MINISTRY OF EDUCATION MINISTRY OF AGRICULTURE
AND TRAINING AND RURUAL DEVELOPMENT
THUYLOI UNIVERSITY
DAO TAN QUY
BUILDING A NUMERICAL MODEL SIMULATING WATER FLOW AND SEDIMENT TRANSPORT IN SMALL
AND MEDIUM WATERSHEDS Specialization: Hydrology
Code no: 62 44 90 01
SUMMARY OF DOCTORAL THESIS HA NOI, 2017
This scientific work has been accomplished at: ThuyLoi University
Advisor 1: Assoc.Prof. Dr. Pham Thi Lan Huong Advisor 2: Assoc.Prof. Dr. Ngo Le Long
Review No.1: Assoc.Prof. Dr. Nguyen Ba Quy
Review No.2: Dr. Nguyen Lap Dan
Review No.3: Assoc.Prof. Dr. Nguyen Hoang Son
This Doctoral Thesis will be defended at the meeting of the University Doctoral Committee at: ................................................................................................................................ ................................................................................................................................ at …………..on……………
This thesis is available at:
- The National Library
- The Library of Thuyloi University
INTRODUCTION
1. Problems statement In recent years, under the negative impact of natural factors as well as the global climate change have caused erosion, sediment transport and land degradation in watersheds, especially inhillslopes. Vietnam is situated in the region of tropical monsoon climate with mountains account forabout 3/4 of the total area, hence erosion is considered as a major threat to the earthen slopes in Vietnam.If we do not have measures to prevent erosion, hundreds of tons of soil and nutrients are lost every year leading tolands become degraded and can no longer cultivated. Therefore, the “building numerical model to simulate the flow and sediment transport on small and medium watershed” is necessary and urgent to apply in calculating the flow and sediment transport on Vietnam’s watersheds.
2. Objectives - Research on the scientific basis to a construct numerical model to simulate the flow and sediment transport on watershed. - Application of the built numerical model in some small and medium watersheds.
3. Objects and Scope - Objects of the study: The numerical models simulating flow and sediment transport. - Scope of the study: Small and medium watersheds.
4. Study contents - Provide an overview of the numerical models simulating erosion and sediment transport in the medium and small watersheds in the world as well as in Vietnam, then evaluate the technical limitation and point out the issues that the thesis should concentrate. - Apply the theoretical basis of the mechanism of erosion and sediment transport to develop anumericalmodel simulating erosion and sediment transport in small and medium watersheds. - Apply the developed numerical model to some small and medium watersheds
1
in Son La province.
5. Methodology - Literature review; - Studying on the basic theory and inheriting other related studies; - Remote sensing and Geographic Information System (GIS); - Mathematical Modelling; - The study results were reported and discussed in many conferences.
6. Implications Scientific significance: The research results of this thesis will reconfirm that buildingnumerical models simulating sediment transport applying remote sensing and GIS is very effective and necessary in the present period. Practical significance: The results of this thesis will provide local data sources and tools that can monitor, evaluate, look up information and monitoring the impact of erosion and sediment transport to the activities of exploitation and using land and water, thenprovideperfectly reasonable solutionsfor water resource planning, and for land use planning.
7. New contributions - Construction of a new numerical model simulating the flow and sediment transport on small and medium watershedsthat usesLax-Friedrich Scheme and addsmore elements of time and space to it to solve flow and sediment transport equation. - Construction of the correlation equationbetweenthe resultsof inter-rill erosion and rill erosion that can predictthe amount of sediment eroded and transported in watershedsbased on the intensity of rainfall.
8. Contents: The thesis consists of 3 main chapters Chapter I:Overview of models simulating sediment transport in small and medium watershed. Chapter II:The scientific basis to develop a numerical model simulating sediment transport in small and medium watersheds. Chapter III: Application of the developed numerical model to simulate flow and sediment transport in several small and medium watersheds.
2
CHAPTER 1 OVERVIEW OF MODELS SIMULATING SEDIMENT TRANSPORT IN SMALL AND MEDIUM WATERSHED 1.1 Overview of erosion and sediment transport inwatershed
1.1.1 Definitions and basic concepts
1.1.1.1 Watershed erosion Erosion is a phenomenon where piece elements, clods, and some time the whole land surface to be eroded and transported away by the wind and water power. Water erosion depends on the energy of the water flow and the resistance to erosion of the soil where water flows.
1.1.1.2 Sediment and deposition Suspended sediment is small-sized sediment particles floatingand driftingdown in the water. The speed of it equals the speed of water flow.
Deposition is a process where the soil particles detached by erosion then
deposited in the ground or inside the water such as lakes, streams and wetlands.
1.1.2 The main causes of erosion and factors affecting erosion
1.1.2.1 The main causes of erosion
a. Group of rainfall factors: Rainfall, rainfall intensity and distribution will decide to the forces dispersing the particles of soil, to the amount of water and to the velocity of runoff. Rainfall in a short time will limit erosion due to the insufficiency of water to form flow. When it rains with a greatintensity in a long time, the very serious erosion will occur.
b. Group of soil mechanical component factors: For soil with heavy mechanical components, the particles of soil are always small, smooth, cohesive and hard to break up, thus the risk of erosion is low. For soil with medium mechanical components, the particles of soil have moderate size and links, and to be porous, hence it is easy to be eroded by surface runoff. Therefore, the risk of erosion is high. For soil with light mechanical components, although having unstable structure but it has a large particle size that is difficult to transport, thus the erosion risk is not high. This soil has good permeability but bad water retention. 1.1.2.2 The factors affecting erosion There are 5 main factors affecting soil erosion include topography, soil type,
3
vegetation, climate and humanity.
1.1.3 Classification of watershed erosion Splash erosion, sheet erosion, rill erosion, gully erosion.
1.1.4 Sediment transport in the watershed Transport is the washing and movement of the soild particles from high lands through rivers and ultimately to the ocean. Sediment transport process in the watershed is complex that depends on the amount of flow, erosion, transport and deposition. 1.2 Studies in the world
1.2.1 Studies on overall assessment of erosion Current erosion process is associated with agricultural activities. Many have said that the land was exploited to become exhausted could be the reason causing past civilizations lost. Therefore, together with land degradation, erosion exists as a problem throughout the development process of humanity.
1.2.2 Studies on models simulating the process of erosion and sediment
transport
1.2.2.1 The empirical model The empirical model is primarily based on the analysis of observations and relationships from measured data. The value of parameters in the empirical model identified through the model calibration process, but usually determined from the verification of the actual monitoring data. The empirical model is not to mention the depositionissues in the watershed and not to calculate a specific rainfall, and also do not consider the erosion in rills, channels and see the depth of the overland flow to be constant.
1.2.2.2 Physically based model Unlike the empirical model, the physically based modelwas developed based on the understanding of the laws of motion and the physical mechanism of the process of erosion, it means that this model based on the understandinghas been theorized as the laws of physics or equation. The physical processes of erosion, including soil particle detachment, transferring and sediment transport. The physically based model simulating erosion and sediment transport is constructed based on the mathematical equationsdescribing the physical phenomena of soil erosion process. The mathematical basis of the model is the
4
continuity equation of Bennett. Continuity equation is commonly used in modeling the dynamics of soil erosion as follows:
(1-11)
1.2.3 Algorithmsin watershed scale erosion and sediment transport models The watershed scale erosion and sediment transport models are based on the continuity equation of Bennet. These are solved by numerical methods such as CREAM, EUROSEMmodels, while othersto be solved by analytical methods such as SWAT, WEPP modelsetc.In a case the models solved by analytical methods, the slopes must be uniform and have rainfall excess. Ina general case the problem is usually solved by numerical methods, using the finite element schemes and a specific grid of time and space. 1.3 Studies in Vietnam
1.3.1 Studies on overall assessment of erosion Vietnam is located in the region of tropical monsoon climate with relatively heavy rainfall (average from 1800 - 2000 mm), but unevenly distributed and concentrated mainly in the rainy season. Heavy rainfall concentrates water into flow ofa great intensity, thus this is a major cause of soil erosion in Vietnam.
1.3.2 Studiesand applications of erosion and sediment transport models The studieson factors causing erosion and on the predictabilityof erosion on hillslopes gave the erosion indicators of rainfall and the erosion coefficient of soil(K) to proceed to apply the universal soil loss equation (USLE) by Wischmeier and Smith, and to predict soil erosion and to preliminary showerosion potential by rainfall on small scale maps. However, these studies did not consider other factors causing erosion such as slope, slope length, diversified cropping systems, soil protection factor. Results of these are still at the level of a forecast of generalized erosion. Besides, the previous studies only used the universal soil loss equation combined with theRemote sensing and Geographic Information System (GIS) to assess the soil erosion potential and to analyse the impacts on agricultural production. This wayresulted ingreat efficiency, helped to save time and costs. Study results in the form of GIS data,so they are very intuitive, easy to use and calibrate. However, the K and Pcoefficients were taken in the reference, hence they would partly affect calculation results.
5
1.4 Gaps in the studies of erosion and sediment transport - Research
orientation in the study
1.4.1 Gaps in the studies of erosion and sediment transport
The researches in Vietnam is mainly apply available mathematical models of
erosion and sediment transport.
Foreign models are usually commercial and costlyones with closed source, thus
applying to river watersheds in Vietnam will not avoid certain errors for the
parameters in the model calibrated according to abroad data.
1.4.2 Research orientation in the study
Research on the theoretical basis of modeling flow, erosion and sediment
transport in the medium and small watershed. Then construction of the diagram
of forming flow and sediment transport, and the diagram simulating process
flow of sediment in the watershed.
The study suggests new stable algorithms with the convergence for continuity
flow equation, momentum equation and erosion and sediment transport
equation. Conclusions of chapter 1:
Author provides an overview of simulation models of sediment transport in the
watershed in Vietnam and worldwide, then points out that in Vietnam, we
mainly use available mathematical models of erosion and sediment transport
which are commercial with high cost and closed source. Moreover, the
parameters of these are constructed based on the physical characteristics of
abroadwatersheds, so when applying to the watersheds in Vietnam will lead to
certain errors. Therefore, to overcome the limitations of the models mentioned
above, author will concentrateon constructing a new model simulating sediment
transport in Vietnam’s small and medium watersheds.
6
CHAPTER 2 THE SCIENTIFIC BASIS TO DEVELOP A NUMERICAL MODEL SIMULATING SEDIMENT TRANSPORT IN SMALL AND MEDIUM WATERSHEDS 2.1 Theoretical foundations – Research and propose algorithm 2.1.1 Theoretical foundations 2.1.1.1 Equation for infiltration
The infiltration process is calculated from a equation calculating losses of
infiltration during rainfall. The Green - Ampt Mein - Larson equation is written
as follows:
(2-1)
2.1.1.2 Equations for Flow
a. Equations for Overland Flow
The continuity equation: (2-2)
The momentum equation: (2-3)
where u = αhm-1. (2-5)
Substituting equation(2-5)into equation(2-2)result is:
` (2-6)
b. Equations for channel/river flow
The continuity equation: (2-10)
The momentum equation: (2-11)
2.1.1.3 Equations for simulation of flow
a. Equations for erosion and sediment transport in watershed
Model simulating erosion and sediment transport is constructed based on
mathematical equations describing the physical phenomena of erosion and
sediment transport. The mathematical basis of physical phenomena isthe
continuity equation of Bennett. The continuity equation for erosion and
7
sediment transport is written as follows:
(2-14)
b. Equations for erosion and sediment transport inchannel/river:
(2-17)
(2-18) (2-19) Boundary conditions: C(0,t) = C0(t) t ≥ 0 Initial conditions: C(x,0) = 0 x ≥ 0 2.1.2 Research and propose algorithm
Currently, there are many methods to solve equations for flow and equations for
erosion and sediment transport. In this study,the thesis uses Lax-Friedrichs
scheme adding time and space components and it called finite difference Lax-
Friedrichs-Weightscheme (LFW) to solve a system of equations for flow and
equations for erosion and sediment transport. The algorithm was proved by
giving the evidence of convergence for the above equations.
2.1.2.1 The algorithm for solving equations for overland flow
a.The algorithm: With xj = jΔx, tn = nΔt, the Lax-Friedrichs scheme (LF)
approximates to the derivative of function u = u(x,t)at the point (xj, tn):
(2-24)
The finite difference Lax-Friedrichs-Weightscheme (LFW) is written as
follows:
8
Applying LFW for equation (2-6), then we get the differential equation of (2-6) as follows:
(2-25)
b.The convergence of algorithm
Study on the stability of LFW difference algorithm for equations (2-6)
corresponds with the stability of LFW difference algorithm for the following
equation:
(2-26)
Where with is constant.
2.1.2.2 The algorithm for solving equations for channel/river flow:
(2-35)
Approximate solution to A, Q at followingthe LFW scheme, but we
have only to solve A by given equation (2-35). Then we have:
(2-41)
2.1.2.3 The algorithm for solving equations for erosion and sediment transport in watershed
a.The algorithm: There are many methods of solving equation (2-14), but
to limit the errors, many used the implicit or explicit finite difference schemes.
The author used the Lax - Friedrichsscheme adding more time and space
components to approximate the derivative of functionu = u(x,t) at a point (xj,
9
tn) with xj = jΔx, tn = nΔt. Then we have:
(2-44)
b.The convergence of algorithm:
The LFW algorithm is stable for equation (2-14), it is easy to inspect the
stability of the algorithm by the way similar to the steps in the process of
checking the stability condition of the LFW difference algorithm to equation (2-
26).
2.1.2.4 The algorithm for solving equations for erosion and sediment transport
inchannel/river
The LFW scheme is applied to erosion and sediment transport equation in
channel/river:
(2-60)
Then we find out from and the approximate values of
:
(2-62)
2.2 Construction of the model components
Based on the theoretical analysis of the formation of flow and sediment
transport in the watershed. Thethesis constructsa diagram describing the flow of
sediment in the river basin as Figure 2-3.
10
Figure 2-3 Diagram of calculation of erosion and sediment transport
11
2.3.1 Inter-rill process
2.3.1.1 The potential of inter-rill erosion
Calculation of erosion and sediment transport in the watershed is very
complicated, thus to simplify it the author chose the following equation:
(2-66) Eil = ail.iket
2.3.1.2 Inter-rill sediment transport
In this model, the author use the Wischmeier - Meyer equations to define inter-
5/3.q5/3.
rill sediment transport capability as follows:
(2-70) Tcl = acl.Io
2.3.2 The process of rill erosion
2.3.2.1 Rill erosion
There are many formulas to calculate rill erosion but they are still too complex
and depends on many factors, therefore theauthordecided to select a equation
having easierrelationships between factors to applyfor the model as follows:
(2-74) Ei = Ciqs KCslr.
2.3.2.2 Rill sediment transport
There are many equations were built to calculate sediment transport capacity,
but the applicability of each equation suit each model. The author chose the
following equation to calculate:
(2-75) Tc = 3600.ac.q.ρs
2.3.3 Processes in channel/river
2.3.3.1 Sediment transport capacityin channel/river
The sediment transport capacity is calculated by the Parsons et al. equation as
follows:
. (2-76)
2.3.3.2 Sediment transport in channel/river
Sediment transport in river is modeled by the Engelund - Hansen equation as
follows:
. (2-77)
12
2.4 Analysis and selection of programming language to construct model C++ programming language is a high-level programming language, the library contains a lot of available functions. Programmers can take advantage of these functions to solve the problem without creating new ones. Moreover, C++ supports many suitable operations to solvetechnical problems with complex formulas. In addition, C++ also allows the programmer to define themselves more abstract data types. Therefore, the C++ programming language is selected for construction of model simulating flow and sediment transport on small
and medium watershed.
2.5 The structure and function of several sub-programes
2.5.1 Structure and key modules
2.5.1.1 The program has two basic components
a. Hydrological component: Processing data rainfall input calculating
precipitation through the Green-Ampt-Lason equation (1973).
b. Erosion and deposition components: Calculating the amount of sediment generated by the effects of rain and flow; Calculating the amount of silt in the watershed and in the channel/river through the sediment continuity equation with the LFW finite difference scheme.
2.5.1.2 Key modules
a. Module calculating inter-rill erosion and sediment transport
b. Module calculating rill erosion and sediment transport.
c. Module calculating erosion and sediment transport from channel/river
to the outlet of watershed.
2.5.2 Functions of several sub-programes To record information and place it into the temporarily store of computer Solving nonlinear implicit equations by using iterative Newton-Rapson scheme. To calculate the amount of sediment generated under the action of rain and sliding shear stress from overland flow; and calculate the amount of silt in the sub-watershed usingLFW finite difference scheme with sediment continuity equation.
13
To calculate the amount of sediment generated by shear stress of the concentrated flow, calculate the amount of silt in the channel/riverusingLFW finite difference scheme with sediment continuity equation.
2.6 User interface
Figure 2-7. Start screen in program
Figure 2-8. Data input interface
Conclusion of Chapter 2
In this chapter,the authoruses the Lax-Friedrichs scheme adding more time and
space components (LFW) to solve flow and sediment transport equations in the
watershed. The author also proves the convergence of the algorithm for the
above equations. Based on this basis, the thesis uses language C++ to construct a new simulation model of flow and sediment transport on small and mediumwatersheds.
14
CHAPTER 3 APPLICATION OF THE DEVELOPED NUMERICAL MODEL TO SIMULATE FLOW AND SEDIMENT TRANSPORT IN SEVERAL SMALL AND MEDIUM WATERSHEDS
3.1 Input data
3.1.1 Data processing
Geographic information system (GIS) will digitizemaps of soil, vegetation,
topography and river systems. Modeling the watershed altitude based on
topographic maps.
Dividing theoriginal watershed into new sub-watersheds based on the division
and overlay of mapsaccording to the consensus on terrain, direction of flow,
soil and vegetation characteristics of the watershed,then determining the
direction of flow in the sub-watersheds.
Generating the files of rainfall data, measured discharge, measured suspended
sediment discharge and the parameters of the model.
3.1.2 Running the model
After entering data into the model andrunning it, we will get the flow and
sediment transport simulation results in watershed in the form of graphs.
Based on the given results, adjusting the parameters of the model according to
measured data. Finally, we have the parameters of the model.
3.2 Application of the developed model to simulate erosion and sediment
transport
3.2.1 Nam Sap watershed
3.2.1.1 Geography Nam Sap stream watershed is located in Moc Chau, Yen Chau, Bac Yen, Mai Son, and Van Ho districts of Son La province and occupies an area of natural 1,085.52km2, which is about 7.66% of the total area of the province. Located in the geographic scope: 104o11'09’’– 104o42'54 longitude, 20o42'8 – 21o10'15 latitude. It borders the Nam Pan stream watershed to the West,Van Ho district to the East, and Laos to the South.
15
3.2.1.2 Topography Lying on 6 National Highway (Hanoi - Son La - Dien Bien)about 40,5km of Son La city, the Nam Sap is an inlandwatershed, the terrain is characterized by complication, being fragmented, steep slope and alternating between mountains 3.2.1.3 Soil characteristics Nam Sap watershed has 4 main groups of soil, includingyellow brown soil, feralit gray soil, grey loam soil on mountain and rocky mountains. 3.2.1.4 Features of vegetation cover and land use status There are many types of soil in Nam Sap watershed suitable for crops. Area of forest is 27,573.5ha, accounting for 25.48% of the entire region, of whichrich natural forest area is about 8,847.06ha, poor natural forest area is 11,752.06ha. The industrial crop land accounts for 0.93% and the agriculture land accounts for 29.86%. 3.2.1.5 Climate features Nam Sap watershed has a tropical climate, characterized by the general climate of the Northwest: Winter is cold, summer is hot and humid. However, the thermal regime, the regime of rain, sunshine hours are different to other sub- regions. The average annual temperature in Co Noi stations ranges from 19.70C -22.20C, in Moc Chau stations from 17.00C – 20.30C.The highest temperature at Co Noi in Juneis 25.80C, at Moc Chau in Julyis 23.40C. Temperatures vary greatly between winter and summer, day and night. However, temperaturehasa rising
trend in recent years.
3.2.1.6 Hydrometeorological characteristics Nam Sap watershed has 5 rainfall stations was built and operated from the 60s of the 20th century, but two of them were shut down. Only the Moc Chau, Co Noi and Ta Lang stations are now working with the reliable quality of measurement data. 3.2.1.7 Request input data for model
- The meteorological dataof watershed in 1962, 1973, 1980, 2010, 2011, 2012
and 2013 include rainfall data (hours), temperature, evaporation and flow
discharge (hours), sediment discharge (hours) at Thac Moc station.
- Data on the land use status, the types of soil in the Nam Sap watershed that
16
area limits toThac Moc station.
- The maps of terrain, river network, meteorological station network, soil and
land use.
3.2.1.8 Model calibration and verification
a. Model calibration
Author chose the August 1962 flood to adjust the parameters of the model. Model calibration results are shown in Figure 3-9 in the form of graphs comparing the calculated and measured results at the location of testing hydrological stations and indicators of corresponding NASH at each station.The model calibration results gave the value of Nash-Sutcliffe is 81%, the disparity between calculated and measured flood peakis 13m3/s. The correlation coefficient between the calculated results and measured data reaches 0.93.
The parameters of the sediment transport model
Inter-rill sediment transport coefficient (acl).
- Inter-rill detached erosion coefficient(ail). - Coefficient of soil erosion potential (ket). - - Coefficient of rill erosion (Ci). - Rill sediment transport coefficient (ac). b. Model validation and verification
Figure 3-9. Data comparison betweenmeasured and computed discharge at Thac Moc Station
17
Figure 3-10. Data comparison betweenmeasured and computed sediment concentration at Thac Moc Station
Figure 3-11.Measureddata of sediment discharge at Thac Moc station
3.2.2 Phieng Hieng watershed
3.2.2.1 Geography Phieng Hieng watershed is located in Bac Yen, Phu Yen districts of Son La province and occupies an area of natural 431km2, account for3.04% of the total area of the province. Located in the geographic scope: 104o24'41’’-104o38'8’’ longitude, 21o24'49’’-21o05'38’latitude. It borders the Suoi Gao stream
18
watershed to the West, Yen Baiprovince to the North, and Chieng Sai commune to the South.
heavily slopes,being fragmented, alternating
3.2.2.2 Topography Phieng Hieng watershed is located in inland region, having complex terrain,steep between mountainsand 100% natural area ofitlying on Da river basin. 3.2.2.3 Soil characteristics Phieng Hieng watershed has 3 main groups of soil, includingferalit gray soil,red brown soil, andgrey loam soil on mountain. 3.2.2.4 Features of vegetation cover and land use status
There are many types of soil in Phieng Hieng watershed suitable for crops.
Area of vacant land accounts for0.26%, Grassland is about 29.60%, poor forest
is 9.07%, young forest is 1.57% etc.
3.2.2.5 Climate features Phieng Hiengwatershed has a tropical climate, characterized by the general climate of the Northwest: Winter is cold, summer is hot and humid. However, the thermal regime, the regime of rain, sunshine hours are different to other sub-regions. There is no measuring station in the watershed, thus the author will use the data of Phu Yen, Bac Yenstations which are in the vicinity of watershed.
The average annual humidity at Phu Yen stations ranges from 77% - 83%, from
78% - 84% at Bac yen. The lowest humidity ranges from 78% - 79% March and
April.
3.2.2.6 Hydrometeorological characteristics There is no rainfall station in the Phieng Hieng watershed, but its vicinity having two rainfall stations, including Bac Yen and Phu Yen stations with reliable data. 3.2.2.7 Request input data for model - The meteorological dataof watershed in1973 and 1976 include rainfall data (hours), temperature, evaporation at Phu Yen station and flow discharge (hours), sediment discharge (hours) at Phieng Hieng station. - Data on the land use status, the types of soil in the Nam Sap watershed that area limits toThac Moc station.
19
- The maps of terrain, river network, meteorological station network, soil and land use.
3.2.2.8 Model calibration and verification Using the sediment discharge at Phieng Hieng station in 1973 to calibrate the set up model. Model calibration results are shown as follows:
Figure 3-20. Data comparison betweenmeasured and computed discharge at Phieng Hieng Station
Figure 3-21. Data comparison betweenmeasured and computed sediment concentration at Phieng Hieng Station
20
3.2.3 Construction of correlation equation
3.2.3.1 Sensitivity analysis of the model parameters
a. Method of sensitivity analysis of the model parameters
Sensitivity analysis is the study on the relationship between the input
information and output information of the model. Sensitivity can be calculated
by several methods or through qualitative or quantitativeanalysis.
To analyze the sensitivity of the parameters in the mathematical model, the
vectorconcept was given to each correspondingparameter (e.g. vector x takes
the values within a b, we write x[a, b]). The number of elements of [a, b]
more or less depends on the exact requirements of the method.
b. The results of the sensitivity analysis of model parameters
There are many parameters involving in the formation of sediment transport on
small and medium watershed which were modeled through model simulating
sediment transport in the watershed. Sensitivity analysis method can indicate
whether the parameters which play important role in the model or not. The
thesis analyses the sensitivity of a group of parameters: rainfall intensity,
detachment erosion coefficientfor linked rills, watershed slope, rill density, land
cover and soil erosion potential.
3.2.3.2 Analysis of the correlationamonginter-rillerosion,watershed slope and
rainfall intensity
Inter-rill sediment yield eroded in watershed is a function of rainfall intensity
and watershed slope. Simulating the problem with different intensities of
rainfall in the watersheds of Phieng Hieng and Nam Sap give us the results of
inter-rill erosion in watershed.
The basic equation for the inter-rillsoil detachmentis written as follows:
b.
The basic equation for the inter-rillsoil detachmentis written as follows:
(3-2) Ei= a.i2.Io
From the computed results,constructing the basic equation for inter-rill soil
detachment insmall and medium watersheds in the province of Son La is based
on calculatedfigures. The equation is written as follows:
21
. (3-3)
3.2.3.3 Analysis of the correlationamong rill erosion,watershed slope and
rainfall intensity
Rill sediment yield eroded in watershed is a function of rainfall intensity and
watershed slope. Simulating the problem with different intensities of rainfall in
the watersheds give us the results of rill erosion in watershed.
d.
The basic equation for the rillsoil detachmentis written as follows
(3-4) Er= c.i2.Io
The results of correlation analysis between the rill erosion in watershed, the
slope and rainfall intensity is shown in the following figures:
Figure 3-26. The correlation among rill erosion, watershed slope and rainfall intensity
22
Conclusions of chapter 3
Author chose the two watersheds to test the reliability of the model. The results
show that the graphs of flow and sediment transport between calculated and
measured data in two watershed is asymptotic. In addition, the author also built
the correlationbetweeninter-rill erosion,watershed slope and rainfall intensity
based on the sensitivity analysis of the parameters.
CONCLUSIONS AND RECOMMENDATIONS
The main findings of the thesis in studying the scientific basis to construct a
simulation model of sediment transport process in small and medium
watersheds, applied to the Northern mountainous watersheds in the province of
Son La in view of approaching the physical processes of sediment transport,
advanced methods and computational tools, especially in solving mathematical equations, combined with the language C++ programming leading to some conclusions as follows:
- Based on a review of simulation models of sediment transport on small and
medium watershed in Vietnam and worldwide, in the context of the watershed with steep terrain, the choice of language C++to simulate the sediment transport is reasonable, appropriate in the current conditions.Construction of tools for
simulation and development of technology calculating sediment discharge in
the small and medium watershed is reliable.
- Son La is a mountainous province in NorthernVietnam with steep terrain
conditions, high risk of erosion, land degradation due to surface sediment
transport.The factors affecting the sediment transport in the watershed include
rainfall, vegetation cover, slope, soil and other activities of humanity. The
factors create the materialsource for transporting sediment on the surface of
watershed is soil.
- Doing literature and empirical research on award scheme through the
difference Lax - Friedrichs method in scale of space and time in order to
improve the accuracy and stability of the model to the highest level.
23
- Applying computational model simulating sediment transport in the watershed
shows that rainfall is the main factor causing the sediment transport. The slope
is an important factor affecting erosion and surface runoff. If the slope is
steeper, the surface erosion is more serious and vice versa. If the slope of the
watershed increase by 30%, the amount of sediment in the watershed will
increase by about 12%.
- From rainfall and slope in the watershed, through correlation equation
betweenrill erosion,inter-rill erosion, rainfall and slope, we can calculate the
amount of rill and inter-rill erosion, then estimate the volume of sediment
transported to the outlet of the watershed.
New contributions
- Construction of a new numerical model simulating the flow and sediment transport on small and medium watersheds that uses Lax-Friedrichs Scheme and adds more elements of time and space to it to solve flow and sediment transport equation.
- Construction of the correlation equationbetweenthe resultsof inter-rill erosion
and rill erosion that can predict the amount of sediment eroded and transported
in watershedsbased on the intensity of rainfall.
Recommendations:
The simulation of sediment transport in river is only based on the sediment
balance, calculating for uniform sediment particles. Therefore, the next time we
should continue to study the erosion ofbed and bank.
The application of the developed model has been done in only two watersheds in
Son La province, thus we should continue to improve the model, and to apply for
other watersheds.
GIS technology should be connected to set up a model on the basis of digitizing
the detailed topography of the watershed, thus it will increase the accuracy of the
model.
24
PUBLICATIONS OF THE AUTHOR
1. Dao Tan Quy and Pham Thi Huong Lan. “Mapping landslide hazards of Son
La province”. Journal of Water Resources and Environmental Engineering. No.
51, December 2015).
2. Dao Tan Quy. “Construction of simulation model of sediment transport in
small and medium watershed. (Apply for Suoi Sap watershed in Son La
province)”. Scientific and Technical Hydro-Meteorological Journal. No. 657,
September 2015.
3. Dao Tan Quy. “Simulation of sediment transport models for small basin
(apply for Suoi Sap basin in Son La province)”. Journal of Water Resources
and Environmental Engineering. No. 48, March 2015.
4. Dao Tan Quy, Tran Kim Chau, Pham Thi Huong Lan and Nguyen The
Toan. “Application of geographic information system (GIS) and ranking
analysis process to map landslide hazards of Son La province”. Proceeding of
the annual conference on water resources, December 2014. Pages 474- 476.
5. Dao Tan Quy and Pham Thi Huong Lan. “Research on the theoretical basis
to construct a numerical model simulating sediment transport on medium and
small watershed”. Proceeding of the annual conference on water resources,
December 2013. Pages 175- 177.
6. Dao Tan Quy and Pham Thi Huong Lan. “Study on analyzing and choosing
the programming language for developing a software to simulate runoff and
sediment transportation in small and medium-sized basin”. Journal of Water
Resources and Environmental Engineering. No.40, March 2013.
