continuum (FAO, 2016).
Figure 23. AquaCrop flowchart
35
The model is a useful tool for • its focus on water;
planning and making man- • the use of canopy cover in-
agement decisions in irrigated stead of leaf area index;
and rainfed crop production. • the use of water productivi-
It helps to develop an irriga- ty (WP) values normalized for
tion schedule for maximum atmospheric evaporative de-
water productivity in differ- mand and CO2 concentration
ent climate scenarios and to that confer the model an ex-
maximize water productivity tended extrapolation capacity
in deficit irrigation. Accord- to diverse locations, seasons,
ing to Raes et al. (2009), the and climate, including future
AquaCrop evolves from the Ky climate scenarios;
approach by separating (i) the • the relatively low number of

actual evapotranspiration (ET) parameters;
into soil evaporation (E) and • input data which requires only
crop transpiration (Tr) and (ii) explicit and mostly intuitive
the final yield (Y) into biomass parameters and variables;
(B) and harvest index (HI). Fur- • a well-developed user inter-
thermore, the separation of ET face;
into soil evaporation and crop • its considerable balance be-
transpiration avoids the con- tween accuracy, simplicity,
founding effect of the non-pro- and robustness;
ductive consumptive use of • its applicability to be used in
water (soil evaporation). This diverse agricultural systems
is important especially when that exist worldwide.
ground cover is incomplete
early in the season or as the FAO (2017) gives the following
result of sparse planting. The limitations of the AquaCrop
separation of yield into bio- model:
mass and harvest index allows • AquaCrop can simulate daily
the partitioning of the corre- biomass production and final
sponding functional relations crop yields for herbaceous
as a response to environmen- crops with single growth cy-
tal conditions. cles only.
Raes et al. (2018) give features • AquaCrop is designed to pre-
that distinguish AquaCrop from dict crop yields at the single
other crop models: field scale (point simulations).
36
The field is assumed to be uni- • Only vertical incoming (rainfall,
form without spatial differenc- irrigation, and capillary rise) and
es in crop development, tran- outgoing (evaporation, transpi-
spiration, soil characteristics, ration, and deep percolation)
or management. water fluxes are considered.
4.2. AquaCrop application

Figure 24 presents the Aqua- agement (irrigation, field), and
Crop model interface with all soil (soil profile, groundwater).

the necessary data entered into At the bottom part of the inter-
the model: climate, crop, man- face is the simulation menu.
Figure 24. AquaCrop interface
37
4.2.1 Climate file be used for the simulation of
crop productivity. Climate data
The AquaCrop model gives the
(sunshine hour, rainfall, mini-
option to import data from an
mum and maximum air tem-
existing file or to create a cli-
peratures, air humidity, wind
mate file if the climate data are
speed) were obtained courtesy
available for a specific location
of Croatian meteorological and
(Figure 25). The model gives the
hydrological service, while the
option to analyze daily, decade,
ETo was calculated by use of ETo
or monthly climate data. In the
calculator (see chapter 3). In the
AquaCrop model following cli-
present analysis, a new climate
mate data are used: solar ra-
file is created for the Osijek loca-
diation (MJ/day), rainfall (mm),

tion, and the year 2018 (January
minimum and maximum air
to December). Plot rainfall data
temperatures (°C), and refer-
for the year 2018 are presented
ence evapotranspiration (ETo,
in Figure 26. Furthermore, ETo
mm/day) and wind speed (km/
(Figure 27) and maximum and
day). For this example, the
minimum air temperature (Fig-
monthly climate data are used.
ure 28) is allowed to be modi-
Some of the locations (weath-
fied or exported as the mean of
er stations) are previously inte-
monthly or yearly values.
grated into the model and can
Figure 25. Climate data interface Figure 26. Rainfall data
38
Figure 27. ETo data Figure 28. Air temperature data
4.2.2 Crop file according to calendar days.

The threshold temperatures
The same as with the climate
for crop development are 8 °C
file, one can choose to create
as base temperature and 30 °C
a new crop file or to use one
as upper temperature. Here,
of the existing files that are
the calendar days file is con-
integrated into the model. For
nected with the temperature
the present analysis crop file
file (in this example from 1 Jan-
is selected for maize with the
uary to 31 December 2018). Di-
beginning of the growing cycle
rect sowing is chosen for maize
on 23 March 2018 (Figure 29).
crop, good initial canopy cover,
In the crop file, the limited (de-
and 75000 plants/ha. Further-
velopment, production, fertility
more, very fast canopy expan-
stress, calendar) and a full set
sion almost entirely covers the
of data with all crop parameters
canopy (96%). Green canopy
(development, ETo, production,
cover (CC) is used to express
water, temperature, salinity,
the foliage development, and
fertility stress, calendar) can be
it represents the fraction of the
chosen. This time, limited crop
soil surface covered by the can-
parameters are used. Crop can-
opy (Figure 30).
opy development is analyzed
39
Figure 29. Crop file Figure 30. Crop development
After the sowing, the CC is low Ks for leaf expansion (Ksexp) with
since 0% of the soil surface is CGC as target crop parameter, (ii)
covered by the crop. During the and Ks for early canopy decline
mid-season, the canopy cover (Kssen). Here the CDC is the target
reaches a maximum value (1) crop parameter and represents
when 100% of the soil surface is Canopy Decline Coefficient (CDC),
covered by the canopy, and it will previously described as the CC
depend on crop type and plant- declination. Ksexp is <1 and can-
ing density (Figure 31). Canopy opy expansion declines when the
Growth coefficient (CGC) describes water content in the root zone
the expansion of CC from germi- drops below the threshold for
nation to extensive growth; low leaf expansion. As for early can-
(germination), maximum (mid-sea- opy decline, Kssen is <1, and can-
son), and declining (late-season). opy decline is triggered in condi-
In this example, crop develop- tions where the soil water content
ment is studied in no water, fer- in the root zone drops below the
tility of salinity stress. Of course, threshold. Lack of water during
water stress can affect the expan- or after the flowering stage can
sion of the ‘leaves and eventually reduce the yield and induce the
result in canopy senescence at reduction of HI whereby the du-
an early stage. Water stress coef- ration and the magnitude of the
ficients (Ks) are used to describe drought stress will determine the
the effect of water stress on CC, (i) reduction.
40
31. Canopy development 32. Flowering and yield formation
If water stress occurs after tion might be adjusted down-

flowering or root enlargement wards if water stress induces
it could result in Harvest index stomatal closure. It is the cor-
(HI) reduction (Figure 33). HI responding reduction in as-
represents the relationship be- similate supply that will slow
tween the grain yield and abo- down the increase in HI. At a
veground biomass. FAO (2017) permanent wilting point the
claims that depending on (i) reduction becomes 100%.
when the water stress occurs At sowing the effective root-
during yield formation, (ii) the ing depth is at a minimum.
magnitude of the stress, and (iii) The maximum effective root
the sensitivity of the crop the depth is 2.30 m (for deep-root-
adjustment can be upwards or ed crops), while the maximum
downwards: rooting depth is on the 108 th
• As long as vegetative growth day after sowing (Figure 34).
is still possible, the daily rate The conditions affecting rooting
with which HI increases might depth are previously described
be adjusted upwards if water by (FAO, 2017):
stress affects leaf expansion. • When the root zone depletion
The enhancement of HI is the exceeds the threshold for sto-
result of a decrease of the matal closure the incremental
competition of leaf growth daily
with reproductive growth; • root deepening under optimal
• The daily rate with which HI conditions is adjusted by mul-
increases during yield forma- tiplying the expansion rate
41
• with the water stress coef- the root deepening will slow
ficient for stomatal closure down and can even stop if the
(Kssto). The water stress alters soil water content at the front
the shoot/root ratio since is at a permanent wilting point;
water stress affecting canopy • Root deepening is slow in soil
expansion occurs earlier than layers with low penetrability
water stress affecting stoma- (expressed as a percentage be-
tal closure (which affects the tween 100 and 0%). Below the
expansion of the root zone); restrictive layer, the root zone
• If the sub-soil at the front of expansion is normal again.
root zone expansion is too dry,
33. Harvest index 34. Root deepening
Biomass water productivity is and C4 crops. This is because the

shown in Figure 34, and basical- C3 plants are able to moderately
ly, it represents the amount of increase crop water productiv-
water that is transpired in rela- ity by restricting the transpira-
tion to a certain amount of bio- tion process. C4 plants (maize,
mass (kg/m3). Since soil evapora- sorghum, sugarcane) tend to be
tion should be considered then more productive in high light
ET water productivity is intro- and temperature environments
duced and it shows the kg of crop which is an important fact dry in
yield per unit of evapotranspired improving of water productivity
water (m3). As it can be seen (WP) in regions with arid climate
from Figure 35, different indica- or drought conditions. WP is nor-
tive ranges are presented for C3 malized for the climate by divid-
42
ing the transpiration and ETo, centrations is analyzed, then WP
so that it could be applicable to is multiplied with the correction
different locations and seasons. factor (fco2). The fCO2 is >1 for
While the normalization for CO2 CO 2 concentration higher than
refers to considering the WP for the reference year, and <1 for
a reference atmospheric CO 2 lower CO2 concentration. Finally,
concentration (369.41 ppm, the the crop characteristics file ends
atmospheric CO2 concentration with the crop calendar with the
for the year 2000, FAO 2018). If given date and length of specific
the year with different CO2 con- growth stages (Figure 36).

35. Water productivity 36. Crop calendar
4.2.3 Management file amount of water for each irriga-

tion event. In this example, the
The management file is divid-
maize crop is irrigated, so the
ed into categories, irrigation,
irrigation schedule is used from
and field (Figure 24). In the
data base. Sprinkler irrigation
AquaCrop model are choices
is chosen method with 80% of
of rainfed (irrigation is not con-
wetted soil surface.
sidered) or irrigated crop. As
Irrigation time can be deter-
for irrigation methods, the user
mined by:
chooses sprinkler irrigation,
• fixed interval, that is the num-
surface, and drip irrigation. Fur-
ber of days between two irriga-
thermore, the fraction of wetted
tions events. The interval will
surface, and specification of irri-
depend on the ETc, effective
gation water quality, timing, and
43
rainfall, soil water holding ca- weed management, mulches,
pacity, and allowable depletion. soil bunds, and field surfaces.
• allowable depletion (mm) Basically, parameters that have
• allowable depletion as % of an impact on soil water status,
readily available water (RAW). run-off, or evaporation. The dif-
It is assumed that the crop is ference between the CROPWAT
irrigated so that the soil water and AQUACROP models is visi-
content is back to field water ble here since in the first model
capacity, while the irrigation only climate, soil, and crop data
water quality is excellent (Figure are considered, while in the sec-
37). The irrigation file is com- ond model also the crop man-
pleted by specifying irrigation agement is taken into consid-
events, i.e. day, water depletion eration. The soil fertility classes
as % of RAW, and date. Under are non-limiting, near-optimal,
Field management practices moderate, about half, poor, and
users chose soil fertility level, very poor.
Figure 37. Irrigation schedule
44
Soil fertility stress corresponds sumed that the soil surface is
to CC, CGC reduction, canopy not covered with mulch (Figure
cover declination, and WP re- 38). As for field surface practic-
duction. Classes of soil cover by es, the user can choose situa-
mulches are: none, spars, about tions field surface practices do
half, significant and complete. If not affect the run-off where the
the mulch is taken into consider- estimation of surface runoff is
ation, then the user can choose based on soil profile character-
the organic or plastic mulch or istics. Thant the situation where
user-specified mulch. The type the field surface practice affects
of mulch determines the reduc- surface runoff, prevent surface
tion of evaporation losses, e.g. run-offs such as tied ridges and

50% for organic mulches, 100% closed-end furrows. Also, the
for synthetic mulches, and 75% situation with soil bund that
for user-specified mulch. For prevents surface run-off, and
the present analysis, it is as- store excess water.
Figure 38. Mulches file
45
4.2.4 Soil file
Similar to previous datasets, a roots, or otherwise provide an
new soil file can be created or unfavorable root environment.
one of the indicative values for Examples are bedrock, cement-
soil datasets (sand, sandy clay, ed layers, dense layers, and
sandy clay loam, sandy loam, frozen layers. ,The next step is
silt, silt clay loam, silt loam, silty to adjust soil capillary rise. Ac-
clay) can be used. The user can cording to Lu and Likos (2004),
choose up to five soil horizons the capillary rise is a well-known
where each soil horizon has its unsaturated soil phenomenon
own specified characteristic. that describes the movement
The characteristics are (Figure of pore water from lower eleva-

39) soil thickness, soil water tion to higher elevation driven
content, that is permanent wilt- by the hydraulic head gradient
ing point (PWP), field capacity acting across the curved pore
(FC), saturation (SAT), and total air/pore water interface. The
available water (TAW). User can author states three fundamen-
choose between the soil with- tal physical characteristics re-
out the restrictive soil layer and lated to capillary rise that are of
specify the soil depth on which primary practical concern: the
the restrictive soil layer is locat- maximum height of capillary
ed. Restrictive layer means hori- rise, the fluid storage capacity
zon or layer that restricts one or of capillary rise, and the rate
more physical, chemical, or ther- of capillary rise (Figure 40). The
mal properties that significant- soil file is completed by specify-
ly reduce movement of water ing groundwater. As in previous
and air through the soil, restrict datasets, a new groundwater
roots or provide an unfavorable file can be created or one of the
root environment. USDA (2016) indicative values for ground-
states that a “restrictive layer” water can be used. User can
is a nearly continuous layer choose the condition without or
that has one or more physical, with the present groundwater
chemical, or thermal proper- table. In late one, groundwa-
ties that significantly impede ter table can vary with the soil
the movement of water and air depth, or it can have constant
through the soil or that restrict depth and salinity.
46
Figure 39. Soil water content Figure 40. Soil capillary rise
characteristics
4.2.5 Simulation ously taken samples are taken,

and in that case, the simula-
At the beginning of a simulation starts on the day when the
tion, the initial conditions for samples were taken. If data on
soil water content (Figure 41) soil water content are not avail-
and salt balance (Figure 412) in able, they are estimated so that
the root zone needs to be spec- it is assumed that the initial
ified. The model calculates the water content in the soil was
amount of water in soil and salt close to the field water capac-
accumulation at the end of the ity during wintertime. That way
growing cycle based upon the the simulation will start for ex-
initial situation. The best way ample during January although
to establish the initial situation the sowing was in springtime.
is to measure the soil water For the Mediterranean region,
content and salt accumulation, the middle months of August
ideally one day after sowing. are taken as the initial period
In case the samples cannot when the water content in the
be taken on the day of sowing soil is close to the value of the
then the values from the previ- wilting point.
47
Figure 41. Initial soil water Figure 42. Initial salt balance
content
The rainfall data before sowing with increased salt content or

can be evaluated according to raising the level of groundwater
cumulative rainfall, 10-days or with increased salt content re-
as a few successive days (Figure sults in the accumulation of salt
43). The salt content of the soil in the surface layer of the soil.
is an important factor in crop Increased salt concentration
production, especially in an irri- has a negative effect on crop
gated field. Irrigation with water growth and yield.
Figure 43. Rainfall data Figure 44. Canopy cover (CC)
In the AquaCrop model, the the root zone. The drop in yield
soil salinity is simulated using due to the stress caused by the
soil salinity stress coefficient increased salt content comes
(Kssalt) and average electrical because of canopy cover (CC)
conductivity of the saturation reduction as well as stomata
soil-paste extract (ECe) from closure (Figure 44).
48
In such conditions, the CC is re- degrees days (GGD). Specifi-
duced because the growing ca- cation of the first and last day
pacity of the crop is decreased. could be used for the estima-
Furthermore, when there is a tion of planting dates in future
higher salt concentration in soil, conditions. Furthermore, simu-
soil water potential is lower due lation for future conditions, i.e.
to the osmotic potential and the climate change gives the oppor-
water is less available to the crop tunity to analyze the impact of
as well as stomatal closure which higher air temperature, rainfall
clearly affects transpiration. pattern, and CO2 concentration
Different criteria for evaluation (Figure 46) on crop productivi-
of air temperature before sow- ty. For example, the increase in

ing can be specified. That is, CO2 concentration will reduce
daily minimum air temperature crop transpiration, yet signifi-
for each day of a given number cantly less than it will increase
of consecutive days. Then, cu- the biomass water productivity
mulative and sum of growing (Figure 47).
Figure 46. CO2 concentration Figure 47. Biomass
49
50
5.
management
Databases in water
Available information is of great bases are now available, which
importance for improving water provide information and ser-
management as well as agricul- vices that facilitate the planning
tural productivity in general. The and management in agriculture
ability to access and share data generally. Databases are free,
has improved the use of ICT (In- easy to access, and up to date
formation and communication so that the farmers, scientists,
technology) in irrigation man- and professionals can easily get
agement by improved water the information they are looking
use. This refers not only to farm for. Some of the databases that
levels but to governments in are used in irrigation planning
planning and implementation and water management are

of irrigation strategies. Sever- Aquastat, Aquamap, Climwat,
al agricultural-irrigation data- and FAOSTAT.
5.1. AQUASTAT
Aquastat is a useful tool in bles, regional water resources,
water management and global and water uses. Furthermore, it
water information system pro- provides information about irri-
vided by FAO (Land and Water). gation water use, a global map
It provides data, information, of irrigation areas by source of
and fact sheets related to water water, water, and gender, wa-
resources, water uses, and ag- ter-related country-level institu-
ricultural water management. tional framework, wastewater,
The following information are etc.
available in Aquastat: country The Aquastat is available on:
database and profiles, data- http://www.fao.org/aquastat/
sets and fact sheets, maps, ta- en/.
5.2. AQUAMAPS
Aquamaps is developed by FAO spatial database on water and
experts and it is complimentary agriculture that are commonly
to Aquastat’s statistical data. used for GIS so that the water
It provides regional and global actions, decision-making, and
51
investments are more efficient. for irrigation, geo-referenced
Spatial datasets in Aquamaps database on dams (Africa,
are regional, global, and orga- Central Asia, Middle East,
nized in the following themes Southern eastern Asia), irri-
(Figure 48): gation areas, percentages of
• Irrigation and infrastructures - the irrigated area serviced by
Dataset shows area equipped groundwater.
Figure 48. Irrigation infrastructure dataset
• Climate dataset – Data- • Analyses human pressure on

set shows a map of aridity, land and water, major agricul-
monthly and yearly grids of tural system, land salinization
precipitation and ETo, and due to irrigation (Figure 49),
rain days/month. percentages of area equipped
• Hydrological basins by conti- for irrigation, etc.
nent Aquamaps is available on:
• Inland water bodies and major https://data.apps.fao.org/aqua-
rivers of the world maps/.
52
Figure 49. Proportional of land salinized due to irrigation
(https://data.apps.fao.org/aquamaps/)
5.3. CLIMWAT
CLIMWAT is database formulat- fall (mm/month) and month-
ed by the Water Development ly effective rainfall (mm). The
and Management Unit and the Reference evapotranspiration
Climate Change and Bioenergy (ETo) is calculated with the Pen-
Unit of FAO that provides the man-Monteith method and ex-
monthly mean values of the fol- pressed as mm/day.
lowing climatic parameters:the CLIMWAT can be combined with
mean of daily maximum tem- the CROPWAT model which is
perature (°C), the mean of daily previously presented in chap-
minimum temperature (°C), the ter 2. The CLIMWAT data can be
mean of relative humidity (%), extracted for single or multiple
the mean of wind speed (km/ stations in the format suitable
day), the mean of sunshine (h/ for their use with the CROPWAT
day), the mean of solar radia- model. For this example, the
tion (MJ/m2/day), monthly rain- Osijek station is chosen (Figure
53
50). When the specific climate stations within a selected coun-
station is chosen, the interface try or within and around the se-
presents the location’s longi- lected country. Afterward, data
tude and latitude. After the are exported as a file that can
country is selected, one can be used as a climate file in the
choose to display all weather CROPWAT model.
Figure 50. CLIMWAT interface
54
55
Partial equilibrium
6.
models
6.1. Partial equilibrium models and their
applications
Economic models provide a sim- interactions of different agri-
plified view and analyze com- cultural products. They also in-
plex phenomena of economic corporate exogenous variables
reality and are also considered such as technological change,
useful tools for linking mathe- world population, and house-
matical theory with statistical hold income into their sup-
data. For the analysis of eco- ply-demand relationships.
nomic phenomena in agricultur- International trade in agricul-
al markets and policy, various ture in these types of models

methods of economic modeling generally focuses on trade in
have been developed, and one primary goods. That is, they
of the most commonly used is cover agricultural supply, de-
modelling using partial equilib- mand, and trade for unpro-
rium models. cessed agricultural products or
Partial equilibrium models usu- first-degree agricultural prod-
ally simulate international mar- ucts processing without taking
kets for a particular group of into account trade in processed
market goods, e.g., agricultural food products.
goods. Usually, partial equilib- The standard agri-food market
rium models see the agricul- partial equilibrium model must
tural sector as a closed system have the following characteristics:
with no links to the rest of the • global coverage
economy. Nevertheless, the in- • parametric differences be-
fluences of the rest of the do- tween countries represented
mestic and world economy on in the model
the agricultural sector can be • dynamic feature
included in a top-down manner • policy representation
by changing parameters and ex- • theoretical consistency
ogenous variables. • exogenous market factors
Partial equilibrium models can that are non-agricultural
be the single or multi-product, Quantitative models are import-
multi-product models can cap- ant sophisticated tools for ana-
ture the supply and demand lyzing the impact of agricultural
56
policies and other exogenous tails on production and policy
changes on key agricultural mar- instruments and describe one
kets. One of the modelling ap- sector or a group of closely re-
proaches used to analyze such lated products (Salvatici et al.,
impacts is AGMEMOD (AGricul- 2001). Generally, PE models are
tural MEmber states MODel- used for quantitative impact as-
ling), an econometric, dynamic, sessments of agricultural policy
partial equilibrium, multi-coun- reforms, but they can be used in
try, multi-market model. AGME- various simulations of the com-
MOD model provides extensive modity markets within the ag-
details of the agricultural sector ricultural sector such as grains
in the individual EU Member and oilseeds markets. Such sim-

States over common country ulations of various agricultural
model templates and the EU as commodity markets are called
a whole through the combined “outlooks” and they present de-
EU 27 model. Typical AGME- tailed medium-term (10 years)
MOD applications refer to the: projections of key agrarian-pol-
• Medium-term market projec- icy indicators (production, yield,
tions imports, exports, etc.) within in-
• Set-aside/ ecological focus dividual agri-food markets.
areas and payments Various PE models are devel-
• Modelling of Production quo- oped so far, but today there are
tas whenever applied for the three core partial equilibrium
countries models in the Integrated Model-
• Resource demand for biofuels ling Platform for Agro-economic
• Fertilizer policies for some Commodity and Policy Analysis
countries (iMAP) hosted by the EU Joint
• Rainfall data for some coun- Research Center (JRC) (M’barek
tries et al., 2012; M’barek and
• Landing Obligation Delincé, 2015). One of the core
Simulations of the impact of ir- models in the iMAP platform is
rigation on the growth of crops PE model AGMEMOD - (Agricul-
production volume due to the ture Member State Modelling)
increase in productivity can ef- which application in simulation
fectively be projected by Par- of the impact of irrigation on
tial Equilibrium (PE) models. PE agricultural production will be
models incorporate more de- presented (Figure 51).
57
Figure 51. Role of AGMEMOD in the framework of the iMAP modelling

platform at the JRC (Salamon et al., 2017)
6.1.1 AGMEMOD (PE) model soft and durum wheat, barley,

and corn; for oilseeds: rapeseed,
The AGMEMOD model is an
soybean, and sunflower; for live-
econometric, dynamic, multi-
stock and meat: cattle, beef, pigs,
product, multi-country partial
pork, poultry, sheep and mutton;
equilibrium (PE) model. The
main purpose of the model is to and for milk and dairy products:
produce medium-term projec- cheese, butter, whole milk pow-
tions or market outlooks for key der and skimmed milk powder),
agricultural products by 2030. where each commodity market is
The modelling strategy uses the based on annual time-series data.
bottom-up approach, based on In the case of Croatia, the data
country-level models, using a range from 1995 to 2018 and are
common country model tem- compiled mainly from national
plate, which is combined into a sources (Croatian Bureau of Sta-
composite EU model (Chantreuil tistics and Ministry of Agriculture).
et al., 2012). The Member State The database covers data on pro-
model is composed of commod- duction, food and feed consump-
ity market sub-models (for grains: tion, imports, and exports in the
58
form of market balance sheets. from industry sources, but the da-
As mentioned, data is collected tabase in the model itself has to
from national statistical institutes, be organized in a specific manner
EUROSTAT, research institutes, or (Figure 52).

Figure 52. AGMEMOD data sources
A plausible and consistent da- currency exchange rates);

tabase is necessary in order to • availability of the projections
modelling process to perform for the macroeconomic vari-
well. The criteria for assessing ables that drive the models
the admissibility of data include: (e.g., GDP, inflation, popula-
• reliability and accessibility tion growth);
of the data series and their • relevance of the data to the
up-dating; users of the results.
• additivity of variables: the coun- The model uses two specific types
try level numbers for many vari- of variables which can generally
ables have to add-up to accept- be divided into endogenous and
able totals for the EU as a whole exogenous and each variable has
(e.g., the national commodity a unique mnemonic and can only
balance sheets must add-up to be allocated to one parameter
that for the EU); type. The general logic behind
• data consistency across all mod- variable type and how mnemonic
els for the variables that drive will look inside the model is pre-
the individual models, (e.g., the sented in Table 2.
59
Table 2. Parameter types used in AGMEMOD
Parameter type Meaning Variable type
Product P and Activity A in Country C endogenous
V2(P_A,C,T1)
in year T1; endogenously calculated in model
Product P and Activity A in Country C
exogenous
V(P_A,C,T1) in year T1; fixed on the last observed
in model
year
World market price for Product P and exogenous
VWP(P_A,T1)
Activity A in year T1 in model
EU dependent Policy variable for Prod- exogenous
VPOL(P_A,T1)
uct P and Activity A in year T1 in model
Policy variable for Product P and Activ- exogenous

VPOLC(P_A,C,T1)
ity A in Country C in year T1 in model
Policy Harmonization variable for
exogenous
Vph(P,A,C,T1) Product P and Activity A in Country C
in model
in year T1
Single Farm Payment policy variable
exogenous
VPOL_SFP(P_A,T1) for Product P and Activity A in year T1;
in model
exogenous in model
Macroeconomic variable for Activity A exogenous
VMAC(A,C,T1)
in Country C in year T1 in model
exogenous
DUM(D,T1) Dummy D in year T1
in model
exogenous
TREND (T1) Trend T in year T1
in model
Commodity markets are mod- demonstrate changes in the be-

elled as interrelated in such a havior of economic stakeholders
way that they reflect the com- (producers and users), changes
petition between different prod- in exogenous data (macroeco-
ucts for resources, various in- nomic variables, technical prog-
teractions between crop and ress, policy instruments), and
livestock markets, etc. Supply prices. Using sets of economet-
and demand, international trade, rically estimated equations, the
and prices are endogenously de- model generates projections of
termined in the commodity mar- endogenous variables from ex-
ket sub-models (Chantreuil et al., ogenous and endogenous data.
2012). Country-specific models In order for the model to satisfy
60
the partial equilibrium condition, al market a market equilibrium,
it is necessary to establish for which implies the following equal-
each individual key agricultur- ity at a certain product price:
Productiont+Beginning stockst+Importst= Domestic uset+Ending stockst+Exportst
In the commodity market and prices are endogenously

sub-models (Figure 53), in this determined inside the model
case for soft wheat, supply and (Chantreuil et al., 2012).
demand, international trade,
Figure 53. Croatian AGMEMOD partial equilibrium model scheme
Country-level models demon- technical progress, policy in-

strate changes in the behavior struments), and prices. From
of producers and consumers, exogenous and endogenous
changes in exogenous data data using sets of econometri-
(macroeconomic variables, cally estimated equations, the
61
model generates simulations output variables is the regres-
of endogenous variables. sion equation. Therefore, the
The general form of the econo- general equation of the model
metric equation according to can be written as:
which the model derives the
Y=α+εβ1X1+εβ2 X2+εβ3 X3…+є (1)
Equation 1. presents standard the General Algebraic Modeling

regression equation, where Y System (GAMS) computer pro-
is the dependent indicator α gram (McCarl, 2010). Before the
is the intercept, ε the elasticity realization of the AGMEMOD

coefficient, β the regression co- model in GAMS (Figure 54), all
efficients, Х123 are independent estimated and calibrated equa-
factors affecting on change of tions, as well as the data enter-
the dependent indicator and є ing the model, must be speci-
is presenting an error term. fied in the MS-Excel program in
AGMEMOD model software ap- separate sheets.
plication was realized through
Figure 54. AGMEMOD-GAMS software
62
The model has its own interface program. These excel sheets
that contains several MS Excel are located above in the drop-
documents that are intercon- down menus under “InputData”
nected or loaded via the GAMS and “ModelSources” (Figure 55).

Figure 55. AGMEMOD interface
“InputData” drop-down menu Currently AGMEMOD model

contains MS Excel documents has data and the corresponding
that cover commodity markets equations for all EU 27 Member
historical data, assumptions States. There are also partners
input data, and policy harmo- from other European countries
nization data. While, “Model- (Serbia, Bosnia and Herzegovina,
Sources” drop-down menu con- Russia, Ukraine, and Kazakhstan)
tain an MS Excel document with and some other countries out-
model equations for each coun- side Europe: Brazil, Ghana, Ethi-
try, commodity, and variable opia, Kenya, and Rwanda. The
within the model itself. One Croatian country-level model is a
more key feature of the model result of the authors’ standalone
is under the “ModelSources” model according to the AGME-
drop-down menu which re- MOD Partnership rules. The na-
lates to the R program in which tional commodity market mod-
econometric behavioral equa- els of the key Croatian crop and
tions are estimated. livestock markets, are estimated,
63
calibrated, and then validated Countries data ranges from the
based on experts’ knowledge year 2000.
and literature afterward. Historical data MS-Excel sheet
Each country model is based is found in “InputData” drop-
on a database template of down menu within AGMEMOD
annual time series, which de- interface and holds annual data
pending on the country rang- on initial stocks, production, im-
es from as early as 1973 to the ports, human consumption and
latest available. Usually for the animal feed, industrial consump-
oldest EU Member states data tion, exports, ending stocks, and
ranges from 1973 and for New domestic prices for all key agri-
Member States from the 1990s cultural markets in each country
in the case of Western Balkan model template (Figure 56).
Figure 56. AGMEMOD historical data input sheet
Where possible, the AGME- compilation has become more

MOD Partnership uses Eurostat complicated, time-consuming.
sources such as the Agricultural National teams responsible for
Information System and New- country models and data com-
Cronos to populate the AG- pilation are required to turn to
MEMOD database (Levert and diverse national data sources.
Chantreuil, 2006). The second MS-Excel sheet
As Eurostat no longer provides under the “InputData” drop-
supply and use balances, data down menu holds a set of data
64
related to the exogenous as- real Gross Domestic Product,
sumptions that lie beyond the GDP deflator, real GDP per cap-
combined model (Figure 57). ita, average annual exchange
This set of data contains annual rate HRK / EUR, EU policy instru-
data about world market prices, ments, dummies, and trends.
population estimates mid-year,

Figure 57. AGMEMOD assumptions input datasheet
For assumptions on the de- ers the EU in its current compo-

velopment of macroeconomic sition (EU-27), following Brex-
conditions, the AGMEMOD uses it, which took place in 2020. It
the same sources as the EU Ag- presents the medium-term
ricultural Outlook, which itself outlook for the EU agricultural
relies on a combination of the markets, income, and environ-
European Commission’s short- ment up to 2030, based on a set
term outlook, forecasts from of coherent macroeconomic as-
IHS Markit, and other sourc- sumptions deemed most plausi-
es including the International ble at the time of the analysis.
Monetary Fund, the World Bank Country-specific policy instru-
and the OECD. Their respective ments are contained in the pol-
major assumptions include data icy harmonization sheet also to
on economic growth, inflation, be found under the “InputData”
and exchange rates (EC, 2016). drop-down menu in the AGME-
The EU Agricultural Outlook cov- MOD interface (Figure 58).
65
Figure 58. Policy harmonization datasheet
Policy harmonized approach as- price addon to producer prices

sumes the inclusion of current for each observed agricultur-
2015-2020 CAP measures and al products market, in order to
instruments, which relate to form a reaction price that af-
regional and voluntary coupled fects the level of production,
supports (Salputra et al., 2011). sown area, average slaughter
Regional, coupled payments as weight, and numerous other
well as state aids are recalcu- variables in the Croatian nation-
lated and included as a policy al AGMEMOD model (Figure 59).
Figure 59. Policy harmonized approach in the AGMEMOD model
66
Finally, the last sheet found torical datasheet. Estimated
in the “ModelSources” drop- variables come from the R ap-
down menu is containing sets plication program which is also
of econometrically estimated an integral part of the AGME-
equations. Some variables in MOD model and also it is found
the model are estimated and under the “ModelSources”
some are derived from a his- drop-down menu.

Figure 60. AGMEMOD equation sheet
Equation type (EQ) means that numerical variables. In some

the variable has been estimat- cases, in addition to numerical
ed or calibrated, identity type variables, qualitative variables
(IDEN) indicates that the vari- are included in the model as
able is defined as a derived regressors in order to describe
variable of other variables. the impact of qualitative infor-
Fixed variable type (FX) - indi- mation, and thus expand the
cates that the variable was de- scope of regression analysis.
termined by its last observa- Qualitative variables are includ-
tion in the historical data and ed in the model using dummy
is therefore treated as an ex- variables. A regression model in
ogenous variable in the model. which the dependent variable is
The variables included in the numerical, and which contains
regression model are usually one or more dummy variables
67
as independent variables in ad- user has to go to “ModelSourc-
dition to the numerical ones, es” drop-down menu of model
is analyzed in the same way as interface and find “AGMEMOD
the standard regression model. Gams Model and Analyzing Tool
Model is generating results in for” then click on “TI GDX2Excel
MS-Excel tables, and to gen- Tool” then the model will gener-
erate results in such manner ate tables shown in Figure 61.
Figure 61. AGMEMOD Excel results table
Results are generated in sep- tain simulated data results on

arate documents which relate harvested areas, animal slaugh-
to crop markets results (soft ters, slaughter weights, yields,
and durum wheat, corn, barley, production, domestic use (food,
rapeseed, sunflower, and soy- feed, and processing use), im-
beans), livestock market results ports, exports, stock change,
(cattle, pigs, poultry, sheep, and market prices, and self-suffi-
goats with associated markets ciency rates by 2030.
for meat, dairy cows, milk and In addition to the tabular dis-
dairy products). play, the model is capable of
The results are presented in the generating graphical results for
form of production-consump- the same listed simulated re-
tion balance sheets and con- sults (Figure 62).
68
Figure 62. AGMEMOD Excel results graphs
Also, it is possible to generate GAMS user has to go to the “Re-

results in GAMS for each indi- sults” drop-down menu and find
vidual variable for each country “OutPut Tables for all Scenarios
in the AGMEMOD model (Fig- (GDX/Gamside)”.
ure 63). To produce results in
Figure 63. AGMEMOD GAMS results
69
6.1.2 AGMEMOD (PE) statistical office and cover data
model in the practical on production, consumption, im-
use of simulating ports, exports, beginning stocks,
and ending stocks in the form
future soft wheat
of balance sheets (CBS, 2021).
yield and production
The Croatian soft wheat market
volumes under the modelled using an appropriate
irrigation econometric methodology as
This section will show the actual described by the general rules
modelling results of the Croa- of the AGMEMOD modelling ap-
tian soft wheat market (harvest- proach (Hanrahan, 2001; Erjavec
ed area, yield, and production) and Donnellan, 2005). The sim-

via the AGMEMOD model by ulation assumes stable market
2030. Two simulations will be conditions without market dis-
made, one without the effect tortions with stable climatic and
of irrigation and the other with weather conditions (ceteris pari-
the effect of irrigation on all bus) and with the continuation
observed soft wheat is sown. of existing agricultural policy
Namely, no country or region measures up to 2030 (baseline)
irrigates all sown areas, but the (Jansik et al., 2006).
intention is to test the applica- Next, the general forms of
tion of the model and its capa- econometric equations in the
bility to simulate the impact of model used for the simulation
irrigation on soft wheat yields of the soft wheat market (har-
and production. When creating vested area, yield, and pro-
a real simulation, it is possible duction) in Croatia will be pre-
to place part of the areas that sented. Future simulated total
are truly irrigated or planned to wheat production is determined
be irrigated in the future in the by simulation of future soft
model itself. wheat sown areas and yields.
Croatian soft wheat market his- Therefore, the equation for the
torical balance sheet data ranges total harvested area of cereals,
from 1995 up to 2018, and they oilseeds, and root crops can be
are compiled from the national written as:
ahi,t = f(p ji,t-1, ahl,t-1, V) j = 1,...,n; i,l = 1,...,3; i ≠ l (3)
70
Where: vested area of culture (e.g.
ahi,t = harvested area in year t of certain policy instruments
crop group i such as coupled payments)
p ji,t-1 = the actual price of culture
j belonging to the group of The share of culture k that be-
crops i in year t-1 longs to the group of cultures
V = vector of exogenous vari- and is determined by the follow-
ables that may affect the har- ing equation:
sh ki,t = f(p ji,t-1 , sh ki,t-1) j,k =1,...,n (4)
The share of the harvested area crop = Total harvested area ×

of an individual crop is calcu- share of nth crop.
lated according to principle 1 - The yield of culture k which is
part of the area of all other (n-1) in the group of cultures i can be
crops, which means that the expressed as:
harvested area of the observed
r ki,t = f(p ji,t-1 , r ki,t-1 , V) j,k =1,...,n (5)
Where: econometric behavioral equa-

r ki,t = yield per hectare of crop k tions is transferred to the AG-
which is in the group of crops i MEMOD program via mnemon-
p ji,t-1 = the actual price of culture ics to the R program through
j belonging to the group of which equations are economet-
crops i in year t-1 rically estimated using national
r i,t-1 = yield per hectare of crop k
k
annual time series (Figure 64).
which is in the group of crops Using sets of econometrically
i in the year t-1 estimated equations, the model
V = a vector of exogenous vari- creates projections of endoge-
ables that may affect the nous variables from exogenous
yield per hectare of crop k lo- and endogenous model data.
cated in the crop group i. The corresponding economet-
Based on the previous equa- ric equations are estimated
tions, the production of ob- using standard econometric
served crop culture is deter- techniques and simulation will
mined: crop production = be the results of the mathemat-
harvested crop area × crop yield. ical solutions to this set of equa-
When the general form of tions (Figures 64 and 65).
71
Figure 64. Estimation process in R program
The evaluation of the model in the coefficient of determination

terms of testing the significance R2. Then, the estimated results
of regression parameters is per- are evaluated by standard sta-
formed by T-test, and then per- tistical tests for heteroskedas-
forms the significance test of the ticity (white heteroskedasticity)
entire regression by F-test with and autocorrelation.
Figure 65. Estimation process in R program – estimation graph
72
After the process of estimation, Simulation results of the Cro-
estimated equations are cali- atian soft wheat market under
brated so that they correspond baseline assumption are mod-
to standard tests that assess the elled producing medium-term
statistical coherency and quality, projections of soft wheat har-
as well as their compliance with vested area, yield and produc-
necessary biological constraints. tion by 2030. In order to simu-
In order to produce plausible late the effects of irrigation on
simulation validation of the pro- observed variables calibration
duced simulation results has to technique was used. In the
be conducted. This means that first step simulation with the
commodity market experts and equation shown in Figure xy

agricultural economists will ex- was used to produce regular
amine the provisional model baseline results. Irrigation ef-
baseline simulation results and fects are simulated simply by
will provide feedback on the adjusting the regression coef-
model’s projections. Regarding ficient of the trend variable in
their comments, further calibra- the estimated equation in order
tion steps might be undertaken. to produce results with irriga-
By following previously de- tion effect simulation. Namely,
scribed steps which relate to: the importance of irrigation in
• Data collection economic terms is most evi-
• Assumption’s determination dent in the increase in yields
• Setting up econometric be- of cultivated crops where it is
havioral equations estimated that irrigation of the
• Estimation of econometric be- area with average annual rain-
havioral equations fall can have up to 20% higher
• Calibration of econometric yields, while in years with low
behavioral equations annual rainfall up to 40% high-
• Validation of econometric be- er yields (Marković et al., 2018;
havioral equations Marković et al., 2017; Marković
• Re-calibration regarding vali- et al., 2016; Čagalj et al., 2015;
dation if needed Šoštarić et al. 2014). Therefore,
We get a good medium-term we took an arithmetic mean of
simulation that can serve as 30%, and by this percentage, we
a decision-making basis, this increased the regression coef-
approach also relates to ev- ficient of the trend in the esti-
idence-based policy making mated equation for soft wheat
(EBPM) (Cairney, 2016). yield in Croatia.
73
Baseline simulation without ir- ments and mechanisms, chang-
rigation effect indicates that es in the landowner structure
soft wheat yields in Croatia will in favor of larger and more ef-
continue to grow by the end of ficient farms should significant-
the simulated period (Figure ly contribute to further conver-
66). The introduction and devel- gence of productivity per unit
opment of new technologies in area, where Croatia should be
arable farming production, EU near the old Member States in
funding sources, Common Agri- terms of productivity levels.
cultural Policy (CAP) direct pay-
Figure 66. AGMEMOD modelling results – soft wheat yield
However, when we increase the are below the levels of average

trend regression coefficient by producer prices in the EU are the
30% in order to simulate irriga- cause of soft wheat sown area
tion effect on soft wheat yields declining trend (Iljkić et al., 2019).
increase up to 7.6 t/ha com- Therefore, a part of domestic
pared to 6.3 t/ha by 2030. This soft wheat producers has reori-
means 20.5% increase in soft ented their production to other
wheat yields compared to base- crops production whose average
line simulation results by the producer prices are closer to the
end of the simulated period. EU levels (Kranjac et al., 2020).
The soft wheat sown areas in According to the AGMEMOD
Croatia have a declining trend model baseline simulation results,
particularly after Croatia’s acces- it is expected stable amounts of
sion to the EU (Figure 67). Rel- soft wheat sown areas in the fu-
atively low soft wheat average ture with a slight growth trend
producer prices in Croatia, which until 2030 compared to 2018.
74
Figure 67. AGMEMOD modelling results – soft wheat harvested area
When we simulate the irrigation yields provide more incentive to

effect in the trend variable, we producers so they become more
can observe that soft wheat har- willing to sow soft wheat. Given
vested areas slightly increase the expected growth in yields
compared to baseline simulation and stable levels of sown areas,
results. Simulated results show soft wheat production is also ex-
an increase of 1.4% in a harvest- pected to approximately grow
ed area compared to baseline by 25% under the baseline sce-
by 2030. This can be explained nario by 2030 compared to 2018
by fact that higher soft wheat (Figure 68).
75
Figure 68. AGMEMOD modelling results – soft wheat production
Under the simulated effect of type can help decision-makers

irrigation soft wheat production in their decisions to support
volumes in Croatia is expected investment in irrigation infra-
to grow by 22.2% compared to structure projects. As such in-
baseline modelling results by vestments are extremely ex-
2030, and by respectively 53.1% pensive, good impact analysis
compared to 2018 production on production indicators is
quantity. needed and partial equilibrium
In a similar way, this kind of models as sophisticated tools
simulation of irrigation impact can deliver impact assessment
on other key crop markets can analyzes that will simulate such
be applied. Simulations of this effects.
76
77
References
7.
AGMEMOD official website (2021): About AGMEMOD model. Available on:
https://agmemod.eu/
Allen R.G., Pereira L.S., Raes D., Smith M. (1998): Crop evapotranspiration:
Guidelines for computing crop water requirements. Irrigation and drain-
age paper No. 56. Rome, Italy: Food and Agriculture Organization on the
United Nations (FAO).
Brouwer C., Prins K., Kay M., Heibloem M. (1985): Irrigation Water Manage-
ment: Irrigation Methods. Training manual No.5. FAO, Rome, Italy.
Čagalj M., Strikić F., Vitanović E., Ivanković M. (2015): Povećanje poljop-
rivredne proizvodnje u uvjetima navodnjavanja u Imotsko - bekijskom
polju. Zbornik radova "50-i Hrvatski i 10-i Međunarodni Simpozij Agrono-
ma" Opatija, Hrvatska, 16-20. 02. 2015, 133-137.
Cairney P. (2016): The Politics of Evidence-Based Policymaking. London:

Palgrave Pivot.
Chantreuil F., Hanrahan K., van Leeuwen M. (2012): The future of EU agricul-
tural markets by AGMEMOD. Springer, Dordrecht.
Croatian Bureau of statistics (CBS) (2021): Available on: https://www.dzs.hr/
hrv/system/stat_databases.htm
Doorenbos J., Pruit W.O. (1977): Guidelines for predicting crop water re-
quirements. FAO Irrigation and Drainage Paper No. 24. Rome.
EC (European Commission) (2016): EU Agricultural Outlook: prospect for the
EU agricultural markets and income 2016-2026, Publications Office of
the European Union, Luxembourg.
Erjavec E., Donnellan T. (2005): Development of the AG-MEMOD country
level agricultural policy analysis tool in the New Members States of EU.
Paper presented at the 89th EAAE Seminar. Parma, Italy.
Fereres E., Soriano M. A. (2007): Deficit irrigation for reducing agricultural
use. Journal of Experimental Botany, 58(2): 147-159.
Food and agricultural organisation of United Nations, FAO (2016): Introduc-
ing AquaCrop. Available on: http://www.fao.org/3/a-i6321e.pdf
Food and agricultural organisation of United Nations, FAO (2017): Book I.
Understanding Aquacrop. Available on: http://www.fao.org/3/a-i6051e.
pdf
Food and agricultural organisation of United Nations, FAO (2018): AquaCrop
Version 6.0-6.1. FAO crop-water productivity model to simulate yield
response to water. Reference manual.
Food and agricultural organisation of United Nations, FAO (2021):
AquaCrop. Available on: http://www.fao.org/aquacrop/overview/limita-
tions/en/
Hanrahan K.F. (2001): The EU GOLD model 2.1: An Introductory Manual.
Dublin: Teagasc, Rural Economy Research Centre.
78
Iljkić D., Kranjac D., Zebec V., Varga I., Rastija M., Antunović M., Kovačević
V. (2019): Stanje i perspektiva proizvodnje žitarica i uljarica u Republici
Hrvatskoj. Glasnik zaštite bilja, 3: 58-67.
Jansik C., Kettunen L., Lehtonen H., Niemi J. (2006): Agricultural policy anal-
ysis with the AGMEMOD model: A new super model takes the stage?
Suomen Maataloustieteellisen Seuran Tiedote, 21: 1-7.
Kranjac D., Zmaić K., Grgić I., Salamon P., Erjavec E. (2020): Accession impact
and outlook for Croatian and EU crop and livestock markets. Spanish
Journal of Agricultural Research, 18(1): e0103.
Levert F., Chantreuil F. (2006): Building the AGMEMOD database: use of
Eurostat data and common rules for a coherent database (revision 6) —
notes and guidelines 4, Paris.

Lu N., Likos J. (2004): Rate of Capillary Rise in Soil. Technical notes. Available
on: https://www.pc-progress.com/Images/Personal/NLu/Publications/
NLJ32200403.pdf
M’Barek R., Britz W., Burrell A., Delincé J. (2012): An integrated modelling
platform for agro-economic commodity and policy analysis (iMAP) - a
look back and the way forward. Publications Office of the European
Union, Luxembourg, 9–91.
M’Barek R., Delincé J. (2015): An integrated modelling platform for agro-eco-
nomic commodity and policy analysis—new developments and policy
support 2012–2014. Publications Office of the European Union, Luxem-
bourg, 1–64.
Marković M., Josipović M., Ravlić M., Josipović A., Zebec V. (2016): Deficit irri-
gation of soybean (Glycine max. (L.) Merr.) based on monitoring of soil
moisture, in sub-humid area of Eastern Croatia. Romanian Agricultural
Research, 33: 1-8.
Marković M., Josipović M., Šoštarić J., Jambrović A., Brkić A. (2017): Response
of maize (Zea mays L.) grain yield and yield components to irrigation
and nitrogen fertilization. Journal of central European agriculture, 18 (1):
55-72.
Marković M., Šoštarić J., Josipović M., Barač Ž., Brkić A. (2018): Yield and
yield components of maize (Zea Mays L.) hybrids as affected by irriga-
tion scheduling and meteorological conditions. Agriculturae conspectus
scintificus, 83 (1): 39-44.
McCarl B.A. (2010): McCarls GAMS user guide, Version 24.1. Available on:
https://www.gams.com/latest/docs/userguides/mccarl/mccarlgamsuser-
guide.pdf
Raes D. (2009): The ETo calculator, Reference Manual. Version 3.1.
Raes D., Steduto P., Hsiao T.C., Fereres E. (2009): AquaCrop-The FAO crop
model to simulate yield response to water: II. Main algorithms and soft-
ware description. Agronomy Journal, 101(3): 438-447.
79
Raes D., Steduto P., Hsiao T.C., Fereres E. (2018): AquaCrop Version 6.0 –
6.1. Reference manual. Available on: http://www.fao.org/3/a-br246e.pdf
Rai R.K., Singh V.P., Upadhyay A. (2017): Planning and Evaluation of Irriga-
tion Projects. Methods and Implementation, 505-523.
Salamon P., Banse M., Barreiro-Hurlé J., Chaloupka O., Donnellan T., Erja-
vec E., Fellmann T., Hanrahan K., Hass M., Jongeneel R., Laquai V., van
Leeuwen M., Molnár A., Pechrová M., Salputra G., Baltussen W., Efken
J., Hélaine S., Jungehülsing J., von Ledebur O., Rac I., Santini F. (2017):
Unveiling diversity in agricultural markets projections: from EU to
Member States. A medium-term outlook with the AGMEMOD model. JRC
Technical Report. 29025 EUR, Publications Office of the European Union,
Luxembourg.
Salputra G., Chantreuil F., Hanrahan K., Donnellan T., Leeuwen van M., Erja-
vec E. (2011): ‘Policy harmonized approach for the EU agricultural sector

modelling’, Agricultural and Food Science, 20: 119-130.
Salvatici L., Anania G., Arfini F., Conforti P., De Muro P., Londero P. (2001):
Recent developments in modelling the CAP: hype or hope? Vo T. Hecke-
lei, H. P. Witzke, & W. Henrichsmeyer, Agricultural Sector Modelling and
Policy Information Systems. Proceedins of the 65th European Seminar
of EAAE Kiel: Wissenschaftsverlag Vauk Kiel KG, 8-26.
Soil Science Glossary Terms Committee (2008): Glossary of Soil Science
Terms 2008. Madison: SSSA, p. 92.
Šoštarić J., Begić S., Salkić B., Kovačević V., Marković M. (2014): Variation of
winter wheat yields in Croatia and Bosnia and Herzegovina among years
with aspect of climatic changes. Turkish Journal of Agricultural and Nat-
ural Sciences, special issue (2014), 1: 1364-1368.
USDA, United States Department of Agriculture (1997): Irrigation guide.
Chapter 9, Irrigation Water Management.
USDA, United States Department of Agriculture (2016): Restrictive layer
depth. Available on: https://datagateway.nrcs.usda.gov/
Walker W.R. (1989): Guidelines for designing and evaluating surface irriga-
tion systems. FAO irrigation and drainage paper 45. Rome, Italy.
80

CO7 - Application of Computer Models in Irrigation Management

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CO7 - Application of Computer Models in Irrigation Management

Uploaded by

Copyright:

Available Formats

APPLICATION OF

For the publisher

© All rights reserved. No parts of this book may be reprinted or transferred

Josip Juraj Strossmayer University of Osijek

The approval for publishing the university textbook was issued

5. Databases in water management 50

6. Partial equilibrium models 55

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

er station networks, soil, and to calibrate and validate the

1.2. Wireless sensors

1.3. GIS (Geographic Information System)

management in irrigated agri-

The CropWat 8.0 model pro- can be found: http://www.

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

Figure 3. Station and climate data

agement purposes (Fereres and can be used by the plants

Figure 5. Effective rainfall calculation

full production under the given stressed, well-managed crops in

Figure 6. Crop data

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

• Crop development stage • Critical depletion fraction - is

remaining in a soil 2 or 3 days

Figure 8. Soil water content

Total available water differs as the depth of water that has

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

Figure 9. Soil properties

Table 1. Infiltration rates for various soil types

ing moisture content and multi- requires a thorough evaluation

2.3. Analysis of Crop Water Requirements

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

Now the CWR is expressed as mm of effective rain during the

Figure 11. Irrigation scheduling

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

Figure 12. Scheduling criteria for non-rice crops

In the irrigation scheduling for the entire growing season

Figure 13. Irrigation scheduling graph

Figure 14. Reference grass ETo

Also, the soil is not considered as a reference method (Raes,

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

Figure 15. The interface of the ETo calculator

Figure 16. Station description and location description and

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

For this example, the mean air sured at 2 m height. Sunshine

Figure 18. Input climate data limits

Once given period can be cor- as daily, then in this step it is

Figure 19. Time range determination

Figure 20. The input of meteorological data

After climate data have been software calculates daily ETo.

Figure 21. Reference evapotranspiration presented as a plot

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

Figure 23. AquaCrop flowchart

approach by separating (i) the • the relatively low number of

4.2. AquaCrop application

APPLICATION OF COMPUTER MODELS IN IRRIGATION MANAGEMENT

Figure 24. AquaCrop interface

diation (MJ/day), rainfall (mm),

Figure 25. Climate data interface Figure 26. Rainfall data

4.2.2 Crop file according to calendar days.

Figure 29. Crop file Figure 30. Crop development

If water stress occurs after tion might be adjusted down-

33. Harvest index 34. Root deepening