This paper was peer-reviewed for scientific content. [608220]

This paper was peer-reviewed for scientific content.
Pages 1000-1007. In: D.E. Stott, R.H. Mohtar and G.C. Steinhardt (eds). 2001. Sustaining the Global Farm. Selected papers from the 10th International Soil
Conservation Organization Meeting held May 24-29, 1999 at Purdue Un iversity and the USDA-ARS National Soil Erosion Research Lab oratory.
The Usefulness of a New Model for the Gully -Control Structures Effects Prediction
Lucia Otlacan Nedelcu*
Motto: Models are to be used, bu t not to be believed. (H. Theil)

*Lucia Otlacan Nedelcu , Universi ty of Agricultural Sciences and Veterinary Me dicine, 59 MARASTI Blvd; Bucharest; Romania Zip:
71331; Fax No. 40-1-224.28.15, Sponsors: The 10th ISCO Conference Organizers, TAROM, Ro manian Flight Company. [anonimizat] ABSTRACT
Gully erosion has potential negative effects over the
gully catchments, the gully channels and/or over the
downstream confluent. The channel evolution together
with the associated landslid es can cause damages, which
were detailed in the cause-effects matrix (Fig. 6).
One of the engineering so lutions for the potential
risks diminution consists of the gully equipping with
check dams.
The paper deals, first of a ll, with the possibility of
forecasting the stable slope of the deposition profile
above gully- control structures. The prediction is based on a study of 43 cases. The research finished with a
model, which is able to anticipate the equilibrium
gradient magnitude. (Formula 16)
The author considers that the interest for the stable
slope value accuracy is justified because this gradient
serves for the “H”- height of dams- calculation. From an economic point of view, th e structure costs and the
upstream-trapped sediment volumes increase with H
3.
The equilibrium slope is also important because it
offers the possibility of assessing the size of sediment
diameters, which continue to be transported and
delivered in the emissary. This assessment is necessary in an environmental impact study and in the decision
regarding the best gully tr imming alternative choice.
INTRODUCTION
Romania is a European country lying on 237 thousand
km2. Romania has natural conditions, which determine the
intrinsic vulnerability at water erosion and landslides on 63.8
thousand km2. These degradations affect 43 percent of
Romanian agricultural surface.
The gross sediment yield pro ceeded from various sources
is 126 million tones per year. (Motoc, 1982). The delivery
ratio of sediments is 44.6 million tones per year. It is supplied in a proportion of 36 .2 percent by surface erosion
and 36.9 percent by gully erosion. In Romania, the gully
erosion is the most important component of the sediment amount discharged from the lateral small watersheds
towards the permanent river network. Gullies can be
intermediaries either between the hillslopes and the river, or sediment suppliers from own sources. There is gully erosion
in the majority of watersheds with sloped agricultural lands.
This process shapes quickly the depression relief producing damaging effects by degradations, aggradations and
generating of associated processe s. (e.g. landslides). The soil
and water users of felt both inside the small watershed, and the impact downstream. Exte rnalities can exceed, most of
times, the damages from inner watershed.
In a systematic approach, th e natural streambed of a
gully is an open, dynamic, probabilistic system of a process
– response in cascade pattern. The aggressive inputs, which
produce changes in the system, are organized both at the
level of watershed and have natural streambed.
The present work is focused on the energetic aspects of
this process for the natural system and for the anthropically
altered system, too. As concerns the natural system, the
powerful inputs during the out rush are: the velocity of the
biphasic flow, the shear stress, the power of the flow and its
loading and transport capacity. The inputs are able to induce
changes in the system. The di rection of the modifications
depends on the ratio between the available energy of the flow and the strengths opposed by the stream support. The
final tendency is to reach the state of dynamic equilibrium.
The equilibrium is settled dow n between the energy entered
the system and the energy used inside the system.
The ways for the energy consumption and dissipating
are: a mechanical work effe ctuation on the stream support
and the sediments transport on one hand and the changes
from a kind of energy to an other one on the other hand. The main means of intensification the energy dissipation are:
ensuring conditions for an in creased turbulence generation,
the growth of fluid viscosity or the change of the flow
conditions by inducing some disturbances with dissipative
effect.
The last method application will be depicted. The
engineering solution is to do some barriers and to plant
streambeds. The check dam is a perturbation for the natural
gully erosion process. It is ab le to produce significant effects
in the phase of high flood propagation along the erosion streambed and in the gully er osion mechanism. The main
goals had in view are that the dam should control the ravine
dynamics and should partly protect the downstream objectives against silting. The check dam can fulfill these
roles because it is permeable for water but it traps a part of
sediments contained in the out rush. In the upper pond of the dam a sediments deposit is materialized. The deposit
gradually reaches the quota of the weir inlet. The original
streambed thawed raising up to the new local basis of erosion represented by the weir inlet, leads to important
changes in the geometrical feat ures of the wetted section and
in the sediment delivery ratio.
I was interested in the final gradient of the deposition
profile above gully – control structures. This slope value
assures the stabilization of the banks and of the new created

longitudinal profile.
RESEARCH GOALS
1. Uncertainty management in the value prediction for
the stable gradient among the gully – control structures.
2. Usefulness of the barring effects prediction.
RESEARCH METHOD
For the first objective achievement the steps were. The
informational system creation for the data base by:
documentation and gathering of the materials existing at the designers; expeditionary prospecting of the site; field
measurements; samplings; grading analysis. Then, the
informatics system was settled up by data interpretation and
processing.
For the second objective, the actions consisted of: the
quantitative assessment of some effects; the matrix method
utilization for the impacts inventorying; social-economic
investigations accomplishment.
Romania has natural areas distinctly affected by the gully
erosion. I have placed my research disposition in the Argesel
watershed, which is representative because of the big density of the gullies, their strong activity and number of check
dams executed on each gully.
River Argesel is 65 km long and has a watershed of 250
km
2. On the 50 km distance, trav ersing an agricultural area,
the river has 80 tributaries, 69 being gullies. The total length
of the gullies is 79.6 km and their density is great, namely 2-5 km/km
2. All the 69 gullies have check dams in various
stages of sediment basins filling. Being concerned in the
equilibrium gradient upstream the dams. I identified on 30 gullies 66 dams with the above channels reached at stability.
(Fig. 1).
The those 30 gullies watersheds are of 0.25-6.74 km
2 (8
under 0.5 km2 and 3 over 5.5 km2) the altitude is between
320 m-850 m height above the Black Sea level. Other specific features of the watershe ds are: annual rainfall from
690 to 1000 mm increasing from South to North. The
torrential core on 15 minutes is in the first or middle third part in the rain duration. The main land uses are: meadows,
forests and orchards. They are not affected by surface
erosion.
RESULTS
The solutions will be presented in the same order as the
objectives were formulated.
Results in the management of uncertainty in the
stable gradient value prediction
Uncertainty is an unavoidable component of all
predictions. (De Jongh 1988). Th ere are sources of errors at
the level of input data, of prediction models and of
probability that the effects are about to be settled down. The
various types of uncertainty need to be handled in different ways in order to decrease the risk of getting some unrealistic
values.
The process of prediction has six stages. There are
specific ways of reduction the uncertainty in each of these
stages. Thus, the management of uncertainty becomes possible and the risk in prediction is reduced to an
acceptable level.

Figure 1. Argesel watershed with 30 investigated gullies.

The six stages mentioned above are:
1. Inventorying and selecting of the significant
inputs;
2. Data gathering;
3. Choice or creation of a method for obtaining
the prediction;
4. Preparation of the method for use;
5. Application of the method;
6. Presentation of the results.
I have applied all these six phases in my thesis for the
doctorial degree (Otlacan, 1989). First, I selected those
variables, which have a great participation in the
stabilization of the gully longitudinal profile. These are: the
geometrical specific features of the wetted section of the
original streambed and after the deposition; the hydraulic
conditions of the flow (peak discharge, strengths opposed by
the streambed, average veloc ity, and bottom velocity); the
specific features of the stream bed material (granulometry,
critical velocity for initiation of motion of sediments), as
well as the height of the structure.
In the second phase, of data gathering, there are the
following techniques for the uncertainty reduction: the measurement method selection for the magnitudes
previously sorted and the sampling program design in order
to achieve a required level of accuracy and significance.
Some of the variables mentioned in the first phase were
measured directly. Thus, by m eans of a topographic survey
there were obtained the specific geometrical features of the
streambed after the finish of sediment deposition. In order to
increase the reliance on these data the measurements were

repeated 2-5 times in 6 years, for the slopes upstream the
dams and for the width of the channel at the structure. The
height of the structure spillway inlet above the original
thawed was also measured.
From the initial 66 cases, 23 were eliminated because
they presented distortions due to some local factors, like:
bridges, alluvial cones of the tributaries superposed on the deposition, remainders of the construction materials, lateral
inflow, etc.
As concerns the discharges, I had at my disposal the
values calculated by the designer for the probability of 5
percent, in various original sections of streambed. I could
reconstitute some discharges, directly after the rain by
measuring the wetted section depth. The traces let by the out
rush on the vegetation, on the channel slabs, or on the spillways were used.
In order to establish the three kind of velocities the
vegetation presence or absence and the sort of vegetation were noticed. Besides, samples of sediments at the dams and
at the 30 m, 50 m and 100 m distance upstream the dams
were drawn. The sample weight was complying with Romanian standard, namely 0.1-2 kg for fine sand until
gravel. After the grading analysis it resulted the diameters
D
20, D 50, D 90.
Despite of the care for the inputs accuracy some sources
of errors still persist. They are difficult to remove having in
view the spatial-temporal variability of the magnitudes to be ascertained and the conditions that the studies should limit in
some acceptable costs. The inputs real values deviation in
comparison to those used in the models of calculation explains part of the miss of similitude between the calculated
and measured slope of equilibrium.
In the prediction process, the third phase follows in
which the uncertainty management is necessary. In this
phase it is required to choose or to create a forecast model. I
have used both ways: I made my option for some existing
models but I created a new model, too.
There is a unanimous agreement that the digital models,
which simulate the natural processes, are uncertain, first,
because they can not exactly reproduce what happens in
reality. Thus, some simplifications must be accepted. Secondly, the model limits may occur from the chosen
relationship. This type of erro r arises from the insufficient
knowledge of the process physics.
I selected 3 models among those existing in the specialty
literature for the calculation of the longitudinal profile stable
gradient. Common to the models which I dealt with is the fact that they refer to the ephemeral flows, strongly loaded
and torrential. At the present level of knowledge the specific
features of this flow regime are very poor studied. The differences among the models regard the start assumption
and the used parameters. Each author accompanies the
recommended formula by values for the factors entering in
the model structure.
Supino Model
(1965) is a black box type of model. It is
in the shape of morph metrical relations empirically
obtained. It is recommended for the Italian torrents. The stable slope S s can be calculated using the formula:
()
4/32 4/3 10/3csQnBυuS= (1)
where:υ- ratio between the mean velocity of water and the
corresponding velocity at the river bottom (nearly equal to
1.3-1.5); u c – maximum permissible velocity, depending on
the size of bed materials, at which the erosion of river bed starts (suggested by Fortier and Scobey) (m/s); B – wetted
perimeter, which can be considered equal to the width of the
river (m); n – coefficient of roughness of the river (values
extracted from Ven Te Chow) (m
-1/3/s); Q – flood discharge
according to which the torrent training is designed (m3/s).
The method applied to those 43 cases lasted for analysis
leads to the results represented in Fig. 2. The other 2 models
used in this work are based on physical laws and hydraulic theory applied to the sediments movement mechanism.

Line of equal
values G. Supino Model
Figure 2. Comparison of measured (x I) and computed (y I)
slopes. Regression coefficient r yx = 0.005.

Brahms Model (1753) attributed by Lelliavsky (1955).
In this case the condition for the longitudinal profile stability
is theoretically expressed by the equilibrium of the forces
acting on a grain in bed of sloping stream. (Formula 2).
()()αcosfPFαsinPF P y y x −≤ −+ (2)
where: P x – the drag force parallel to the bed; F – the
submerged weight of the particle with diameter D; P y – the
lift force; α – the original slope angle of the bed; f – the
friction factor.
Assuming that
y x22PP0,5;f0;αtg1; cosα S; tgα sinα ====≅≅
and after some simplifications, the final expression for the
stable slope S s is:

2
d2p2o2o2s
DBDB
S1
0.4S12S141S



∗−−+= (3)
where: B o, B p – are the top width in the dam location (m),
the index “o” represents the orig inal situation and the present
value (after the deposition) is marked by the index “p”; D o –
the sediment diameter most often arise in the original
situation (before the gully trimming); D d – the desirable
diameter, adopted so that downstream damages should
decrease. The basin is capable of trapping only the
sediments with diameter D d or bigger. I considered D o=D 50
and D d=D 20. The calculated values S s are represented in the
Fig. 3 in comparison with the measured values.

Line of equal
values A. Brahms Model
Figure 3. Comparison of measured (x I) and computed (y I)
slopes. Regression coefficient r yx=0.168

Baloiu Model (1972) V. Baloiu is a Romanian professor.
He starts from the theoretical hypothesis that the
hydrodynamic equilibrium between the flow and the stream
is provided with an out rush velocity equal to the critical velocity corresponding to the material from which the
stream is consisting of. It results the formula (4) for the
stable slope S
s:

RCvS22cs= (4)
where v c is the critical velocity calculated in accordance to a
Russian recommendation (v c=v0.1h0.2), C is the resistance
coefficient computed with the Bazin’s formula:
RγR87C + =
where R is the hydraulic ra dius corresponding to the
discharge having a return period of 2 years.
In the results represented in Fig. 4 the values for the
stable slopes were calculated with the maximum permissible velocity for the diameter D
50. VV.
V. Baloiu Model
Line of equal
values

Figure 4. Comparis on of measured (x I) and computed (y I)
slopes. Regression coefficient r yx = 0.213.

As concerns the model obtained as a result of my researches
developed between 1982 ÷ 1988, it started with an American
experiment performed by D.A. Woolhiser and A.T. Lenz
(1965). The theoretic support is a deterministic model, based
on fundamental physics law, namely the principle of continuity as applied to the sediment carried by a stream and
the sediment forming the bed and banks.

()[]() bgBg Bfdtd
s−= (5)
in which f(B) is an arbitrary function describing the channel
boundary; g(B) is the sediment discharge capacity of the
channel expressed as a function of the channel boundary configuration, and the product “g
sb” is the rate of sediment
supply to the channel section under consideration; b is the
channel width.
()[] 0 Bfdtd= (6)
and
()Bgbgs= (7)
for channels in equilibrium.
The Du Boys formula was selected for the sediment
discharge capacity. It is convenient for its simplicity.
()()c ooδδψbδ Bg − = (8)
in which: ψ is an experimentally determined coefficient; δo
is the bed shear stress; δc – critical bed shear stress.
Utilizing the relationship
3/5
1/2ee obsQnR;γRSδ



==
in the equation (8) and substituting the result into equation
(7), it can be written:

0.7e0.6
c0.7e0.6
s s SbQnγδ SbQnγψb Gbg 






−

== (9)
Equation (9) is the theoretic support for the stable slope
prediction model substantiation. The method for the
equation solution consists of its simplification and in the
pertinent variables identification. The values of these variables are going to be determined by field measurement.
Assuming that δ
o – δc ≅ δo and S e≅ S (S e – energy slope; S –
channel slope), the slopes ratio in terms of present and
original values of the pertinent variables is:
1/7
op6/7
po6/7
po5/7
po5/7
sosp
op
bb
nn
QQ
ψψ
GG
SS



















= (10)
Two comments are necessary. The first one is that the
product of the first four ratios on the right side may be
considered fairly constant. The second observation refers to
the fact that under equilibrium, the average original width b o
could well be correlated with the original slope S o. As
concerns the width b p, it might be related to b a (the width of
the channel at the dam) and to H (the height of the inlet above the original ch annel quota). For example, considering
a trapezoidal section,

ZHb b a p−=
where Z is the side-slope. It results that:
bp = f(b a, H) (11)
Incorporating the above observations in the Equation 10
we get the following expression:
()[]2 1 bab
o p H,bf CS S= (12)
where C is a coefficient and b 1 and b 2 are arbitrary
exponents.
In a first approximation, a linear regression model may
be used for the function f(b a, H) explicitness:
Hb bbaSS
3 a2 1op++= (13)
where b a and H are the independent variables.
Then, a new multiple not linear regression model may be
used if some deviations between the computed and measured slopes are found.
() 5bH3bab2b1a4b
oS1apS ++ = (14)
In the preparation phase for the model utilization, the
software for the statistical techniques was applied to the same 43 input data used in the other models. The results are:

0.113H 0.0017b 0.497SS
aos − += (15)
and
() 0.113H 0.0017b 0.497 0.663SS a0.89o s − + = (16)
In the last phase of the uncertainty management I
answered to the exigency of presenting the results and of
deciding which model should be recommended. In order to
have some objective arguments, the 4 models for the stable slope simulation were statistically compared. The statistical
analysis allows the comparison between the observed set of
values x i and the homologous computed set of values y i.
This analysis leads to the conclusion that the fourth method adjusts the best value of the equilibrium slope. Having a
correlation coefficient r
yx closer to 1.0, this model reflects a
stronger relation between the two sets of variables.
In order to recommend the 4th method for the design
activity the statistical test was continued. The result was that
ryx is significant even at the 0.999 confidence level. The
plotting of the confidence interval bounds for the regression
line proved that all the couples of values (y i, xi) are included
in the interval for the 0. 99 assurances. (Fig. 5).
The recommendation of using the 4th model relies, also,
on the conclusion of other trial. The method was applied to 25 data of South Western Wisconsin (N.E. Minshall, 1953),
which did not participate in the model calibration. The
correlation coefficient is r
yx=0.915, though.
The explanation for the multiple regression method
ability to anticipate a more correct value can be found, first,
in the fact that it operates with directly measurable sizes. Thus, the accuracy of the inpu t data may be guaranteed.
Secondly, having in view that the studied events are
aleatory, it is suitable to promote a theoretical substantiation, which should lead to the pertinent variables selection. A
statistical processing must be applied to these variables.
The validity field of the model is: the original thawed
slope up to 15 percent; dams from concrete or rubble
masonry with a height less than 3.5 m; non-cohesive
streambed material of D
90=0.2 ÷ 20.0 mm; values of the ratio
stable slope/original slope S s/So=0.05 ÷ 0.6. Besides, the
model is applicable in watersheds comparable to Argesel
watershed with respect to climate, topography, soils and
vegetal cover.

Figure 5. Comparison of measured (x I) and computed (y I)
slopes. Regression coefficient r yx = 0.770.

Results in the assessment of the usefulness of the
barring effects prediction
The equilibrium slope value is a resultant, which
integrates the consequences of the gully erosion process
anthropically modified. This final gradient of the sediments deposit serves to the gully control structures dimensioning.
Thus, the structure height H depends on the stable slope S
s
computed value:
H = L(S o – S s) (17)
where: L – equipped sector length (m); S o – original slope
(m/m).
From an economic point of view, the structure costs and
the upstream-trapped sedime nts volumes increase with H3.
Therefore the stable slope prediction is a useful tool needed
in the economic analysis fo r choosing the best gully-
trimming variant.
The works reliability is a technical aspect influenced by
the concordance between the computed slopes and those
produced in fact. In the ca se of the structure location
according to their mutual suppo rting principle, the final
slope underestimation leads to the partial silting of the
upstream work. Thus, the investments become unjustified great. On the contrary, if the an ticipated grade is steeper than
the actual one, a part of the stream remains unprotected and
erosions are developing. The upstream dam undermining increases the expenditures fo r the maintenance works.
From a power efficiency point of view, the gully control
structures create a new stream with a bigger hydraulic radius than in natural conditions. The longitudinal profile gradient
reduction counter balances (compensates) this increase and
the final velocity v
f is smaller than the initial velocity v i.
There is a significant reductio n of the kinetic component
of flow energy. The values of the ratios 2
f2i/vv is of 1.8 ÷ 9.
In 56.8 percent of cases the decrease is of 2 ÷ 3.5 times. This
effect may be quantified by a correct prediction of the
equilibrium slope. The assessment of the size of diameters, which continue to be transported and delivered in the
emissary, becomes possible in this way.
The gully trimming opportunity must be judged from
more points of view than those already mentioned. In the
decision taking it is important to establish what damages are
avoided by intervention with works in comparison with the no-intervention alternative. I tried to complete the inventory
of the potential harms generated by the gully erosion in the
case of no intervention. The first step was to highlight the
gully specific features. Some of them are as follows:
1. Often, gullies have neither major streambed nor
meanders, so they dispose of a greater specific energy
than the rivers in the same zone.
2. The watershed is smaller than 10 km
2, therefore in the
discharge value calculation the rain intensity may be
considered uniform and the factors, which
individualize the catchments area, must be taken into account.
3. In majority of cases the runoff is ephemeral or
intermittent. This makes suitable the formative discharge calculation with the 50 percent insurance.
Thus, it complies with the concept of effectiveness of a modeling force.
4. The out rushes are torrential, having a sudden variation
of the discharge in time and a heavy load of solid
material.
5. The streambed dynamics happens either during the out
rushes or between them. The active factor of the flood
is the liquid and solid discharge variation.
Based on the enumerated spec ific elements I drew up the
cause-effects matrix. (Fig. 6). It is obviously that the gully
erosion induces numerous negative effects on the biotope, biocoenosis, human settlements, people welfare and on the
landscape. Thus, this natural process introduces restrictions
in the soil and water resources utilization.
The recently appeared ecological economics is interested
in some additional indicators in comparison with the traditional economics. They are: the damages produced by
the externalities; the tota l economic value of the
environmental goods and services; specific valuation techniques of the projects fo r the environment protection.
The works effects pred iction is very useful in the damages
assessment.
In a case study damages values produced downstream
the watershed were 2.45 times bigger than those suffered
inside the watershed. In order to obtain the total economic value of soil and water in a watershed I started from this
value definition:
Total economic value = (18)
Actual use value + Existence value
The first term may be computed by cost-benefit method.
The technique used for the other 2 terms was the contingent valuation. This method uses as tool the socioeconomic
investigation. A team of professors and students performed
such a study in 3 rural settlements in Argesel watershed. We
had in view to test the opportunities of maintaining the
population in these places and of encouraging new landowners to come from their residential towns to the
localities where they have just received a plot of land. From
the point of view of eroded soils management, it was a practical work for facilitati ng the communication among the
groups involved in this field. (L. Nedelcu, R. Sofronie,
1998).
The questionnaire contained questions by which we
checked: the level of awareness about the quantitative and
qualitative deterioration of the resources in the watershed; the degree of the community knowledge about the watershed
working like a system. In this system, any change is felt by
all the categories of public. Also , we wanted to find out the
farmers opinions about the ways for their benefits and yields
increasing and about their w illingness to pay in order to
improve the soil and water quality.
The sample included about 10 percent of husbandries. It
resulted a relative low perception about the damages
generated by the erosion. Thus, only 20-33.3 percent of
those interviewed think that the anti-erosion works are
necessary for the yields and benefits increasing. On the other hand, they would consent in 85-95 percent to contribute by
means of work and money in order to improving the soil and
water quality.
In our opinion, there are potential conflicts between

Figure 6. Cause – effect matrix.

Stocking on
slopes and
downstream
Stocking in and
downstream
gullies
Detachment by
slides
Detachment by
gully erosion
A 1 Physical changes of soil characteristics † † †
2 Increase of slope dryness †
3 Soil settling on slopes †
4 Marshes formation on slopes (swamps) † N

5 Groundwater hydrological regime modification † † † †
6 Torrentiality increase † † † † E
7 River bed morphometrical dynamics † †
8 Lakes and bottom lands silting † † G
9 River turbidity increase † †
10 Earth masses redi stribution on slopes † † A
11 Definitive ground loss † † † †
12 Increase of dust particles concentration † † † † T
13 Land slides starting † † †
14 Slope breaking up † I
15 Soil fertility decrease † † †
16 Oxygen concentration decrease in river water † † V
77 Increase of CO concentration in air † †
18 Increase of ground water mineralization † † E
19 Soil pollution †
20 Water pollution † †
21 Mobilization of the previous alluvial stocks † † † † E
B 22 Habitat modification † †
23 Embarrassing of the flora and fauna development † † † † F
24 Human population distress † † † †
C 25 Pedogenesis inhibition † † † F
26 Trophic chains unbalancing †
27 Biocoenoses modification † † E
28 Starting of water anaerobiosis and deterioration † †
D 29 Water inlets and land reclamation works silting † † C
30 Flood risk amplification † †
31 Decrease of the soil and water resources productivity † † T
32 Landscape degradation † † † †
33 Lines of communication depreciation † † † † S
34 Damages in settlements † † † †
35 Hindering of the tillages mechanized execution † † † †
36 Water storage dams damaging † † †

1 Increase of groundwater storing capacity z POSITIVE
2 Soil fertilizers spreading z EFFECTS
3 Enlivening of the supply with construction materials z z z

LEGEND
Impact on:
A Physical-chemical specific features
B Biological conditions
C Ecological reactions
D Living standard

farmers located in the same watershed or between farmers
and other economic activities; between sustainable
development and short-term interests. We are sure that we
must do a lobby for the anti-erosion works.
CONCLUSIONS
The research work presented in the paper had in view
two main objectives:
1. Uncertainty management in the value prediction for the
stable gradient among the gully – control structures.
2. Usefulness of the barring effects prediction.
In order to answer to the first goal, the author applied six
steps of a forecast process recommended by P.E. De Jongh
(1988). A model that allows the calculation of the final gradient above the gully- control structures has been
obtained. In order to have some objective arguments for this
model use, it was statistically compared with 3 other models known in the scientific literature.
As concerns the second obj ective of the research, the
usefulness of a tool for the final gradient of the sediment deposits was highlighted.
It has already been known that this slope serves to the
calculation of the check dams height (Formula 17). Besides, the author considers that once the final slope of the new
thawed being calculated, the level of the gully erosion
damages reduction can be anticipated. In this way, it is possible to decide on the gully trimming opportunity and the
chance the project is accepted by beneficiaries might
increase.
Using investigation methods specific to the ecological
economics, the author tested the level of the awareness about
the gully erosion damages and the farmers’ opinions about
the gully trimming project opportunity. It resulted in a
relative low perception about the relation between the farmers’ benefits and the rate of the gully erosion process.
ACKNOWLEDGMENT
The present paper is based on two researches. My first
work was the thesis for the doctorial degree, conducted by
the professor in hydraulics S. Hancu and the academic
professor in erosion control M. Motoc. I am grateful for their professionalism and exigency. The second research work
was accomplished by a Romanian -British team (1995-1999).
The main aspects had been the environmental and the socioeconomic issues of the eroded soils. I led the Romanian
team, which rejoiced of the support of the British Council
and the Romanian Ministry of Research and Technology. I could initiate and direct this project due to the training
courses organized in Romania by the University of
Minnesota, the National Polytechnic Institute of France and the World Bank. Regarding the financial support, I am
grateful to the mentioned sponsors, which offered me the
opportunity of attending this conference.
A thought to my professor D. Barbulescu that convinced
me to learn English and to my understanding husband. REFERENCES
Baloiu, V. 1972. Researches a bout the technical efficiency
of the anti-erosional works in the Chineja watershed,
Galati county – Thesis for the doctorial degree, Brasov,
Romania.
Chow, V.T. 1959. Open Channel Hydraulics, Mc.Graw Hill
Book Comp, New York
Fortier, S. and F.C. Scobey. 1926. Permissible Canal
Velocities, Transactions ASCE vol 89
De Jongh, P.E. 1988. Uncertainty in Environmental Impact
Assessment – Chapter II.4 in Environmental Impact
Assessment, Theory and Practice edited by P. Wathern,
Library of Congress.
Lelliavsky, S. 1955. An introduction to fluvial hydraulics,
London.
Minshall, N.E. 1953. Evaluation of gully –control structures
in Southwestern Wisconsin, U.S. Dept. of Agric., Soil
Conservation Service, Washington D.C., September.
Motoc, M. 1982. The average annual rate of erosional
degradation of the territory of Romania – Bulletin of
Academy of Agricultural and Forestry Sciences.
Nedelcu, L. and R. Sofronie. 1998. Setting up of the
Environmental Engineering Department. The catalytic
actions of this Department in the development of co-
operation in the eroded soils management in Romania – International Workshop on Community – Based Natural
Resource Management Economic Development Institute
of the World Bank, Washington D.C.
Otlacan, L. 1989. Methods and works for the concentrated
flow energy dissipation on gullies – Thesis for the
doctorial degree, Bucharest, Romania.
Supino, G. 1965. Le reti idrauliche – Patron, Bologna.
Woolhiser, D.A. and A.T. Lenz. 1965. Channel gradients
above gully-control structures – Journal of the
Hydraulics Division, Proceedings of the American
Society of Civil Engineers, Vol. 91, No. HY 3, May.