Hohai University

Water Environmental Mo
del and Inverse Problem

Dr. Xiaodong Liu

1

Outline of the Lecture

Introduction to WEM
A personal perspective on WEM
Mixing in Rivers
Inverse Problem of WEM

Outline of the Lecture

Introduction to WEM
A personal perspective on WEM
Mixing in Rivers
Inverse Problem of WEM

Outline of the Lecture

Introduction to WEM
A personal perspective on WEM
Mixing in Rivers
Inverse Problem of WEM

Outline of the Lecture

Introduction to WEM
A personal perspective on WEM
Mixing in Rivers
Inverse Problem of WEM

Methods for studying water environment

Observation is often the first step in as

king environmental questions.

Questions may form the first step in d

esigning experiments and models.

1 Methods field
observation
Field work at Lake Tai 2003

Lake Tai or Lake Taihu is a large freshwater lake in the Yangtze Delta plain
near Shanghai, China. The lake belongs to Jiangsu province With an area of
2,250 square kilometers and an average depth of 2 meters (6.6 ft), it is the
third-largest freshwater lake in China, after Poyang and Dongting. In
Taihu recent years, Lake Tai has been plagued by pollution as a result of rapid
October 2007
economic growth in the surrounding region.

Location
Lake
Taihu

Ya
ng
t

Taihu

ze

Ri

ve
r

Taihu Sensors
Vaisala
WXT 510
Air T,
WD, WS,
Rain, AP

ZebraTech
DO

SeaPoint
Turbidity
and
Chlorophyll
a
Apogee
Solar
Radiation

Instrument
Northwest
pH
iQuest
Datalogger
and GPRS

NexSens
Temp
Chain

10

Lake high
watertech
quality
of monitoring!
field observation
1 Real time monitoring 2 Remote Sensor

11

Remote Sensing

12

2 Experimentation

Experiments can be used to test hypothe

ses.

An ecologist may set up an artificial envir

onment in a laboratory or greenhouse, or ca
refully alter conditions in selected parts of n
atural ecosystems.

13

Methods lab experiments

Sediment flume

Sediment flume

Phytoplankton growth
14

3 Modeling

Many environmental events occur over suc

h long periods of time or over such large dista
nces that they are difficult to study directly.

Ecologists make models to help them unde

rstand these phenomena.

15

Advantages of Environmental Modelin

g

A good model can reveal more about a environme

ntal processes and responses than we might other
wise learn through conventional (i.e., limited num
ber) sampling techniques. Spatio-temporal co

ntinuity & large-scale.

Modeling can predict how water quality might beh

ave before any pollution occurs. Prediction.
Modeling can be used to simulate different mitigat
ive measures to minimize potential impacts from
development activities. Scenario.
16

What is a model?
- All models are wrong, but some are us
eful.
George Edward Pelham Box FRS
(born 18 October 1919) is a
statistician, who has made
important contributions in the areas
of quality control, time-series
analysis, design of experiments,
and Bayesian inference. Box
famously wrote that "essentially, all
models are wrong, but some are
useful" in his book on response
surface methodology with
Norman R. Draper.
17

Understanding Models

Surplus Yield Model (Lackey

and Hubert 1978)

What is a model?
A model is a simplified r

Equilibrium Yield

epresentation of the real

world
There are two types of m
odels
Conceptual
Mathematical

Fishing Effort
Developed by: Hagley

Updated: May 30, 2004

U5-m21a-s18

Conceptual Models
What are they?
Qualitative, usually based on graphs
Represent important system:
components
processes
linkages
Interactions

Developed by: Hagley

Updated: May 30, 2004

U5-m21a-s19

Conceptual Models
When should they be used?
As an initial step
For hypothesis testing
For mathematical model development

As a framework
For future monitoring, research, and management
actions at a site

Developed by: Hagley

Updated: May 30, 2004

U5-m21a-s20

Conceptual Models
How can they be used?
Design field sampling and monitoring programs
Ensure that all important system attributes are measured
Determine causes of environmental problems
Identify system linkages and possible cause and effect relati

onships
Identify potential conflicts among management objectives
Anticipate the full range of possible system responses to

management actions
Including potential negative effects

Developed by: Hagley

Updated: May 30, 2004

U5-m21a-s21

Conceptual Model Example

Increased nutrient loading

Primary productivity

Macrophytes

Algal biomass

Increased pH

Transparency

+
% blue-green algae

Fish cover

Grazing impact

+
Sedimentation rate

Nutrient release
due to anoxia

Mean zooplankton
size

Zooplankton refuges

Hypolimnetic oxygen depletion

Developed by: Hagley

Updated: May 30, 2004

U5-m21a-s22

Mathematical Models
What are they?
Mathematical equations that translate a conceptual unde

rstanding of a system or process into quantitative terms

(Reckhow and Chapra 1983)

How are they used?

Diagnosis
E.g., What is the cause of reduced water quality in a lake?

Prediction
E.g., How long will it take for lake water quality to improve, o
nce controls are in place?

Developed by: Hagley

Updated: May 30, 2004

U5-m21a-s23

Categories of Mathematical Models

Type
Empirical
Based on data analysis

Mechanistic
Mathematical descriptions based
on theory

Time Factor
Static or steady-state
Time-independent

Dynamic
Describe or predict system
behavior over time

Treatment of Data Uncertainty and Variability

Deterministic
Do not address data variability

Developed by: Hagley

Stochastic
Address variability/uncertainty

Updated: May 30, 2004

U5-m21a-s24

Mathematical Models
When should you not use a model?
If you do not understand the problem or system well eno

ugh to express it in concise, quantitative terms

If the model has not been tested and verified for situation
s and conditions similar to your resource
It is important to understand model:
Structure
Assumptions
Limitations

Developed by: Hagley

Updated: May 30, 2004

U5-m21a-s25

Mathematical modeling
.

Mathematical modeling involves teamwork

26

Mathematical Modeling Proce

ss

27

Development & application

28

Teamwork!!!

29

Common used
Environmental model
1 EFDC (EPA)
2 CAEDYM (CWR)
3 WASP (EPA)
4 ECOPATH (
)

5 MIKE (DHI)
6 QUAL2E (EPA)
7 Delft3D
30

Outline of the Lecture

Introduction to WEM
A personal perspective on WEM
Mixing in Rivers
Inverse Problem of WEM

31

Basic Principle of Mechanistic Models

Laws of Conservation

Conservative properties are those that are no

t gained or lost through ordinary reactions. T
herefore we can account for any change by si
mply keeping track of all those processes that
can cause change

Examples of conservative properties

Mass (water mass, constituent mass)
Momentum
Heat

Model Equations NS equation

HYDRODYNAMIC EQUATIONS
The equation of continuity, based on the
conservation of water mass, predicts water
heights (heads) and velocities

33

Methods model
CASE-Lake Tai

wind-driven circulation

34

Water Quality Model

Conservation of constituent mass
Control
Volume

y
x

Three Dimensional Pollutants Transport

Equation

C U x C U y C U z C

t
x
y
z
x

Ex

C

C

C

z
S
y

x y
y z
z

36

Typical waste materials

Natural Inorganic Salts and Sediments

Waste Heat
Organic Wastes
Trace Metals
Synthetic Organic Chemicals
Radioactive Materials
Chemical and Biological Warfare Agents
37

WEM

Solve
rs
Output
38

Categories of WEM

39

Outline of the Lecture

Introduction to WEM
A personal perspective on WEM
Mixing in Rivers
Inverse Problem of WEM

40

What is Mixing Process

Hydrologic Transport Processes
Advection
Diffusion (molecular)
Diffusion (Turbulent)
Dispersion
Mixing
Evaporation
Particle Settling
41

Hydrologic Transport Processes

Advection: Transport by an imposed curre
nt system, as in a river waters.

Diffusion (molecular): The

scattering of particles by
random molecular
motions,which may be
described by Ficks law.

42

Hydrologic Transport Processes

Diffusion (Turbulent):The random scatteri
ng of particles by turbulent motion. (whic
h are much larger than molecular diffusio
n.)
Laminar flow pattern

Reynolds
test

Straight line
Turbulence flow pattern

43

Hydrologic Transport Processes

44

Hydrologic Transport Processes

Dispersion: The scattering of particles by t
he combined effects of shear and transve
rse diffusion.

uniform
flow
shear
flow

45

Consider flow between two paralllel plates of

infinite extent
Uniform

Shear

u
O

Dispersion
y

y
u(y)

O
y

y
c(y)

c(y)
X

O
6/30/15 [Link] PM

u(y)

47

Hydrologic Transport Processes

Mixing: Diffusion of dispersion as described
above; any process which causes one parc
el of water to be mingled with or diluted by
another.

C U x C U y C U z C
C

C

C
E z

Ex
E y
S
t
x
y
z
x
x y
y z
z

Advection

Diffution

48

Diffusion Coefficients

E hu
Ex x hu

E y y hu

Ez z hu
49

x y z

Ez 0.067hu*
E y a y hu*
0.1 ~ 0.2 0.15
ay
0.4 ~ 0.8 0.6

Ex f E y

Ez

Mixing in Rivers
Consider a stream of effluent discharged into a river,
as sketched in the figure.
C

B
A

A-

B-

C-

Three stages in the mixing of an effluent into a river

Stage 1 Vertical Mixing

Effluent ~ Vertical mixing uniform
section

Mixing: The initial momentum of the discha

rge determine the rate of waste dilution.
3D model shouid be applied to simulate the
mixing process of pollutants.
C U x C U y C U z C
C

C

C

E
x

z
S
y

t
x
y
z
x
x y
y z
z

Stage 2 Transverse Mixing

Vertical mixing uniform section ~ Transverse mixing uniform section

2D model shouid be applied to simul

ate the mixing process of pollutants.
C U x C U y C
C

C

E
x

t
x
y
x
x y
y

Stage 3 longitudinal Mixing

Transverse mixing uniform section ~
Up stream

Stage 1

Stage 2

Stage 3

effluent

To simulate the mixing process of pollutan

ts, 1D model should be applied.

Outline of the Lecture

Introduction to WEM
A personal perspective on WEM
Mixing in Rivers
Inverse Problem of WEM

54

WEM

Output
C(x,y,z
,t)

55

Inverse Problems of WEM

Output
C(x,y,z
,t)

56

Determination of Ex
The one-dimensional dispersion equation
M
V

Under the conditions of instantaneous source and

steady flow, the concentration in downstream can be
obtained.

x
o
57

Determination of Ex

Inverse
Model
Ex=?
58

CASE STUDY...

59

1 Case Study: Lake Chao Water Diversion

Water Environmental Model

3D Lake Model

1D
river
model

2D
Yangzi
River
model

Model Verification
Water quality models must be tested with real field data
under baseline conditions to ensure that they work!
Field sampling must consider several dimensions:

Depth
Sampling location
Seasonality
Annual variation

Analytical error and natural

variability must be considered

Model Verification

x y 150m
3
Sampling
location

Mesh

Model Verification
TN
V3

V1

V4

V2

V5
V6

The development of Management System

Thanks
Lake Hamilton, NZ
66