3 Spatial Averaging

3-D equations of motion Scaling => simplifications Spatial averaging Shear dispersion Magnitudes/time scales of diffusion/dispersion Examples

Navier-Stokes Eqns
Conservative form momentum eqn; x-component only
2u 2u 2u 1 p u 2 + (u ) + (uv) + (uw) fv = + + 2+ 2 2 x x y z t x y z

1 storage or local acceleration 2 advective acceleration 3 Coriolis acceleration [f = Coriolis param = 2sin(), = latitude] 4 pressure gradient 5 viscous stress

Turbulent Reynolds Eqns

u = u + u ' , etc (u = time average) 2 2 [(u + u ')(u + u ')] = u + u ' x x x u v w + + Continuity x y z u 2 + (u ) + (u v ) + (u w ) fv t x y z
Insert & time average (over bar)

e.g., x-component; term 2

x-momentum

3 u u 1 p u 2 = + u' + ' ' ' ' + u v u w x x x y y z z 4 5 2b

2a

Specific terms
Term 2a: could subtract u times continuity eqn

u v w + =u + x y z
u u u +v +w u x x x
Terms 5 + 2b

NC form of momentum eqn (term 2) x-comp of turbulent shear stress xx, xy, xz are turbulent (eddy) kinematic viscosities resulting from closure model (like Exx, Exy, Exz)

u u u '2 = xx u '2 2 xx x x u v u u ' v' = xy u ' v' xy + y y x u u w u ' w' = xz u ' w' xz + z z x

Specific terms, contd

Similar eqns for y and z except z has gravity. For nearly horizontal flow (w ~ 0), z-mom => hydrostatic

1 p 0= g z
4

z z=(x,y,t)

6 z

p = pa + gdz
z

Use in x and y eqns

Specific terms, contd

Term 4
p = pa + gdz x x z

z=(x,y,t)

pa s g + gdz = + x x x z

4a

4b

4c

4a atmospheric pressure gradient (often negligible) 4b barotropic pressure gradient (barotropic => = s = const) 4c baroclinic pressure gradient (baroclinic => density gradients; often negligible)

Simplification
Neglect xx; xy = h; zx = z Neglect all pressure terms except barotropic And drop over bars
u u u 2 h + + z + (u ) + (uv) + (uw) fv = g x y y z z z y t x 1 2a 3 4b 2b1 2b2

Contrast with mass transport eqn

c + (uc) + (vc) + ( wc) t x y z = c c c Ex + y E y y + z E z z r x y

2a

2b1

2b2

Pressure gradient (4) in momentum eqn => viscosity (2b) not always important to balance advection (2a); depends on shear (separation). For mass transport, diffusivity (2b) always needed to balance advection (2a)

Comments
3D models include continuity + three components of momentum eq (z may be hydrostatic approx) + n mass transport eqns Above are primitive eqs (u, v, w); sometimes different form, but physics should be same Sometimes further simplifications Spatial averaging => reduced dimensions

Further possible simplifications

Neglect terms 2a and 2b1

u u + fv = g + z t x z z
Also neglect term 1 u fv = g + z x z z Also neglect term 2b2 fv = g x

Linear shallow water wave eqn Steady Ekman flow

Geostrophic flow

Spatial averaging
3-D equations (x,y,z,t) ocean

z-vertical

y-lateral

x-longitudinal 2-D lateral average (x,z,t) long reservoir; deep estuary/fjord

2-D vertical average (x,y,t) shallow coastal; estuary

1-D vertical & lateral average (x, t) river; narrow/shallow estuary

1-D horizontal average (z, t) deep lake/reservoir ocean

Comments
Models of reduced dimension achieved by spatial averaging or direct formulation (advantages of both) Demonstration of vertical averaging (integrate over depth then divide by depth, leading to 2D depth-averaged models) Discussion of cross-sectional averaging (river models)

Vertical Integration => 2D (depth-averaged) eqns

z
H(x,y,t) (x,y,t)

y
h(x,y,t)

u ( x, y, z , t ) = u ( x, y, t ) + u" ( x, y, z , t ) u ( x , y , z , t ) = U h ( x , y , t ) + u" ( x , y , z , t )
(notes use)

Vertical Integration, contd

z

H -h

u"

c"

1 u ( x, y , t ) = H
concentrations into

u ( x, y, z, t )dz
u + u" , c + c"
, etc. and spatially average

In analogy with Reynolds averaging, decompose velocities and

Depth-averaged eqns
Continuity
t + x (u H ) + (v H ) = 0 y
u v w + kinematic surface bc of , z x y

from

Mass and Momentum

2 = ( ) u dz x x h

2 2 = u H + u" H x x

( )

straight forward except for NL terms 2 2 u + u" u + u" dz = u dz + 2 u u" dz + u" dz x x

)(

momentum dispersion = + ( uc ) dz u c H u" c"H x x x h

mass dispersion

Depth-averaged eqns, contd

x-momentum
(u H ) + u 2 H + (u v H ) fv H = gH t x y x

u H + L x x

v sx bx + H T y x

u 2 u" H H x x x
Depth-ave long. diff

u u" v"H H y y y

Long. dispersion

Depth integrated eqns, contd

Mass Transport
(c H ) + (u c H ) + (v c H ) = t x y c + + HE L x x
c " " u c H H E x x x

c c c + HE E E T z z z s z y y

c " " v c H H E y y y

Depth-ave long. diff

Long dispersion

Comments
L, T, EL , ET are longitudinal and transverse momentum and mass shear viscosity/dispersion coefficients. L >> T and EL >> ET , but relative importance depends on longitudinal gradients Dispersion process represented as Fickian (explained shortly)

Boundary Conditions: momentum

sx 2 2 = C D U w + Vw U w
Surface shear stress due to wind (components Uw and Vw); external input (unless coupled air-water model) Assumes Uw >> us; CD = drag coefficient ~ 10-3 (more in Ch 8)

bx = Cf u 2 +v 2 u
u*
2

Bottom shear stress caused by flow (computed by model) Different models for Cf (DarcyWeisbach f; Manning n; Chezy C), e.g.,

bx

f = u 2 8

Boundary Conditions: mass transport

c Ez z s

Surface mass transfer (air-water exchange)

c =0 z s
No flux (dye, salt)

c >0 z s
Source (DO)

c <0 z s
Sink (VOC)

Boundary Conditions: mass transport

c Ez z b

Benthic mass transfer (sediment-water exchange)

c =0 z b

c >0 z b

c <0 z b
Sink (trace metals bound by anoxic sediments)

No flux (dye, salt)

Source (pore water diffusion)

Magnitude of terms: Ez
E z ~ u ' L ~ u* H u* = shear velocity = b / =
x

f u, 8

f 0.02 => u* = 0.05u

A
Fg = AxSg u* =
2

u
2

= F f = b px = u* px

ASg = R H Sg ; u* gHS ; p R H = A / p = hydraulic radius H

Normal flow; gravity balances friction

Ez contd
E z = 0.07u* H
Seen previously; from analogy of mass and momentum conservation (Reynolds analogy) and log profile for velocity

vm

0.5 H 2 Ez

0.5 H 2 7H 0.07u* H u* 7 Hu ; if u = 20u* u* xvm 150 H

xvm = u vm

Transverse mixing: ET
ET 0.08 0.24 u* H (say 0.15)
Laboratory rectangular channels

ET 0.2 4.6 u* H
tm
xtm

(say 0.6)

Real channels (irregularities, braiding, secondary circulation) B = channel width

0.5B 2 0.5 B 2 = 0.6u* H ET 0.5B 2 = u ; if u = 20u* 0.6u* H

17 B 2 = H

Example
B = 100 m, H = 5m, u = 1 m/s Xvm = 150H = 750 m xtm = 17B2/H = (17)(100)2/5 = 34,000 m (34 km)
It may take quite a while before concentrations can be considered laterally (transversally) uniform

Simplifications
Steady state; depth-averaged; no lateral advection or long dispersion; no boundary fluxes
c c c c (c H ) + (u c H ) + (v c H ) = + HE HE + E E L T z z y t x y x x z s z b y

Hu

c c = HE T x y y

Uniform channel

c ET 2 c = x u y 2
const

Simple diffusion equation: solutions for continuous source at x=y=0 y B u x

1.0 qd - dimensionless cumulative discharge 0.8 0.001 0.6 0.4 0.2 30 0 0.0001 2.0 cd - dimensionless concentration 1.6 1.2 0.8 0.4 0 q = 1 qd = 0 10 0.001 0.01 0.1 0.1 1=cd 3 1.1 1.3 1.01 0.01 0.3 0.7 0.9 0.99

y B

Holly and Jirka (1986)

xET uB2

cBu H & m

Note:

xtm
xET uB2

0.5 B 2 u ET

0.2

0.4

0.6

xd - dimensionless longitudinal distance

Figure by MIT OCW.

A useful extension: cumulative discharge approach (Yotsukura & Sayre, 1976)

Use cumulative discharge (Qc) instead of y as lateral variable Q
Qc = H ( y ' )u ( y ' )dy '
0 y

Q
Qc

c 2 c = H u ET x Qc Qc

D D behaves mathematically like diffusion coefficient, but has different dimensions; can be approximated as constant (cross-sectional average): Q 1 2c c 2 D = H u ET dQc => = D 2 Q 0 x Qc
c

u
H

Can use previous analysis

Longitudinal Shear Dispersion

Why is longitudinal dispersion Fickian? Original analysis by Taylor (1953, 1954) for flow in pipes; following for 2D flow after Elder (1959) z
u
u

Longitudinal Shear dispersion

Why is longitudinal dispersion Fickian? Original analysis by Taylor (1953, 1954) for flow in pipes; following for 2D flow after Elder (1959) z H
to

t
t1 t2

L2

Shear dispersion, contd

z H
to

t
t1 t2

L2

c c c c + u( z) = Ex + Ey x x x y y t u = u + u"; c = c + c"; x = + u t ; t = c" c c" c c" c c c" Ex Ex = + u" + u" + + + Ez + Ez z z z z

1 2 3 4 5 6 7

Shear dispersion, contd

z H
to

t
t1 t2

a ) L >> H

=> longitudinal dispersion << vertical

L2 5, 6 << 7 4 << 3 1, 2 << 3

b) c >> c" => c" << c c) x >> L => << u"

c c" c c" c c" c" Ex Ex + + u" + Ez + u" + = z z

1 2 3 4 5 6 7

Shear dispersion, contd

z H
to

t
t1 t2

c" c = Ez u" z z
3 7 We want

L2 Differential advection balanced by transverse diffusion

u" c"

and to show that it is ~

c EL x

Integrate over z twice to get c; multiply by u; integrate again and divide by H (depth average); add minus sign

Shear dispersion, contd

1 1 c 1 u" u" dzdzdz u" c" = u" c" dz = H 0 H 0 0 E z 0
H H z z

EL = I h

H2 Ez

u"2

Ih = dimensionless triple integration ~ 0.07

EL!!!

u" ~ u* ; E z ~ u* H
Elder (1959) using log profile for u(z) Taylor (1954) turbulent pipe flow (ro= radius)

EL = 5.9u*H EL = 10u*ro

Use EL to compute L or use measured L to deduce EL

Comments
EL involves differential advection (u) with transverse mixing in direction of advection gradient (Ez) EL ~ 1/Ez; perhaps counter-intuitive, but look at time scales:
EL ~ H2 Ez
Tc Uc2

u"2

Recall Taylors Theorem

D ~ u 2 R( )d
0

A thought experiment
Consider the trip on the Mass Turnpike from Boston to the NY border (~150 miles). Assume two lanes in each direction, and that cars in left lane always travel 65 mph, while those in the right lane travel 55 mph. At the start 50 cars in each lane have their tops painted red and a helicopter observes the dispersion in their position as they travel to NY 1) How does this dispersion depend on the frequency of lane changes? 2) Would dispersion increase or decrease if there were a third (middle) lane where cars traveled at 60 mph?

Thought experiment, contd

EL ~ H
2

(u" )

Ez

What are the analogs of H, Ez and u

Thought experiment, contd

H2 (u" ) 2 EL ~ Ez
H ~ number of lanes Ez ~ frequency of lane changing u ~ difference between average and lane-specific speed 1) Decreasing Ez increases dispersion (as long as there is some Ez) 2) Increasing lanes increases H, decreases mean square u,
(65 60) 2 + (55 60) 2 (2 lanes) = 50 / 2 (u" ) = 2 (65 60) 2 + (60 60) 2 + (55 60) 2 2 (u" ) = = 50 / 3 (3 lanes) 3
2

If Ez is constant, net effect is increase in EL by (3/2)2(2/3) = 50%

1D (river) dispersion
u = u + u"; c = c + c" 1 c = cdA AA

Insert into GE and spatial average

Continuity
A + ( Au ) = q L t x

qL = lateral inflow/length [L2/T]

Mass Conservation (conservative form)

c A u c A E + Ari + q L c L ( Ac) + ( Auc) = " " x x x x t

AE L

c x

Longitudinal dispersion again

1D (river) dispersion, contd

NC form from conservative equation minus c times continuity
c ( Ac) + ( Auc) = " " A E A u c + Ari + q L c L x t x x x

A c + ( Au ) = q L t x
Mass Conservation (NC form)

qL c c 1 c (c L c ) +u AE = r + + L i t x A x x A
Note: if cL > c, c increases; if cL < c, c decreases (dilution)

EL for rivers
Elder formula accounts for vertical shear (OK for depth averaged models that resolve lateral shear); here we need to parameterize lateral and vertical shear. Analysis by Fischer (1967)

B 2 2 c u" c" = I b u" ET x

EL

B = river width ET = transverse dispersion coefficient Ib ND triple integration (across A) ~ 0.07

Same form as Elder, but now time scale is B2/ET, rather than H2/Ez. ET > Ez, but B2 >> H2 => this EL is generally much larger

EL for rivers, contd

Using approximations for u, ET, etc. or

u 2B2 E L 0.01 u* H
u B EL 0.01 H u* H u *
2 2

Fischer (1967); useful for reasonably straight, uniform rivers and channels

if u 20u*
EL B 4 u* H H
2

Magnitude of terms, revisited

Ez 0.07 u* H ET 0.6 u* H EL 6 u* H EL 10 u* ro
2 2

Vertical Diffusion Transverse Diffusion in Channels Longitudinal Dispersion (depth-averaged flow) Longitudinal Dispersion (turbulent pipe flow)

u B EL 0.01 Longitudinal Dispersion (rivers) u* H u* H

Previous Example, revisited

B = 100 m, H = 5m, u = 1 m/s, u* = 0.05u =0.05
0.01u 2 B 2 (0.01)(1) 2 (100) 2 EL = = u* H (0.05)(5) = 400 m 2 /s

d x 2 = 2E L dt
xtm 0 .5 B 2 u = ET

=> Gaussian Distribution; but only after cross-sectional mixing xtm

xtm = 34 km if point source on river bank;

Previous Example, revisited

B = 100 m, H = 5m, u = 1 m/s, u* = 0.05u =0.05
0.01u 2 B 2 (0.01)(1) 2 (100) 2 EL = = u* H (0.05)(5) = 400 m 2 /s

d x 2 = 2E L dt
xtm 0 .5 B 2 u = ET

=> Gaussian Distribution; but only after cross-sectional mixing xtm u B

4 times less if point source in mid-stream

Previous Example, revisited

B = 100 m, H = 5m, u = 1 m/s, u* = 0.05u =0.05
0.01u 2 B 2 (0.01)(1) 2 (100) 2 EL = = u* H (0.05)(5) = 400 m 2 /s

d x 2 = 2E L dt
xtm 0 .5 B 2 u = ET

=> Gaussian Distribution; but only after cross-sectional mixing xtm

Less still if distributed across channel (but not zero)

Storage zones
Real channels often have backwater (storage) zones that increase dispersion and give long tails to c(t) distribution

A As

c(x,t)

Storage zones, contd

c q 1) c + u c = 1 AE L + L (c L c ) + (c s c) t x A x x A dc s A (c s c ) = 2) dt As
A(x) = cross-sectional area of main channel As = cross-sectional area of storage zone = storage zone coefficient (rate, t-1; like qL/A) If you multiply 1) by A and 2) by As, the exchange terms are A(cs-c) and A(cs-c) Really the same process as longitudinal dispersion, but instead of cars in either fast or slow lane, some are in the rest stop. Main channel Storage zone