Unsaturated zone modelling: the role of soil database classification

Abstract

Physical and hydraulic soil properties are essential input parameters for models from different sciences (e.g. hydrology, agriculture, water management, nature preservation). Generally texture composition, porosity and other easily measurable physical properties of soils are known. However, saturated hydraulic conductivity and characteristic values of the water retention curve are usually missing information. Therefore, based on the physical similarity of soils (classes), they are substituted by data derived from soil databases. The aim of this study was to assess the currently unknown uncertainties of such classified databases. To do so, a large variety of tests were carried out: (i) static and dynamic, (ii) 1D and 3D (iii) hydraulic and hydrologic applied tests, (iv) real and synthetic soils, parameterized accordingly, and (v) HUNSODA and/or HYPRES databases. The results were sorted with respect to FAO and USDA classification systems. Soil class overlapping was evaluated through the statistics of basic hydraulic parameters (retention curve, hydraulic conductivity). Indicators related to hydrologic extremities (excess water and drought) were used to quantify the uncertainties of soil texture based on parameter substitution. It was concluded that the two evaluated soil classification systems did not sort soils reliably from the hydrologic and hydraulic viewpoint: the test results of classes showed major overlaps. Moreover, in most cases class synthetic parameter combinations poorly represented real soils. As a general consequence the results based on classified soil databases should be accepted only with reservation.

Full text

Unsaturated zone modelling:
the role of soil database classification
Z. Kozma, T. Ács & L. Koncsos
Department of Sanitary and Environmental Engineering,
Budapest University of Technology and Economics, Hungary
Abstract
Physical and hydraulic soil properties are essential input parameters for models
from different sciences (e.g. hydrology, agriculture, water management, nature
preservation). Generally texture composition, porosity and other easily
measurable physical properties of soils are known. However, saturated hydraulic
conductivity and characteristic values of the water retention curve are usually
missing information. Therefore, based on the physical similarity of soils (classes),
they are substituted by data derived from soil databases. The aim of this study was
to assess the currently unknown uncertainties of such classified databases. To do
so, a large variety of tests were carried out: (i) static and dynamic, (ii) 1D and 3D
(iii) hydraulic and hydrologic applied tests, (iv) real and synthetic soils,
parameterized accordingly, and (v) HUNSODA and/or HYPRES databases. The
results were sorted with respect to FAO and USDA classification systems. Soil
class overlapping was evaluated through the statistics of basic hydraulic
parameters (retention curve, hydraulic conductivity). Indicators related to
hydrologic extremities (excess water and drought) were used to quantify the
uncertainties of soil texture based on parameter substitution. It was concluded that
the two evaluated soil classification systems did not sort soils reliably from the
hydrologic and hydraulic viewpoint: the test results of classes showed major
overlaps. Moreover, in most cases class synthetic parameter combinations poorly
represented real soils. As a general consequence the results based on classified soil
databases should be accepted only with reservation.
Keywords: unsaturated zone, soil database, soil classification, hydrological
modelling, agricultural yield estimation.
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
doi:10.2495/SI140181
Sustainable Irrigation and Drainage V 197

1 Introduction
The hydraulic behaviour of unsaturated soils is an important factor in a number of
environmental phenomena. The rate of water and solute transport in the
unsaturated zone directly or indirectly governs most hydrologic processes, as well
as erosion, plant growth, ecosystem functioning, nutrient cycling and diffuse
pollution.
According to its importance, unsaturated zone is assessed in the frame of many
sciences at various scales. Among others hydrologic, water management,
agricultural (e.g. crop yield – McCown et al. [1], van Ittersum and Rabbinge [2])
and climate (e.g. regional atmospheric) models must handle the impacts of
hydraulic processes in the unsaturated zone to a certain extent. The topic has a
great significance in Hungary as well, since the country has to face a unique
combination of extreme hydrologic symptoms: the extensive, regular appearance
of flood, excess water (see definition in e.g. van Leeuwen et al. [3]) and drought.
As a result of increased computational capacity the theoretical insight is
broadening remarkably in the topic. Up-to-date one- or multidimensional solvers
as well as integrated hydrologic models attain great complexity and became
essential parts of research (Harter and Hopmans [4]). However, there are still many
open questions and problems in relation with understanding and predicting the role
of vadose zone in landscape scale hydrology (Pachepsky et al. [5]). Probably the
most important of these issues is data reliability. Considering the quality of the
simulation results, usually the bottleneck is the input data. Soil maps and databases
(Wösten et al. [6], Soil Survey Staff [7]) are used expansively to provide
information about the 3D spatial distribution of various soil types and their
hydraulic parameterization (e.g. water retention and hydraulic conductivity). Such
databases are often developed by using soil classification. Though it is a common
practice, even the magnitude of uncertainty associated to soil categorization is
practically unknown.
Concerning the aspects above, recently a research was started with the aim to
(i) quantify the uncertainty related to classified soil data used for hydrologic
modelling, and to (ii) derive useful information and hydrologic indicators for soil
samples from Hungary. In the first phase of the study (Kozma et al. [8]) the
well-known FAO classification system (FAO [9]) was analysed by using it on the
Unsaturated Soil Hydraulic Database of Hungary (HUNSODA – Nemes [10]).
Three possible directions were outlined for the second phase of the research:
(a) the dynamic analysis of all soils in the database instead of using only
parameter class envelopes,
(b) the evaluation of the USDA classification system (USDA [11]) with the
same methodology, and
(c) the analysis of the FAO and/or USDA methods using the Database of
Hydraulic Properties of European Soils (HYPRES).
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
198 Sustainable Irrigation and Drainage V

In this paper task (a) and (b) will be discussed in detail. Since the complete
HYPRES database is not accessible for the broad scientific community, task (c) is
limited to the published FAO class averages derived for HYPRES.
The time-invariant and dynamic tests applied so far consider only a single
homogeneous soil column (pedon), therefore their ability to analyze complex
landscape scale hydrologic behaviour is limited. For this reason, another type of
test, a simple sensitivity analysis was carried out as well: hydrologic processes and
excess water were simulated with the WateRisk integrated hydrologic model
(Kozma and Koncsos [12]) at a relatively humid Hungarian lowland watershed.
2 Theoretical background
2.1 Classified soil databases
As the direct determination of soil properties is expensive and time-consuming,
such data is often provided by pre-parameterized spatial soil databases. These
contain the 2D/3D distribution of soil types and physical parameters associated to
them. The magnitude of uncertainty related to database elaboration is usually
unknown, but probably there are two major sources of error: (i) spatial
interpolation between sample locations, and (ii) soil classification. As stated
above, in this research only (ii) is analyzed.
The aim of classification is to synthesize field data, so that predictive
estimates/concepts can be set up about soil behaviour. Samples are categorized
into soil types with respect to easily definable properties (e.g. genesis,
morphology, texture). Classes are then characterized with geometric average
values for also such properties that are more difficult to quantify or expensive to
measure and can be missing for many soils (water retention, saturated hydraulic
conductivity, etc.). This way classification allows the extrapolation of scarcely
known soil information.
Figure 1: The soil texture triangle showing class limits for (a) the USDA
method and for (b) the FAO method.
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
Sustainable Irrigation and Drainage V 199

Two of the most widespread systems are the FAO [9] and the USDA [11]
method. Both robust approaches sort mineral soils into classes with respect to their
textural composition (sand, silt and clay content). The applied limit values are
shown on Figure 1. The USDA method uses 12 classes (class names are created
from the words loam, sand, silt and clay, e.g.: sicL – silty clay Loam), while the
FAO system uses only five types (C – coarse; M – medium; MF – medium-fine;
F – fine; VF – very fine).
One would expect that the twelve-class USDA method provides more detailed
and sound representation of soils. Contrary, Wösten et al. [6] and Nemes [10]
suggest that the estimation of hydraulic behaviour is more reliable in case of the
FAO method. This might be a reason, why this method is more accepted in Europe:
among others the HYPRES as well as the HUNSODA was developed with the
FAO method.
2.2 Soil hydraulic functions
Water flow and solute transport in various soils can differ significantly. To
describe hydraulic behaviour, generally two characteristic functions are used: the
WRC and the hydraulic conductivity function (HCF). The WRC provides the
water content (

[m3 m-3]that a soil can hold at various matric pressures
(

[cm]). Figure 2 illustrates the physically interpretable content of the WRC for
an arbitrary medium textured soil (note that pF = log(-

).
Experimentally the WRC is defined by measuring



data pairs. Various
parametric analytical formulas can be used to express the WRC. These functions
have to be fitted to the measured discrete data points. Most accepted among these
is the function introduced by van Genuchten (vG) [13], which was used in this
study as well. In eqn. (1)

and n are fitted parameters, while m = 1 – 1/n.
Figure 2: The water retention curve of a medium textured soil, showing
characteristic water content values.
residual water content:

r
=

(pF 6.2);
water content at wilting point:

wp
=

(pF 4.2);
water content at field capacity:

fc
=

(pF 2.5);
saturated water content:

s
=

(pF 0);
available water content:
AWC =

fc
-

wp;
specific yield:
SY =

s
-

fc
.
0
1
2
3
4
5
6
7
pF [-]
WRC
wilting point
field capacity
AWC

r

s

fc

wp
SY
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
200 Sustainable Irrigation and Drainage V

  


m
n
rsr αψ1θθθψθ 
 (1)
Hydraulic conductivity function describes the variation of the k(

)/Ks [-] ratio
with respect to water content (where k(

) [cm day-1] is the unsaturated hydraulic
conductivity and Ks [cm day-1] is the saturated hydraulic conductivity). Since Ks
values can vary in a wide range within soil classes, usually this is the dominant
source of uncertainty. Moreover, the HCF is strongly non-linear, also ranging over
several orders of magnitude (OMs), and its measurement is heavily biased with
error. In this research for the HCF – as van Genuchten suggested [13] – the
Mualem model was used.
3 Time-invariant analysis
In the first phase of this research FAO classes of the HUNSODA database showed
significant uncertainties in soil hydraulic parameters (Kozma et al. [8]). Therefore
the database has been categorized and analyzed also with the USDA method.
Figure 3: Comparison of (a) FAO and (b) USDA class statistics for specific
yield and available water content and for saturated hydraulic
conductivity.

-0.4
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
CMMFFVF
water content [-]
S
lS
sL
scL
L
siL
cL
Si
C
sicL
siC
-150
-100
-50
0
50
100
150
number of soils [-]
FAO USDA
SY AWC nSoil
1.E-02
1.E-01
1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
C MMFFVF
Ks [cm day
-1
]
S
lS
sL
scL
L
siL
cL
Si
C
sicL
siC
FAO USDA
Mean
Max
Min
USDA 1988
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
Sustainable Irrigation and Drainage V 201

Figure 3 illustrates the number of soils in classes, class averages and their
variation for AWC, SY and Ks values. The AWC and SY are derived from

values
at two pF level, thus they provide the best single value aggregated information
about the WRC shape.
Three remarks can be taken: (i) the number of soils in certain classes is diverse.
The VF class of FAO and four USDA classes contain less than 20 samples, which
questions the reliability of any derived statistical measure for these classes (the sC
USDA class contains no soils, so it is not included on the graph). (ii) The relative
values of class variances are between 20–80% (FAO) and 15–72% (USDA) of the
average, indicating major diversity of the WRC shapes. And finally (iii) variances
often exceed the change between class averages – especially in case of mixed,
medium textured soil types.
The saturated hydraulic conductivity varies within a wide range: values in
certain classes cover 3 to 5 OMs for both methods. The only exceptions are the S,
scL and Si USDA classes with only 2 OMs intervals. However scL and Si contain
only 11 and 9 soils respectively, which in fact emphasizes the diversity of Ks.
Large values of class variances (displayed with error bars) indicate great scatter in
data as well.
At the same time, class averages show surprisingly moderate trends: in case of
the FAO method 2 OMs are barely covered and the physically expected
monotonous decline is biased by the known anomaly of the VF class (Nemes [10],
Wösten et al. [6]). For the USDA method the trend is more obvious: when moving
from coarser to finer textured soils, average Ks decreases over 3 OMs. Only the
sicL and siC mixed classes serve as exceptions: their lower conductivity is usually
attributed to their special textural composition.
The very fine FAO class contains only 13 WRCs and 6 Ks values. Such small
sample demonstrates well, how uncertainties of class averaging arise. As an
obvious consequence, class statistics are rather obscure. A more complex effect
can also be observed: for soils with measured Ks values the WRC decreases
relatively fast: average SY is 0.085, indicating that the theoretical capillary
fringe would reach up only to 100-200 cm above the water table. Thus these soils
– regardless of their Ks – have limited water conveying capacity under unsaturated
conditions. However, for the 7 soils without Ks the lower portion of the WRC is
steep, the average SY is 0.012, and the capillary fringe would extend to more than
10 meters. By averaging all 13 WRCs (determining geometrical means of

s at all
pF levels), the summation curve has a SY of 0.010 (note that the resultant SY is
not a linear combination of individual SY values), which would lead to great
capillary rise. As a result, with its large averaged Ks value the synthetic soil
representing the VF class average shows unexpected and rather unrealistic
hydrologic behaviour (see results of VF class in Figure 5).
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
202 Sustainable Irrigation and Drainage V

4 Dynamic numerical analysis
Beside the time-invariant analyses two transient numerical tests were applied on
soils of HUNSODA database to assess how the revealed uncertainties would affect
the dynamic behaviour of soils and soil classes. In the first one 1D unsaturated
flow was simulated in soil columns with Hydrus-1D (Simunek et al. [14]),
focusing mainly on the hydraulic responses of certain soils to real meteorological
conditions. In the second test the WateRisk model was used in order to evaluate
the hydrologic behaviour of soil types at heterogeneous regional scale. The second
test is discussed under section 4.2.
Due to the climatic, topographic and hydrographical conditions of the Great
Hungarian Plan, excess water and drought occur regularly, even in the same year.
Therefore, assessment of soils was carried out with emphasis on these hydrologic
extremities. Model input data (meteorology for the 1D analysis; topology,
hydrometeorology and HUNSODA based soil map for the 3D analysis) was
selected from affected lowland regions.
4.1 1D unsaturated flow test
The HUNSODA database contains 840 soil samples, of which 576 have ten-point
WRC and 252 have measured Ks values. Due to lack of measured Ks or measured



data pairs, 588 records were neglected from further assessment. Additionally,
the test was applied on two other datasets as well, namely the collection of fictive
soils characterized by the five-five FAO class average parameter sets derived from
the (i) HYPRES and (ii) HUNSODA databases.
Atmospheric boundary condition with 100 cm allowed maximum water cover
depth was set at ground surface of the soil column. Daily precipitation (P) and
temperature time series recorded at Napkor weather station for the period
2010–2012 were used as input. Potential evapotranspiration PET was estimated
by the robust Varga-Haszonits equation (Varga-Haszonits [15]). The evaluated
time period is relevant for both excess water and drought. 2010 was the wettest
year of the last century (P ≈ 900 mm and PET ≈ 740 mm) while 2011 and 2012
were extreme dry years (P ≈ 730 mm and PET ≈ 2180 mm for the two years).
Vegetation with 50 cm root depth was assumed. Plant rooting distribution
decreased linearly with depth from 1 to 0. For the separation of evapotranspiration
to evaporation and transpiration, LAI for crops was used.
Soil columns were homogeneous and depth was set to 20 m. The possibly finest
(1001 nodes), uneven discretization was used to minimize numerical problems.
Constant flux (Q = 0 m3day-1) was used as lower boundary condition.
Considering drought, one of the most important question from agricultural and
nature preservation viewpoint is the ratio of water uptake and water demand of
plants. As a rather practical and simple approach, this ratio was assessed through
transpiration. Among other characteristics current excess water management
practice focuses mainly on spatial and temporal extent of water coverage. In case
of a 1D approach only the latter can be analyzed (see the 3D analyses for
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
Sustainable Irrigation and Drainage V 203

evaluation of other factors). Both extremities are affected by the status of the
groundwater. Considering the above mentioned aspects, evaluation of soils was
carried out through three hydrologic indicators. These are:
(a) T / TP, where T [cm] stands for actual transpiration and TP [cm] is
potential transpiration;
(b) ExWat cover – total duration of excess water [days];
(c) Avg GW level – average groundwater level [m below/above surface].
Figure 4 illustrates the variation of the indicators as a function of time for a
randomly chosen soil profile. Note that evaporation is not displayed on the figure.
Results of converged simulations are summarized on Figure 5. Coloured stripes
represent the ± one sigma interval while vertical black lines show the minimum
and maximum values of the indicators for each class. To emphasize the trends
related to classification, medians were also displayed on the figures.
The median of the T / TP and ExWat cover indicators denote moderate trends
that meet the general expectations: the finer the soil (i) the less water it can
transport to the root zone and (ii) the longer it takes to the water to infiltrate from
the surface. However, considering the other statistical measures there are
remarkable overlaps between the soil classes. More homogeneous is the picture in
case of Avg GW level where even the median does not show significant difference
between classes. Simulation results of HYPRES and HUNSODA class average
parameterization show unexpected anomalies when compared with real soils.
It is to be noted that the evaluation of the results with USDA classification was
also carried out. In harmony with the findings of Wösten et al. [6] and Nemes [10]
it was concluded that even though USDA is a more detailed classification system,
it holds uncertainties at least at the same magnitude.
Figure 4: Simulated boundary water fluxes and groundwater level for an
arbitrary soil.
-600
-500
-400
-300
-200
-100
0
100
200
time [days]
-3
-2
-1
0
1
2
3
4
5
ExWat cover Precip. Infilt.
Pot. Transp Act. Transp GW level
cumulative fluxes [cm]
groundwater level [m]
Avg GW level
Ground surface
ΣTΣT
p
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
204 Sustainable Irrigation and Drainage V

Figure 5: Summarized results for the three hydrologic indicator.
4.2 3D hydrologic sensitivity analysis
WateRisk (Kozma and Koncsos [12]) is a physically based, distributed parameter,
fully coupled hydrologic model. It simulates the major processes of the hydrologic
cycle, including evapotranspiration, channel and overland flow as well as
unsaturated zone and shallow groundwater movement.
To analyze soil data uncertainty, hydrologic processes were simulated at the
Szamos-Kraszna Interfluve, a 457 km2 large sub-catchment in north-eastern
Hungary. The area had been a marshland before it was drained off to be adapted
to arable crop production. Ever since, an extremely long and costly drainage
channel network is maintained on the territory to mitigate excess water. However
the area suffers from this phenomenon almost every year. The heterogeneous but
mostly finer soil composition of the region is a defining factor in this issue.
Sensitivity analysis (SA) was carried out for the five year period between
the summers of 1995 and 2000 (average annual precipitation was 724 mm for the
period). The model was previously calibrated for recent past conditions at this area
(Jolánkai et al. [16]). 3D spatial soil data was based on the FAO method and the
HUNSODA database (Bakacsi et al. [17]). During the SA all boundary conditions
and parameters were unchanged except of porosity and specific yield values.
Porosity represents total soil storage capacity, while specific yield determines
mainly the intensity of groundwater table fluctuations under given boundary
conditions. As physically these two properties are strongly related, their
20
40
60
80
100
ΣT/ΣT
p
[%]
-200
-150
-100
-50
0
50
100
CMMFFVF
Avg GW level [m]
-16
-14
-12
-10
-8
-6
-4
-2
0
2
number
of soils
number of soils median
HUNSODA class avg HYPRES class avg
0
200
400
600
800
1000
Excess Water [days]
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
Sustainable Irrigation and Drainage V 205

simultaneous SA was preferable. The SA included a reference and two perturbed
simulations: the reference (denoted ‘mean’) represented the calibrated model setup
except of porosity and SY. For latter parameters the FAO class averages of the
HUNSODA soils were used. For the two perturbed runs (‘- sigma’ and ‘+ sigma’)
class averages were reduced or increased with class variances (Table 1).
Table 1: Porosity and SY mean and variance values of FAO soil classes.
FAO class Specific yield Porosity
Mean [-] Variance [%] Mean [-] Variance [%]
Coarse 0.215 44.1 0.419 15.7
Medium 0.104 62.0 0.458 14.1
Medium fine 0.069 74.1 0.477 9.9
Fine 0.062 69.9 0.486 9.9
Very fine 0.044 91.1 0.552 6.8
Figure 6: Water coverage time series for the Szamos-Kraszna Interfluve in
case of the three SA simulations.
Table 2: Sensitivity analysis results for aggregated hydrologic indicators.
indicator - sigma mean + sigma
ExWat. coverage
max [ha] +32 % 3596 -32 %
mean [ha] +40 % 1547 -36 %
dur. [day] +7 62 -8
ExWat. pumping [1e6 m3] +28 % 11.7 -11 %
Avr. annual ET [mm] -4 664 +12
GW depth mean [cm] -7 158 +13
var. [cm] +6 60 -7
0
1 000
2 000
3 000
4 000
5 000
6 000
7 000
8 000
2.14 3.12 4 .8 5.5 6.1 6.28
Time [m.dd]
Water coverage [ha]
Observed
+sigma
mean
-sigma
Duration
Mean
Maximum
Threshold
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
206 Sustainable Irrigation and Drainage V

Based on simulation results, Table 2 provides four aggregated indicators,
which quantify (i) the severity of the excess water (denoted ‘ExWat.’) in 2000
and (ii) the hydrologic behaviour of the area for the 5 years period. These
indicators also give information about model sensitivity to soil data. Excess water
related indicators (maximum, mean and duration of water coverage that exceeds a
certain threshold) are displayed in Figure 6.
Several conclusions can be drawn: (i) in accordance with theoretical knowledge
about hydrologic processes and experiences with the WateRisk, the partly
nonlinear model behaviour is reasonable, (ii) all excess water indicators behave as
presumed: as soils get finer, the situation becomes more severe, (iii) as expected
in case of excess water coverage, the variation of mean values exceeds the changes
of maximums, (iv) groundwater response was also reasonable: depth and
fluctuation correlated with porosity and SY.
5 Conclusions
The time-invariant analyses proved that both Ks values and the characteristic
values of WRCs show considerable variances for each soil class. Furthermore, in
harmony with the results of the 1D numerical test it was shown that in some cases
variances exceed the differences between class averages. Also, it was concluded
that class averaging may result in unrealistic parameter combination that poorly
represents soils they stand for. Moreover, hydrologic responses of class average
synthetic soils differed significantly from the majority of soils from the same class.
General consequences are: (i) texture based classification holds considerable
uncertainties concerning hydraulic behaviour and (ii) it is not obvious that
synthesized parameter combinations derived from soil databases appropriately
characterize classes when modelling hydrological processes.
A step forward in the topic could be the establishment of a new soil
classification method that focuses mainly on hydrologic characteristics of soils.
However, it is questionable that such system can be built based on only generally
available physical properties.
References
[1] McCown, R.L., Hammer, G.L., Hargreaves, J.N.G., Holzworth, D.P. &
Freebairn, D.M., APSIM: A novel software system for model development,
model testing and simulation in agricultural systems research. Agricultural
Systems, 50, pp. 255-271, 1996.
[2] van Ittersum, M.K. & Rabbinge, R., Concepts in production ecology for
analysis, quantification of agricultural input-output combinations. Field
Crops Research, 52, pp. 197-208, 1997.
[3] van Leeuwen, B., Tobak, Z. & Szatmári, J., Development of an integrated
ANN-GIS framework for inland excess water monitoring. Journal of Env.
Geogr. I, 3-4. pp. 1-6, 2008.
[4] Harter, T. & Hopmans, J.W., Role of vadose-zone flow processes in
regional-scale hydrology: review, opportunities and challenges. In: Feddes,
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
Sustainable Irrigation and Drainage V 207

R.A., de Rooij, G.H. & van Dam J.C. (eds). Unsaturated-zone Modeling,
Kluwer Academic Publishers, Netherlands, pp. 179-210, 2004.
[5] Pachepsky, Y.A., Smettem, K.R.J., Vanderborght, J., Herbst, M.,
Vereecken, H. & Wösten, J.H.M., Reality and fiction of models and data in
soil hydrology. In: Feddes, R.A., de Rooij, G.H. & van Dam J.C. (eds).
Unsaturated-zone Modeling, Kluwer Academic Publishers, Netherlands,
pp. 231-262, 2004.
[6] Wösten, J.H.M., Lilly, A., Nemes, A. & C. Le Bas, Development and use
of a database of hydraulic properties of European soils. Geoderma, 90,
pp. 169-185, 1999.
[7] Soil Survey Staff, Natural Resources Conservation Service, United States
Department of Agriculture. Web Soil Survey. Available online at
http://websoilsurvey.nrcs.usda.gov/. Accessed [03/23/2013].
[8] Kozma, Zs., Ács, T. & Koncsos. L., Hydrological modelling of the
unsaturated zone – evaluation of uncertainties related to the FAO soil
classification system. Pollack Periodica (accepted), 2013.
[9] Food and Agriculture Organisation (FAO), Guidelines for soil description.
FAO/ISRIC, Rome. 1990. (3rd Ed.)
[10] Nemes, A., Unsaturated Soil Hydraulic Database of Hungary: HUNSODA.
Agrokémia és Talajtan, 51 (1-2), pp. 17-26, 2002.
[11] United States Department of Agriculture (USDA). Soil Survey Manual. US
Dept. Agriculture Handbook No. 18. Washington, DC. 1951.
[12] Kozma, Zs. & Koncsos, L. Methodological Overview of a Coupled Water
Resources Management Model System. Proceedings of the Thirteenth
International Conference on Civil, Structural and Environmental
Engineering Computing, B.H.V. Topping and Y. Tsompanakis, (Editors),
Civil-Comp Press, Stirlingshire, United Kingdom, paper 157, 2011.
[13] van Genuchten, M.T., A closed-form equation for predicting the hydraulic
conductivity of unsaturated soils, Soil Science Society of America Journal,
Vol. 44, pp. 892-898, 1980.
[14] Simunek, J., Sejna, M. & van Genuchten, M.T., The HYDRUS-1D software
package for simulating the one-dimensional movement of water, heat and
multiple solutes in variably-saturated media: version 2.0 IGWMC-TPS-70.
International Groundwater Modeling Center, Colorado School of Miners,
Golden, 1998.
[15] Varga-Haszonits, Z. Agro-meteorológia. Pannon Agrártudományi
Egyetem, Mezőgazdaságtudományi Kar (PATE), Mosonmagyaróvár, 1997.
[16] Jolánkai, Zs., Kardos, M, Koncsos, L., Kozma, Zs. & Muzelák, B., Pilot
Area Studies in Hungary with a Novel Integrated Hydrologic Model –
WateRisk. Proceedings of the 6th IWA International Conference for Young
Water Professionals (IWA YWPC 2012), 10-13 July 2012, Budapest,
Hungary, 2012, (electronic format on CD, id: IWA-9840).
[17] Bakacsi, Zs., Pásztor L. & Szabó J., Compilation of a 3D soil physical
database for the unsaturated zone, 19th World Congress of Soil Science, Soil
Solutions for a Changing World, Brisbane, Australia, 1-6 August 2010,
Published on DVD.
WIT Transactions on Ecology and The Environment, Vol 185,
www.witpress.com, ISSN 1743-3541 (on-line)
© 2014 WIT Press
208 Sustainable Irrigation and Drainage V