Introduction to

Observational Physical Oceanography

12.808

Class 16, 10 November, 2009

1:05 to 2:25

these slides are online at

www.whoi.edu/science/PO/people/jprice/class/miscart/Class16-10Nov09.ppt
 Where are we?

•Stirring and mixing last

•Advection and diffusion class

Section 7: Momentum balance on a rotating Earth,

geostrophy, and the large scale circulation of the oceans

• the Coriolis force

•Autonomous Systems Lab Tour today’s
class
Assigned reading:

-Hakkinen and Rhines, Science, 2004

-Kerr, Science, 2004

To be presented on Tuesday, November 17th

Volunteers?
 the Coriolis force
1) What is the Coriolis force and why or how does it arise?
The Coriolis force is an ‘inertial force’ that arises because the
Earth is rotating with respect to the fixed stars and is therefore
a noninertial reference frame.

2) Is it important? Oh yes, the momentum balance for large

scale, extra-tropical motions of the atmosphere and oceans is:

Coriolis force ~ pressure gradient.

3) A thorough understanding of the Coriolis force requires

mathematical care and detail that we will not attempt here; see

'Topics in Fluid Dynamics', MIT/OCW:

Photo courtesy of UCSD

Consider a projectile fired from

the North Pole towards a target
at the equator:
Eastward velocity of the
surface of the Earth is
FASTEST at the equator.
Initially, the projectile has
same eastward velocity of
the Earth, but as it moves
southward the eastward
velocity of the Earth
becomes larger than that of
the projectile.
In relation to the Earth, the
projectile moves southward
AND westward!
 towards the Coriolis force:

1) Review of inertial forces/noninertial reference frames.

a) linear acceleration
b) rotation at a steady speed, centrifugal force
c) steady rotation of fluid-attached reference frame.

2) The Coriolis force in 3-d; the horizontal components.

3) Modes of the horizontal momentum equation (next class)

a) inertial oscillations
b) steady, forced (geostrophic) motion

4) What am I supposed to know from all of this? (next class)

 The equations of motion

r r Motion of a particle given by the sum

F = ma of all forces on that particle
r r
du
F= m
dt
r 2r
d x There are many forces on a
F= m 2
dt parcel of water, including
pressure gradient, friction,
gravity, and the Coriolis force.
 inertial and noninertial reference frames
Classical dynamics in an inertial reference frame:
for a single particle of mass m (m is constant)
d 2X
m 2  F,
dt
where F is due to interactions with other particles.
F on this particle requires - F on other particles
so that the total momentum is conserved.

If we had the choice, we would always do

mechanics in an inertial reference frame.
However, the Earth is rotating, and so when we
make measurements with repsect to the Earth,
i.e., when we observe winds and ocean currents,
we are observing from a non-inertial reference frame.
 if at rest or moving steadily:
you feel an upward force, W,
d2z
m 2 = 0 = W - mg or
dt
W = mg,
because the force of gravitational
mass attraction, -mg, is balanced
by the rigid floor
W = mg
 when your elevator is accelerated upwards
you feel an increased weight
a
d2z
m 2 = ma = W - mg, or
dt
W = m(g  a),
because of the acceleration, a

W = m(g + a)
 from inside the evelator,
a non-inertial reference frame,
there is no acceleration, and the
consequence appears to you
to be an increased gravity
that seems to have the same
effect on every body, i.e.,
everything weighs more than usual in this
accelerated, and thus noninertial
reference frame

W = m(g + a)
 while in free fall you feel no weight
d2z
m 2 = W - mg,
dt

- m g = W - mg
or W = 0.

a = -g from inside the evelator,

it appears that gravity has vanished

evidently, inertial forces are relative to

a reference frame, but within a given reference frame
they can be quite real, i.e., the change of W.
 dynamics seen from an inertial reference frame

noninertial
frame
viewed from the inertial frame,
the particle's acceleration is

d2X
2
= F/m
dt
inertial frame
 dynamics observed from a noninertial reference frame
X  XO  X ',
noninertial
dX dXO dX'
frame   ,
dt dt dt
d 2 X d 2 XO d 2 X'
2
 2
 2
dt dt dt
d 2X
if 2
F/m
dt
inertial frame then the acceleration seen
from the noninertial reference frame is
d 2X ' d 2 XO
2
 F/m
dt dt 2
d 2 XO
where 2
is the acceleration of the noninertial frame.
dt
The equation of motion in a noninertial reference frame
d 2X ' d 2 XO
m 2 Fm
dt dt 2

an ‘inertial force’

These 'inertial forces ' are in a

direction opposite to the
acceleration of the accelerated reference
frame, and are exactly proportional
to the mass of the particle.
The Earth is rotating and Ω
therefore non-inertial!
Z’

Y’

X’
Z

,Z0
, Y 0
X 0

Y
Our goal: find
X the inertial
accelerations
that we on Earth
experience
 inertial forces due to steady, circular motion

a centrifuge at rest
 W Imagine the view from inside the box,
a noninertial reference frame:
Relative to this frame, you are
at rest, and something exactly like
m2r gravity is pushing outwards with a
magnitude m 2 r.You feel
the reaction force, W,
0 = W + m 2 r ,
r where m 2 r is an inertial force
dubbed 'centrifugal force.'

centrifugal has Latin roots:

 'centri' = central
steady 'fugal' = fleeing
rotation
 W

The view from outside the box:

you are moving in a steady
circular motion, with acceleration
toward the center at m 2 r.
This requires a centripetal (or
solid body center seeking) force,
rotation velocity: m 2 r  W.
V = Ω×r
V = Ω 'cross' r

steady
rotation
 vector cross-product

magnitude:
Ia x bI = IaI IbI sin()
the signed area of a
parallelogram:

direction: for more on the

use the right hand rule vector cross product
to find that the see links in slide 45
the product vector is
perpendicular
to a and b
 the centripetal acceleration associated
with steady, circular motion
In the radial direction:
V  r V  r
 V  r sin( )   r  t ,
V  2 r  t , 
V dV V time = t
  r, or,
2
  2 r.
t dt
r
In vector form : V  Ω  r, and t + t
dV     t
 Ω  Ω  r is the centripetal
dt
acceleration of a particle that is moving
in a circular motion at constant speed, r.
Note that the direction of V changes but
the magnitude does not.
solid body velocity: W
V  Ωr
The view from outside the box:
you are moving in (accelerated)
circular motion, with acceleration
toward the center at m 2 r.
This requires a centripetal force,
acceleration : m 2 r  W in the radial direction.

dV centripetal has Latin roots

 ΩΩr
dt 'centri' = center
'petal' = towards
 Ω 2 r (radial)

steady
rotation
 Inertial forces to this point:

1.) are relative to a reference frame

2.) may be accounted for exactly, provided we know the

acceleration of the reference frame,

3.) an acceleration (change in direction of a velocity vector)

results from steady rotation alone.
 plumb bobs

g
a fluid at rest
 plumb bobs
show the local
vertical
g
a fluid at rest

set the fluid into a steady rotation; a

level surface is then a parabola.


  2r 2 / 2g g
 

r g

the particle will be in equilibrium at a given radius if it is moving at

V  r, or V  Ω  r, the solid body rotation velocity.
dV
The inertial frame description :  Ω  Ω  r   g ,
dt

which is all radial : -  r   g
2
.
r

The rotating frame description : nothing's happening!

How is Earth similar and different from this spinning disk?
Earth is a sphere, and is in a gravitational/rotational balance.
Our definition of level is still a good one.

centrif F = (Req – Rpole)/Req

for Earth, F = 21/6370

=0.0033

for Saturn, F = 0.1

The Earth is out of round just enough so that the

centrifugal force is cancelled for things that are at rest in
the rotating (Earth) frame.
balanced motion in a rotating (parabolic) ref frame

balanced at this r
with V = xR
 balanced motion in a rotating (parabolic) ref frame

balanced at this r
with V = xR

What happens to the balance of forces if we move

outward?
 will not balanced if displaced to this larger r

if balanced at this r = r0

If displaced to a larger r, the particle will be in an environment:

1) having a larger positive radial velocity, so the particle will have

a negative (relative) azimuthal velocity, and

2) it will be accelerated back toward r = r0 because the

radial slope at r > r0 will be larger than required to balance the
centrifugal force of the particle.
In the fixed/inertial frame (looking at the spinning
table from above) it appears that the particle moves
outward on the table and continues to rotate with the
same radial velocity, while the table spins underneath
it.

Video courtesy
Wikipedia!
In the rotating/non-inertial frame, the particle moves
outward on the table but also appears to veer toward
the right (developing a negative radial velocity)

Video courtesy
Wikipedia!
The velocities seen in the two different reference frames
are different! The difference between the two is simply the
solid body rotation velocity!

Video courtesy
Wikipedia!
The velocity of the particle as seen in the non-inertial
reference (rotating) frame can be expressed as follows:
r r ′
dX ' r r ⎛ dX ⎞
= (−Ω × X ') + ⎜ ⎟
dt ⎝ dt ⎠

Velocity in rotating Velocity in fixed

frame frame
€
Solid body rotation

The difference between the two velocities is simply the

solid body rotation!!!
The velocities seen from the inertial and the rotating
(noninertial) reference frames differ by the
solid body rotation,
'
dX'  dX 
 - Ω  X'    ,
dt  dt 

the implied relationship of time derivatives is general,

and applies to all vectors:
'
d( )'  d( ) 
 - Ω  ( )'    ,
dt  dt 

apply this operator to the velocity dX'/dt to find

'
2
d X' dX '  2
d X 
 - 2Ω  - Ω  Ω  X '   2 
dt 2 dt  dt 

(for the details of this, see 'a Coriolis tutorial')

 the relationship between acceleration seen in the rotating
frame and the acceleration seen in the inertial frame :
'
d 2 X' dX '  d2X 
2
 - 2Ω  - Ω  Ω  X '   2 
dt dt  dt 

acceleration seen in the rotating frame

Coriolis force

centrifugal force
cancelled by
Earth's oblate
acceleration seen in the
shape
inertial frame (and rotated)
 the rotating frame equation of motion
when the centrifugal force is cancelled:
Polaris
d2X dX
2
 - 2Ω  F/m
dt dt

acceleration seen other forces

in the rotating frame Coriolis force acting on the parcel

where Ω is Earth's rotation

vector measured with respect to 
the distant, 'fixed' stars,
2π
=
23 hrs 56 min 4.1 sec

 7.292 x105 sec 1

 r
−2Ω× V = Coriolis The Coriolis force has a vertical component
(normal to a level surface) that is much smaller
than gravity and usually ignored; the horizontal
 component of the
Coriolis force is often very important:

Northern latitudes: the

C = -2xV horizontal component of C
is toward the equator, to the
right of V. The Coriolis force
a velocity V tends to deflect winds and
that is east currents to the right

Southern latitudes, the

horizontal component of C
is also toward the equator,
C = -2xV
which is now to the left of V
 r
−2Ω× V = Coriolis
 in the northern hemisphere the
horizontal component of the
Coriolis force is to the right....

a velocity V
that is east ‘what’s it do right on the
equator?’

C in the southern hemisphere the

horizontal component of the
Coriolis force is to the left....
 r
−2Ω× V = Coriolis
the 3-D Coriolis force


the vertical component of C

C = -2xV
a velocity V
that is east
(into the board) horizontal component; in
northern latitudes, the
horizontal component of C
is toward the equator, or to the
right of V
 The 3 - dimensional vector Coriolis force;
dV
m   m 2Ω  V ,
dt
where Ω is Earth' s rotation vector and V is the velocity
as seen from an Earth - attached, rotating reference frame.
 r
−2Ω× V = Coriolis
in the northern hemisphere the
 horizontal component of the
Coriolis force is to the right....

a velocity V so what does it do

that is east C right on the equator??

C
in the southern hemisphere the
horizontal component of the
Coriolis force is to the left....
 r
−2Ω× V = Coriolis
in the northern hemisphere the
 horizontal component of the
Coriolis force is to the right....

a velocity V on the equator, the

that is east C Coriolis force is
exactly vertical; there is
no horizontal component

C
in the southern hemisphere the
horizontal component of the
Coriolis force is to the left....
 sea level pressure, mb, at mid-latitudes
1) outside of the tropics,
winds are almost parallel
L to the lines of constant
pressure (isobars)
480
L 2) speeds are greatest
where the pressure gradient
is largest, i.e., where isobars
H are closer together

H 3) the wind is almost

in geostrophic balance:
H 840 1 P
H  fV  
∂η  x
L -fV = -g
∂x
L ∂η 1 P
fU ∂y
fU = -g

 y
from FNMOC, 'Fleet Numerical'
 r
The momentum equation in vector form:
∂VV r r r r
= −-VV ⋅∇
 V −-22Ω
Ω ×VV+F,F
∂t
t
where F is a force, a pressure gradient or wind stress (later)

The horizontal components are usually of most importance:

€ U U U
east:  -U -V  f V  Fx
t x y

V V V
north:  -U -V  f U  Fy
t x y

f is the Coriolis parameter,

f  2  sin(latitude),   7.29 x 10-5 sec -1
f vanishes on the equator and is maximum,  2, at the poles.
modes of the horizontal momentum eqns:

U U U
 -U -V  f V  Fx
t x y
V V V C
 -U -V  f U  Fy
t x y

V
local, time-dependent
inertial oscillations

the horizontal velocity vector will rotate

at the rate -f = -2sin(latitiude), which is
clockwise in the N hemisphere, c-clockwise in S hemisphere
inertial oscillations observed in ocean currents at
25 m depth following a hurricane

~ /f

data thanks to Zedler, et al., JGR 2003

Modes of the horizontal momentum equations :

U U U
 -U -V  f V - Fx
t x y
V V V
 -U -V  f U - Fy
t x y

steady, forced motion; C + F = 0

V C in this steady forced

balance, the velocity is
normal to the force, F,
and flows to the right
of F (N hemisphere)
F

when F is a pressure gradient, this is known as geostrophic balance

 Coriolis force: what am I supposed to know?
1) Coriolis force, C, is an inertial force that is a consequence
of Earth’s rotation. It is not 'ad hoc'
or 'fictitious' or 'virtual', as is sometimes implied.

2) the 3-d vector form: C = -2 x V.

3) the horizontal components are usually the important part:

Cx = fV; Cy = -fU, where U, V are east and north velocity
components and f is the Coriolis parameter,
f = sin(lat),  is Earth’s rotation rate.

4) the local, time-dependent mode of the horizontal

momentum equations is an inertial oscillation.

5) the steady, forced mode is geostrophic (or Ekman)

balance, both of which are very important.
end of class 17
 some review of the vector cross product

some examples follow, and here are the

links to two very good tutorials:
http://www.phy.syr.edu/courses/java-suite/crosspro.html
http://mathworld.wolfram.com/CrossProduct.html
 vector cross-product

magnitude:
Ia x bI = IaI IbI sin()
the signed area of a
parallelogram:

direction:
use the right hand rule
to find that the
the product vector is
perpendicular
to a and b
 vector cross product examples

torque = r X F
 F

torque = r X F, out of the page, counterclockwise

magnitude = IrIIFI
 F

torque = r X F, into the page, clockwise

magnitude = IrIIFI
 F

torque = r X F, into the page, clockwise

magnitude = ½ IrIIFI
 F

torque = r X F, into the page, clockwise

magnitude = IrIIFIsin () ~ 0.7 IrIIFI

 F

torque = r X F

magnitude = IrIIFIsin () ~ 0 (no torque)

 F

torque = r X F

magnitude = IrIIFIsin () ~ 0 (no torque)

backup slides follow
height (m) of the 925 mb pressure surface

1) winds are almost parallel

L to the lines of constant
480 height
L
L 2) speeds are greatest
where the isolines are
closest together

3) the wind seen here is

almost geostrophic:
HH
H
H 840 
H  fV   g
-fV = -g
∂η x
∂x
∂η 
L fU 
fU = -g
∂y  g
y

from FNMOC
height (m) of the 925 mb pressure surface
the slope is
L  100 m
 '
y 400 km
 480
L the acceleration is
L

g  2.5 x 103 m sec -2 .
y
in 1 hour, this would
cause a change of speed of
HH about 9 m sec -1
H 840 which does not happen;
H H evidently this slope-induced
∂η
-fV = -g
acceleration is nearly balanced by
∂x
∂η
L fU = -g
something else -
∂y

the Coriolis force = -2 x V

from FNMOC
 a brief diversion to geophysics............

eastward
westward

moving westward 1 mGal = 10-5 m sec-2

slowly

westward
faster eastward

on the equator, Eotvos acceleration = 15 mGal/(m/sec),

which is large compared to important geophsyical signals.