Basic Conservation Laws

Debasis Sengupta

Center for Atmospheric and Oceanic Sciences, Indian Institute of Science, Bangalore
dsen@iisc.ac.in

February 3, 2020

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 1 / 25

 Outline

1 Total Differentiation
Total Differentiation of a Vector in a Rotating System

2 The Vectorial form of the momentum equation in rotating coordinates

3 Component Equations in Spherical Coordinates

4 Scale analysis of the equations of motion

Geostrophic Approximation and Geostrophic Wind
Approximate Prognostic Equations; the Rossby Number
The Hydrostatic Approximation

5 The Continuity Equation

An Eulerian Derivation
A Lagrangian Derivation
Scale Analysis of the Continuity Equation

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 2 / 25

 Basic Conservation Laws

Atmospheric motions are governed by three fundamental physical

principles: conservation of mass, conservation of momentum, and
conservation of energy.
The laws may be derived by performing the budgets of mass,
momentum, and energy for an infinitesimal control volume in the
fluid.
Two types of control volume are commonly used in fluid dynamics.
In the Eulerian frame of reference the control volume consists of a
parallelepiped of sides δx , δy , δz, whose position is fixed relative to
the coordinate axes. Mass, momentum, and energy budgets will
depend on fluxes caused by the flow of fluid through the boundaries
of the control volume.
In the Lagrangian frame, the control volume consists of an
infinitesimal mass of “tagged" fluid particles; thus, the control volume
moves about following the motion of the fluid, always containing the
same fluid particles.
Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 3 / 25
 Basic Conservation Laws

The Lagrangian frame is particularly useful for deriving conservation

laws, as such laws may be stated most simply in terms of a particular
mass element of the fluid.
The Eulerian system is, however, more convenient for solving most
problems because in that system the field variables are related by a
set of partial differential equations in which the independent variables
are the coordinates x, y, z, and t.
In the Lagrangian system, however, it is necessary to follow the time
evolution of the fields for various individual fluid parcels. Thus the
independent variables are x0 , y0 , z0 , and t, where x0 , y0 , and z0
designate the position that a particular parcel passed through at a
reference time t0 .

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 4 / 25

 Total Differentiation

D
Dt is called the substantial, the total, or the material derivative.
∂
∂t is called local derivative.
To derive a relationship between the total derivative and the local
derivative, it is convenient to refer to a particular field variable
(temperature, for example). For a given air parcel the location
(x , y , z) is a function of t so that x = x (t), y = y (t), z = z(t).
Following the parcel, T may then be considered as truly a function
only of time, and its rate of change is just the total derivative DT
Dt .
In order to relate the total derivative to the local rate of change at a
fixed point, we consider the temperature measured on a balloon that
moves with the wind.

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 5 / 25

 Total Differentiation

Suppose that this temperature is T0 at the point x0 , y0 , z0 and time

t0 . If the balloon moves to the point x0 + δx , y0 + δy , z0 + δz in a
time increment δt, then the temperature change recorded on the
balloon, δT , can be expressed in a Taylor series expansion as :
∂T ∂T ∂T ∂T
δT = ( )δt + ( )δx + ( )δy + ( )δz + (higher order terms)
∂t ∂x ∂y ∂z
(1)
DT δT
≡ lim (2)
Dt δt→0 δt

DT ∂T ∂T Dx ∂T Dy ∂T Dz
=( )+( ) +( ) +( ) +(higher order terms)
Dt ∂t ∂x Dt ∂y Dt ∂z Dt
(3)
Dx Dy Dz
≡ u, ≡ v, ≡w (4)
Dt Dt Dt
u, v, w are the velocity components in the x, y, z directions,
respectively.
Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 6 / 25
 Total Differentiation

The expression reads as

DT ∂T ∂T ∂T ∂T
= +u +v +w (5)
Dt ∂t ∂x ∂y ∂z
∂T DT
= − U.∇T (6)
∂t Dt
where U = iu + jv + kw is the velocity vector.
−U.∇T is called the temperature advection. It gives the contribution
to the local temperature change due to air motion.
For example, if the wind is blowing from a cold region toward a warm
region −U.∇T will be negative (cold advection) and the advection
term will contribute negatively to the local temperature change.

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 7 / 25

 Total Differentiation
Question : The surface pressure decreases by 3 hPa per 180 km in the
eastward direction. A ship steaming eastward at 10 km/h measures a
pressure fall of 1 hPa per 3 h. What is the pressure change on an island
that the ship is passing ?
∂p Dp ∂p
= −u , (7)
∂t Dt ∂x
where Dp/Dt is the pressure change observed by the ship and u is the
velocity of the ship.
∂p 1 hpa km −3 hPa 1 hpa
=− − (10 )( )=− (8)
∂t 3h h 180 Km 6h
The rate of pressure fall on the island is only half the rate measured on the
moving ship.
If the total derivative of a field variable is zero, then that variable is a
conservative quantity and the local change is then due to advection
only.
Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 8 / 25
 Total Differentiation of a Vector in a Rotating System

The conservation law for momentum (Newton’s second law of

motion) relates the rate of change of the absolute momentum
following the motion in an inertial reference frame to the sum of the
forces acting on the fluid.
It is desirable to refer the motion to a reference frame rotating with
the earth.
To derive this relationship, let A be an arbitrary vector whose Cartesian
components in an inertial frame are given by :

A = i 0 A0x + j 0 A0y + k 0 A0z (9)

and whose components in a frame rotating with an angular velocity Ω are :

A = iAx + jAy + kAz (10)

Da A
Letting Dt be the total derivative of A in the inertial frame,

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 9 / 25

 Total Differentiation of a Vector in a Rotating System

Da A DA0x DA0y DA0z

= i0 + j0 + k0
Dt Dt Dt Dt (11)
DAx DAy DAz Da i Da j Da k
=i +j +k + Ax + Ay + Az
Dt Dt Dt Dt Dt Dt
The first three terms on the line above can be combined to give :
DA DAx DAy DAz
≡i +j +k (12)
Dt Dt Dt Dt
which is just the total derivative of A as viewed in the rotating coordinates
(i.e., the rate of change of A following the relative motion).

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 10 / 25

 Total Differentiation of a Vector in a Rotating System

Figure 2: Longitudinal dependence of the

Figure 1: Definition of a local Cartesian unit vector i.
framework of reference on a spherical earth.
The coordinate x is directed eastward, y
northward, and z upward.

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 11 / 25

 Total Differentiation of a Vector in a Rotating System

Figure 3: Resolution of δi into northward

and vertical components.

∂i ∂i ∂φ
δi = δλ + δφ + δz (13)
∂λ ∂φ ∂z
δi ∂i δλ
For solid body rotation, δλ = Ωδt, δφ = 0, δz = 0, so that δt = ∂λ ( δt )
and taking the limit δt → ∞.
Da i ∂i
=Ω (14)
Dt ∂λ
Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 12 / 25
 The longitudinal derivative of i can be expressed as :
∂i
= jsinφ − kcosφ (15)
∂λ
Ω = (0, Ωcosφ, Ωsinφ) (16)
Da i
= Ω(jsinφ − kcosφ) = Ω × i (17)
Dt
Da j Da k
In a similar fashion, it can be shown that Dt = Ω × j and Dt = Ω × k.
The total derivative for a vector in an inertial frame is related to that
in a rotating frame by the expression :
Da A DA
= +Ω×A (18)
Dt Dt

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 13 / 25

 The Vectorial form of the momentum equation in rotating
coordinates

In an inertial reference frame, Newton’s second law of motion may be

written symbolically as :
Da Ua X
= F (19)
Dt
In order to transform the expression to rotating coordinates, we must
first find a relationship between Ua and the velocity relative to
rotating system, which we will designate by U. This relationship is
obtained by applying it to the position vector r for an air parcel on
the rotating earth.
Da r Dr
= +Ω×r (20)
Dt Dt
Ua = U + Ω × r (21)
Similarly,
Da Ua DUa
= + Ω × Ua (22)
Dt Dt
Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 14 / 25
 The Vectorial form of the momentum equation in rotating
coordinates

Da Ua D
= (U + Ω × r ) + Ω × (U + Ω × r )
Dt Dt (23)
DU
= + 2Ω × U − Ω2 R
Dt

Ω is a constant. R is a vector perpendicular to the axis of rotation,

with magnitude equal to the distance to the axis of rotation, so that
with the aid of a vector identity.

Ω × (Ω × r ) = Ω × (Ω × R) = −Ω2 R (24)

DU 1
= −2Ω × U − ∇p + g + Fr (25)
Dt ρ
where Fr designates the frictional force, and the centrifugal force has
been combined with gravitation in the gravity term g.
Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 15 / 25
 Component Equations in Spherical Coordinates

Figure 5: Dependence of unit vector j on

latitude
Figure 4: Dependence of unit vector j on
longitude.

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 16 / 25

 Component Equations in Spherical Coordinates

Du uvtanφ uw 1 ∂p
− + =− + 2Ωvsinφ − 2Ωwcosφ + Frx (26)
Dt a a ρ ∂x
Dv u 2 tanφ vw 1 ∂p
+ + =− − 2Ωusinφ + Fry (27)
Dt a a ρ ∂y
Dw u2 + v 2 1 ∂p
− =− + 2Ωucosφ + Frz (28)
Dt a ρ ∂z
which are the eastward, northward, and vertical component momentum
equations.

The terms proportional to 1a are called curvature terms as they arise

due to the curvature of the earth.
Du ∂u ∂u ∂u ∂u
= +u +v +w (29)
Dt ∂t ∂x ∂y ∂z
With similar expressions for Dv/Dt and Dw/Dt.
Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 17 / 25
 Scale analysis of the equations of motion

Figure 6: Typical value of the variable in atmosphere

f0 = 2Ωsinφ0 = 2Ωcosφ0 ∼
= 10−4 s −1 , φ0 = 45◦ (30)

Figure 7: Scale Analysis of the Horizontal Momentum Equations

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 18 / 25

 Geostrophic Approximation and Geostrophic Wind

1 ∂p
−fv ≈ − ; (31)
ρ ∂x
1 ∂p
fu ≈ − (32)
ρ ∂y
where, f ≡ 2Ωsinφ is called the coriolis parameter.
1
Vg ≡ k × ∇p (33)
ρf

V g ≡ i ug + j vg , called the geostrophic wind.

The geostrophic relationship is called a diagnostic relationship.
The geostrophic wind approximates the true horizontal velocity to
within 10-15 % in midlatitudes.

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 19 / 25

 Approximate Prognostic Equations; the Rossby Number

Du 1 ∂p
= fv − = f (v − vg ) (34)
Dt ρ ∂x
Dv 1 ∂p
= −fu − = −f (u − ug ) (35)
Dt ρ ∂y

A convenient measure of the magnitude of the acceleration compared

to the Coriolis force may be obtained by forming the ratio of the
characteristic scales for the acceleration and the Coriolis force terms :
2
( UL )/(f0 L).
This ratio is a non-dimensional number called the Rossby number
after the Swedish meteorologist C. G. Rossby (1898–1957).
The smallness of the Rossby number is a measure of the validity of
the geostrophic approximation.

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 20 / 25

 The Hydrostatic Approximation

Figure 8: Scale Analysis of the Vertical Momentum Equation

1 dp0
≡ −g (36)
ρ0 dz

We define standard pressure p0 (z), which is the horizontally averaged

pressure at each height, and a corresponding standard density ρ0 (z),
defined so that p0 (z) and ρ0 (z) are in exact hydrostatic balance :

p(x , y , z, t) = p0 (z) + p 0 (x , y , z, t) (37)

0
ρ(x , y , z, t) = ρ0 (z) + ρ (x , y , z, t) (38)

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 21 / 25

 The Hydrostatic Approximation
where p 0 and ρ0 are deviations from the standard values of pressure
and density.
For an atmosphere at rest, p 0 and ρ0 would thus be zero.
0
Assuming that ρρ0 is much lesser than unity in magnitude so that
0
(ρ0 + ρ0 )−1 ∼
= ρ0 −1 (1 − ρ ), we find that
ρ0

1 ∂p 1 ∂
− −g =− 0
(p0 + p 0 ) − g
ρ ∂z (ρ0 + ρ ) ∂z
(39)
1 ρ0 dp0 ∂p 0 1 ∂p 0
≈ [ − ] = − [ρ0 g + ]
ρ0 ρ0 dz ∂z ρ0 ∂z
For synoptic scale motions, the terms have the magnitudes :
1 ∂p 0 δP ρ0 g
∼[ ] ∼ 10−1 ms −2 , ∼ 10−1 ms −2 (40)
ρ0 ∂z ρ0 H ρ0
∂p 0
+ ρ0 g = 0 (41)
∂z
Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 22 / 25
 An Eulerian Derivation

Figure 9: Mass inflow into a fixed (Eulerian) control volume due to motion parallel to
the x axis.

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 23 / 25

 A Lagrangian Derivation

Figure 10: Change in Lagrangian control volume (shown by shading) due to fluid
motion parallel to the x axis.

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 24 / 25

 Scale Analysis of the Continuity Equation

Debasis Sengupta (CAOS, IISc, Bangalore) February 3, 2020 25 / 25