AN INTEGRATED MODELLING APPROACH FOR IRRIGATION WATER

MANAGEMENT USING SALINE AND NON-SALINE WATER: THE

SALTMED MODEL
R. RAGAB1

Centre for Ecology and Hydrology, CEH, Wallingford, OXON, OX10, 8 BB, UK
E-mail: Rag@ceh.ac.uk

ABSTRACT

A successful water management scheme for irrigated crops requires an integrated

approach that accounts for water, plant, soil and field management. For that purpose,
the SALTMED model has been developed. The model runs on a PC under Windows
95/98 operation System. The model's input consists of: Climate data, Soils data, crop
data, irrigation data (System, amount, salinity), soil parameters, crop parameters, and
other model parameters. The model has default values and includes database for soils
and crops. In the model, the Richards Equation and the Convection-Dispersion Equation
describe the water and solute movements respectively. The daily potential and actual
evapotranspiration were calculated using Penman-Monteith equation according to FAO
Irrigation & Drainage paper No. 56. The model runs for a variety of irrigation systems,
crops, soils, and water salinity levels. The daily model output ( graphs and data files)
includes, yield, potential and actual water uptake, salinity, soil matric potential and soil
moisture profiles, crop water requirements, leaching requirements, plant growth
parameters, Potential and actual evapotranspiration, bare soil evaporation and plant
transpiration. The model is friendly and easy to use benefiting from the windows
environment.

Key words: SALTMED, Salinity, Integrated Management

I. INTRODUCTION

The limited fresh water resources and the increasingly need for food and fibre for the growing
populations in so many parts of the world, has put a lot of pressure on these resources. In a
competition with other sectors for fresh water, agriculture sector usually looses out. As a result,
many agronomists all over the world sought other alternative resources some times known as non-
conventional e.g. re-use of drainage water, use of brackish water (groundwater), use of seawater,
desalinate saline water, etc. In all these non-conventional methods a great care should be paid when
applying saline waters for irrigation. In order to protect the environment and secure satisfactory
level of food and fibre production, an integrated approach for management is the key factor.

1
Centre for Ecology and Hydrology ( formerly Institute of Hydrology), Wallingford, Oxon, OX10 8BB, UK
Chairman of the Work Group on the " Sustainable use of Natural resources for crop production" of the International
Commission on Irrigation and Drainage, ICID
Chairman of the Work Team on the "Use of poor quality waters for irrigation", ICID

6th International Micro-irrigation Congress South Africa, 22 – 27 October 2000

‘Micro-irrigation Technology for Developing Agriculture’ Organised by: Conference Planners
Conference Papers CD Produced by: Document Transformation Technologies
Management of soil or crop or water in isolation would lead to soil deterioration, increasing water
table levels causing further soil salinization, reducing oxygen levels in the root zone and affecting
nutrients availability and possibly creating wet lands which could be a health hazard (e.g. Malaria,
etc.). As salinity impact on soil and the environment is a long term process and therefore short term
experiments are unable to show that, a modelling activity to predict the long term effect on soil and
environment, crop yield, soil water and salinity profiles under different strategies of water
management (blending & alternative use of fresh and saline waters), water and leaching
requirements are being developed in parallel with the field and greenhouse experiments.

The objective of the SALTMED project is to increase productivity and sustainability of irrigated
crops in areas which have intrinsically, or potentially, saline water supplies. Part of its objectives is
to develop the SALTMED model which when successfully calibrated and tested could be run with
"what if scenarios" to help in producing a set of guidelines. A generic approach was chosen by
employing. different crops, soils, irrigation systems and water management strategies. The latter has
two options A- blending: apply mixed saline and fresh water with different ratios and B- Cyclic:
alternatively apply fresh and saline water i.e. fresh water when the crop is at sensitive stage for
salinity and saline water when the crop is at a more salinity tolerant stage. Water and solute flow
were treated mathematically as one-dimensional flow if irrigation system is sprinkler, basin, rain
fed and two-dimensional if irrigation system is furrow, trickle line source and Three-dimensional if
irrigation is a trickle point source.

II. THE SALTMED MODEL

The SALTMED model includes the following key processes and possibly more (if required).
Evapotranspiration, Plant water uptake, Water and solute transport under different irrigation
systems and Crop yield - water use relationship. Brief description is given here under:

I. Evapotranspiration

The evapotranspiration has been calculated using the modified FAO ( 1998) version of Penman-
Monteith equation of 1965 in the following form:

900
0.408∆ ( Rn − G ) + γ U 2 ( eS − ea )
T + 273
ET 0 = (1)
∆ + γ (1 + 0.34U 2)

where ETo is the reference evapotranspiration ( mm day-1) Rn is the net radiation,

(MJ m-2 day-1 ), G is soil heat flux density, (MJ m-2 day-1 ), T is the mean daily air temperature at 2 m
height, (oC ), ∆ is the slope of the saturated vapour pressure curve, (kPa oC-1 ), γ is the psychrometric
constant, 66 Pa oC -1, es is the saturated vapour pressure at air temperature, kPa, ea is the prevailing
vapour pressure, kPa, and U2 is the wind speed at 2 m height, m s-1, The ETo reference calculated
here is for short well-watered green grass.

The crop evapotranspiration ETc is calculated as:

ETc = ET 0( Kcb + Ke ) (2)

where Kcb is the crop transpiration coefficient ( known sometimes as basal crop coeeficient) and Ke
is the soil evaporation coefficient.
II. Plant water uptake in presence of salts

There are several models to estimate the plant water uptake however there is a limited number of
them which deals with water uptake under saline conditions. An example of these is the one
suggested by Cardon and Letey (1992) where the water uptake S (d-1) is estimated as:

 
 
S max (t )
S ( z, t ) =  
3 λ ( z, t ) (3)
  a (t )h + π  
1 +   
  π50 (t )  

where Smax (t) is the maximum potential root water uptake (d-1) at the time t (d), z is the vertical
depth taken positive downwards (cm), λ(z,t) is the depth-and time-dependent fraction of total root
mass, h is the matrix pressure head (cm), π is the osmotic pressure head (cm), π50 (t) is the time-
dependent value of the osmotic pressure at which Smax(t) is reduced by 50%, and a(t) is a weighing
coefficient that accounts for the differential response of a crop to matrix and solute pressure. The
coefficient a(t) equals π50(t)/h50(t) where h50(t) is the matrix pressure at which Smax(t) is reduced by
50%. Equation 3 has been implemented in the model. The growth rate and final yield depend very
much on the water uptake, therefore, a crop-water-yield function could be developed based on the
results obtained from this equation.

III. Water and solute flow

The vertical transient-state flow water in a stable and uniform segment of the root zone can be
described by a Richard's type equation as:

∂θ ∂  ∂ (ψ + z ) 
= −  K (θ )  − Sw (4)
∂t ∂z  ∂z 

where θ is volume wetness; t is the time; z is the depth; K (θ) is the hydraulic conductivity (a
function of wetness); ψ is the matrix suction head; and Sw is the sink term representing extraction
by plant roots.

The movement of solute in the soil system, its rate and direction, depend greatly on the path of
water movement, but it is also determined by diffusion and hydrodynamic dispersion. If the latter
effects are negligible, solute flow convection can be formulated as:

J c = qc = v θc (5)

where Jc is the flux density; q is the flux density of the water; c the concentration of solute in the
flowing water and v the average velocity of flow. The rate of a diffusion of a solute (Jd) in bulk
water at rest is related by Fick’s law to the concentration gradient as:

J d = Do (∂c ∂x ) (6)

where D0 is the diffusion coefficient. In the soil the diffusion coefficient Ds is decreased due to the
fact that the liquid phase occupies only a fraction of soil volume, and also due to the tortuosity
nature of the path, therefore, it can be expressed according to the following equation:
 Ds = D0θξ (7)

where ξ is the tortuosity, it is an empirical factor smaller than unity, which can be expected to
decrease with decreasing θ.

The convection flux generally causes hydrodynamic dispersion too, an effect that depends on the
microscopic non-uniformity of flow velocity in the various pores. Thus a sharp boundary between
two miscible solutions becomes increasingly diffuse about the mean position of the front. The
importance of the diffusion coefficient bas been found by Bresler (1973) to depend linearly on the
average flow velocity v , as follows:

Dh = αv (8)

where α is an empirical coefficient. By the combination of the diffusion and the dispersive
coefficient with the convective transport equation an overall of the flux of solute can be obtained as:

J = −( Dh + Ds )(∂c ∂x ) + v θc (9)

If one takes the continuity equation into consideration, one-dimensional transient movement of a
non-interacting solute in soil can be expressed as:

∂ (θc) ∂  ∂c  ∂ (qc)
=  Da  − − Ss (10)
∂t ∂z  ∂z  ∂z

in which c is the concentration of the solute in the soil solution, q is the convective flux of the
solution, Da is a combined diffusion and dispersion coefficient, and Ss is a sink term for the solute
representing root adsorption/uptake.

IV. Water and solute transport under irrigation systems

Under irrigation from a trickle line source, the water and solute transport can be viewed as two-
dimensional flow and can be simulated by one of the following:

1) a “plane flow” model involving the Cartesian co-ordinates x and z. Plane flow takes place if one
considers a set of trickle sources at equal distance and are close enough to each other so that their
wetting fronts overlap after a short time from the start of the irrigation.

2) a “cylindrical flow” model described by the cylindrical co-ordinates r and z.

Cylindrical flow takes place if one considers the case of a single trickle nozzle or a number of
nozzles spaced far enough apart so that the overlap of the wetting fronts of the adjacent sources
does not take place.

For a stable, isotropic and homogeneous porous, the two-dim. flow of water in the soil can be
describes according to Bresler (1975) as:

∂θ ∂  ∂ψ  ∂  ∂ (ψ + z ) 
=  K (θ )  +  K (θ ) (11)
∂t ∂x  ∂x  ∂z  ∂z 
where x is the horizontal co-ordinates; z is the vertical-ordinate; K (θ) is the hydraulic conductivity
of the soil. Considering isotropic and homogeneous porous media with principal axes of dispersion
oriented parallel and perpendicular to the mean direction of flow, the hydrodynamic dispersion
coefficient Dij can be defined as follows:

λ T V δ ij + (λ L − λT )ViV j
Dij = (12)
V + D p (θ )

where λL is the longitudinal dispersivity of the medium; λT is the transversal dispersivity of the
medium; δij is Kronecker delta; Vi is the i component of the average interstitial solution velocity V;
and Dp (θ) is the soil diffusion coefficient as defined by Bresler (1973). If one considers only two
dimensions and substituting Dij , the salt flow equation becomes:

∂ (Cθ ) ∂  ∂C ∂C  ∂  ∂C ∂C 
=  Dxx + Dxz − q x C +  Dzz + Dzx − q z C (13)
∂t ∂x  ∂x ∂z  ∂z  ∂z ∂x 

Bresler (1975) considered that in presence of a trickle source the water flow and solute transport can
be treated as a two-dimensional problem using axi-symmetric cylindrical model where the
cylindrical co-ordinates (x, r) are considered ( e.g. the radius r replaces z in equation 13).

V. Relative crop yield

The relative crop yield, RY is estimated as the sum of the actual water uptake over the season
divided by the sum of the maximum water uptake (under no stress condition) as:

RY =
∑ S ( x, z, t ) (14)
∑ S ( x, z, t )
max

The actual yield, AY is simply obtainable by :

AY = RY *Y max (15)

where Ymax is the maximum yield obtainable in a given region under optimum condition.

VI. Leaching requirements

This simply calculated as a ratio between the salt concentration of the irrigation water to that of the
drainage water or the mean salinity level of the root zone.

VII. Draiange

At this stage free drainage at the bottom of the root zone is considered.

VIII. Data requirements

1: Plant characteristics these include the crop coefficient , Kc or the Leaf area index. LAI
(periodical measurements), root depth and lateral expansion, crop height ( periodical
measurements) and maximum / potential final yield
2: Soil characteristics these include, soil horizon depths and texture, saturated hydraulic
conductivity, infiltration rate, salt diffusion, hydrodynamic dispersion, initial soil moisture and
salinity profiles.

3: Meteorological data these include daily values of Temperature (max.), Temperature

(Minimum), relative humidity, net radiation, wind speed, etc. and daily rainfall

4: Water Management data these include the date and amount of irrigation water applied and
the salinity level of each irrigation applied

5: Default Data in the Data base

The model has built -in data base which contains, different crops from different regions and their
growth stages, Kc values for each growth stages ( based largely on the FAO paper 56, 1999),
maximum height and maximum rooting depth. In addition, there is at least 14 different soil types
and their hydraulic parameters. Also included, default solute transport parameters.

IX. Model output.

The model output results as text and graphical files.

These include: horizontal and vertical soil moisture, soil salinity, relative concentration and, soil matric
potential profiles , time series of reference, crop transpiration, bare soil evaporation, leaching
requirements, irrigation amounts, Kc , Kcb , Root depth and yield (ton ha-1 ). The model is being tested
and calibrated and a future paper will present the final results. Examples of the input /output are shown
here under.

III. CONCLUSION

In this paper the principals and the objectives of the SALTMED model have been highlighted. It
was shown how the model adopts integrated management approach and allows different water
application strategies. The model is friendly and easy to use benefiting from the windows
environment, however it is physically based model uses the well known water and solute transport
equations. A follow up paper not in a distant future will show the results of this model

IV. REFERENCES

BRESLER, E. 1973. Simultaneous transport of solutes and water under transient unsaturated flow
conditions. Water resources research, Vol. 9, No.4, 975-985.

BRESLER, E. 1975. Two-dimensional transport of solute during nonsteady infiltration for a trickle
source. Soil Sci. Soc. Amer. Proc, Vol. 39, 604-613.

CARDON, E.G., J. LETEY. 1992. Plant water uptake terms evaluated for soil water and solute
movement models. Soil Sci. Soc. Am. J. 32:1876-1880.

FAO, 1998. Crop evapotranspiration, Irrigation and Drainage paper No 56., Rome, Italy

MONTEITH, J.L. 1965. Evaporation and environment. In G. E. FOGG (ed.) Proc. Symp. Soc. Exp.
Biol. 19: 205-234. Univ. College of Swansea. 8-12 Sept. 1964. Cambridge Univ. Press, London