Date post: | 04-Apr-2018 |
Category: |
Documents |
Upload: | krishnagdeshpande |
View: | 224 times |
Download: | 0 times |
of 58
7/30/2019 Ionosphere(v1.6.2)
1/58
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
EC3630 Radiowave Propagation
IONOSPHERIC WAVE PROPAGATION
by Professor David Jenn
EARTHS SURFACE
IONOSPHERE
(version 1.6.2)
7/30/2019 Ionosphere(v1.6.2)
2/58
1
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (1)
The ionosphere refers to the upper regions of the atmosphere (90 to 1000 km). This regionis highly ionized, that is, it has a high density of free electrons (negative charges) andpositively charged ions. The charges have several important effects on EM propagation:
1. Variations in the electron density ( eN ) cause waves to bend back towards Earth, but
only if specific frequency and angle criteria are satisfied. Some examples are shownbelow. Multiple skips are common thereby making global communication possible.
EARTHS SURFACE
1
2
3
4
TX
IONOSPHERE
SKIP DISTANCE
maxeN
7/30/2019 Ionosphere(v1.6.2)
3/58
2
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (2)2. The Earths magnetic field causes the ionosphere to behave like an anisotropic medium.
Wave propagation is characterized by two polarizations (denoted as ordinary andextraordinary waves). The propagation constants of the two waves are different. Anarbitrarily polarized wave can be decomposed into these two polarizations uponentering the ionosphere and recombined on exiting. The recombined wave polarizationangle will be different that the incident wave polarization angle. This effect is calledFaraday rotation.
The electron density distribution has the general characteristics shown on the next page.The detailed features vary with
location on Earth, time of day, time of year, and sunspot activity.
The regions around peaks in the density are referred to as layers. The F layer often splitsinto the 1F and 2F layers.
7/30/2019 Ionosphere(v1.6.2)
4/58
3
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Electron Density of the Ionosphere
(From Rohan. Note unit is per cubic centimeter.)
Day
Night
7/30/2019 Ionosphere(v1.6.2)
5/58
4
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
The Earths Magnetosphere
(From George Parks, Physics of Space Plasmas, Addison-Wesley)
Interaction of the sun andEarth magnetic fields
Interaction of the sun andEarth magnetic fields
7/30/2019 Ionosphere(v1.6.2)
6/58
5
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Barometric Law (1)Ionized layers are generated by solar radiation, mainly ultraviolet, which interacts withneutral molecules to produce electron-ion pairs. As a first step in predicting the ionizedlayers it is necessary to know the distribution of molecules with height.
A
Altitude
Column of
Atmosphere
dhh +
h p
dpp +
Consider a differential volume of air, as shown in the figure.The pressurep decreases with altitude, so dp is negative.There is a net buoyant force on the volume of cross sectionalareaA
AdpdppAApFb =+= )( .
This force must be balanced by the gravitational force of theweight of the gas in the volume
gAdhgdVFg ==
where is the mass density of the gas (mass per unit volume)and g is the gravitational constant. Ifm is the mean molecular
massgAdhmNgdVFg ==
Equating the last two equations gives gdhmNdp = . N is thenumber of molecules per unit volume.
7/30/2019 Ionosphere(v1.6.2)
7/58
6
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Barometric Law (2)Pressure and temperature are related by the ideal gas law: TBp Nk= (kB is Boltzmansconstant and T is temperature). Use the ideal gas law to substitute forNin dp:
TB
dp mgdhp k =
then integrate from starting height oh to a final height h
( )ln
( ) To
h
ho B
p h mgdh
p h k=
Finally, assuming that all of the quantities are approximately constant over the heights of
concern, and defining a new constant1
,TB
mg
H k= gives the barometric law for pressure vs.
height:
( ) ( )exp ( )exp ( )expT
o o
h ho
o o oh hB
h hmg dhp h p h dh p h p hk H H
= = =
His called the local scale height and thus the exponent, when normalized byH, is in Hunits. The reference height is arbitrary. The pressure, mass density, and number density
are seen to vary exponentially with height difference measured inHunits.
7/30/2019 Ionosphere(v1.6.2)
8/58
7
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Chapmans Theory (1)Chapman was the first to quantify the formation of ionized regions due to flux from thesun. In general we expect ion production to vary as follows:
1.Small at high altitudes solar flux is high, but the number of molecules availablefor ionization is small
2.Small at low altitudes the number of molecules available is large, but solar fluxis low due to attenuation at higher altitudes
3.High at intermediate altitudes sufficient number of molecules and fluxChapmans derivation proceeds along the following steps:
1.Find the amount of flux penetrating to an altitude h.2.By differentiation, find the decrease in flux with incremental height dh.3.The decrease in flux represents absorbed energy. From the result of step 2 an
expression for ion production per unit volume at height h is derived.4.The barometric law is used to find the optical depth at height h (optical depth is a
measure of the opacity of the atmosphere above h).5.The maximum height of electron-ion production is set as the reference height.6.Assuming equilibrium (no vertical wind) laws of the variation of the electron
density are derived based on the rates of electron production and loss.7.A scaling law is developed so that ionization can be represented by a single
variable rather than two (altitude and solar zenith angle).
7/30/2019 Ionosphere(v1.6.2)
9/58
8
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Chapmans Theory (2)Only the final results are presented here. Sdenotes the solar flux and is the zenith angle.
The electron density at height h and zenith angle is given by
( ) = zoe ezNhN sec1
21exp),(
where Hhhz o /)( = and oN is the electrondensity at the reference height oh . This equation
specifies a recombination type of layer (as
opposed to an attachment type layer1). It can berewritten (by a substitution of variables andrescaling) as
)secln,0(cos),( = zNzN ee
The value for any can be obtained from theo0= curve.
A
Column ofAtmosphere
dhh +
h
dSS+
S
1There are several reactions which can remove electrons from the ionosphere. The two most important classes of recombination are (1)the electron attaches itself to a positive ion to form a neutral molecule, and (2) it attaches itself to a neutral molecule to form a negativeion.
7/30/2019 Ionosphere(v1.6.2)
10/58
9
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Chapmans Theory (3)
It is found that the E and F1 regions behave closely to what is predicted by Chapmansequation.
7/30/2019 Ionosphere(v1.6.2)
11/58
10
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Dielectric Constant of a Plasma (1)
When an electric field is applied, charges in the ionosphere are accelerated. The mass ofthe ions (positive charges) is much greater than the electrons, and therefore the motion ofthe ions can be neglected in comparison with that of the electrons. The polarization vector
isdeNpNP eerrr
==
where =eN electron density3
m/ ,19
1059.1 =e C, electron charge, and dr
is the
average displacement vector between the positive and negative charges. The relative
dielectric constant is
E
P
E
PE
E
D
oo
o
or r
r
r
rr
r
r
+=
+== 1
The propagation constant is
njkc
jjk orc ===
where rn = is the index of refraction. Both the dielectric constant and index of
refraction can be complex ( rrr j = and njnn = ).
7/30/2019 Ionosphere(v1.6.2)
12/58
11
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Dielectric Constant of a Plasma (2)
Assume ax polarized electric field. The electrons must follow the electric field, and
therefore, xxd =r
. The equation of motion of an electron istj
xeeExmxm =+
&&& where 31100.9 =m kg, electron mass, and = collision frequency. The solution of the
equation of motion is of the form tjaex = . Substituting this back into the equation givestj
xtjtj
eeEaejmeam =+ )()( 2
with( ) /12 jm
eEa x
=. The polarization vector is
( ) /1
2
2
jm
xeEeNP
tjxe
=
r
Now define a plasma frequency,o
ep
m
eN
2= and constants
2
=
pX and
=Z so
that
( )jZ
XxeEP tjxo
=
1
r
.
7/30/2019 Ionosphere(v1.6.2)
13/58
12
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Dielectric Constant of a Plasma (3)
The ratio of Pr
to Er
gives the complex dielectric constant, which is equal to the square ofthe index of refraction
( ) ( )
jjZXnj prrr
=
===
2
2 11
1
Separating into real and imaginary terms gives an equivalent conductivity
orr j
= where
( )222
1
+
=m
eN
o
er and ( )22
2
+=
m
eNe
For the special case of no collisions, 0= , and the corresponding propagation constant
is real
2
2
1
pooroc kk ==
with oook = .
7/30/2019 Ionosphere(v1.6.2)
14/58
13
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Dielectric Constant of a Plasma (4)
Consider three cases:
1. p > : ck is real andzkjzjk cc ee
= is a propagating wave
2. p< : ck is imaginary and zkzjk cc ee = is an evanescent wave3. p = : 0=ck and this value of is called the critical frequency
1, c
At the critical frequency the wave is reflected. Note that c depends on altitude because
the electron density is a function of altitude. For electrons, the highest frequency at
EARTHS
SURFACE
TX
IONOSPHERE h
REFLECTION
POINT0== rc
which a reflection occurs is
max92
ec
c Nf =
Reflection at normal incidence requires
the greatest eN .
1The critical frequency is where the propagation constant iszero. Neglecting the Earths magnetic field, this occurs atthe plasma frequency, and hence the two terms are oftenused interchangeably.
7/30/2019 Ionosphere(v1.6.2)
15/58
14
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (1)
If the waves frequency and angle of incidence on the ionosphere are chosen correctly,the wave will curve back to the surface, allowing for very long distance communication.For oblique incidence, at a point of the ionosphere where the critical frequency is cf ,
the ionosphere can reflect waves of higher frequencies than the critical one. When thewave is incident from a non-normal direction, the reflection appears to occur at a virtualreflection point, h, that depends on the frequency and angle of incidence.
EARTHSSURFACE
TX
IONOSPHERE
h
VIRTUAL
HEIGHT
SKIP DISTANCE
7/30/2019 Ionosphere(v1.6.2)
16/58
15
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (2)To predict the bending of the ray we use a layered approximation to the ionosphere just aswe did for the troposphere.
1z2
z3z
1
i
2
3
A
LTITUDE
LAYERED
IONOSPHERE
APPROXIMATION
)( 1zr
)( 2zr
)( 3zr
M
1=r
Snells law applies at each layer boundary( ) L== )(sinsin 11 zri
The ray is turned back when 2/)( =z , or )(sin zri =
7/30/2019 Ionosphere(v1.6.2)
17/58
16
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (3)
Note that:
1. For constant i , a higher eN is required for higher frequencies if the ray is to return
to Earth (because r decreases with ).2. Similarly, for a given maximum eN ( maxeN ), the maximum value of i that results
in the ray returning to Earth increases with increasing .
There is an upper limit on frequency that will result in the wave being returned back toEarth. Given
maxe
N the required relationship betweeni
andfcan be obtained
i
eie
ei
pi
ri
Nf
fN
f
N
z
2max
max
22
max
2max2
2
22
cos
81
81
cos
811cos1
1sin
)(sin
==
=
=
=
7/30/2019 Ionosphere(v1.6.2)
18/58
17
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (4)
Examples:
1. o45=i ,310
max m/102 =eN : 8.1)707.0/()102)(81(210
max ==f MHz
2. o60=i , 310max m/102 =eN : 5.2)5.0/()102)(81( 210max ==f MHz
The value offthat makes 0=r for a given value of maxeN is the critical frequencydefined earlier:
max9 ec Nf =
Use the maxeN expression from the previous page and solve forf
icie fNf secsec9 max ==
This is called the secant law or Martyns law. When isec has its maximum value
(maximum angle of incidence on the ionosphere), the frequency is called the maximum
usable frequency (MUF). A typical value is less than 40 MHz. It can drop as low as 25MHz during periods of low solar activity. The optimum usable frequency1 (OUF) is 50%to 80% of the MUF. Data for the MUF is available. This is only correct for horizontallystratified media.
1Sometimes called the optimum working frequency (OWF).
7/30/2019 Ionosphere(v1.6.2)
19/58
18
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Maximum Usable Frequency
Shown below is the MUF in wintertime for different skip distances. The MUF is lower inthe summertime.
Fig. 6.43 in R. E. Collin,Antennas and Radiowave Propagation, McGraw-Hill, 1985
7/30/2019 Ionosphere(v1.6.2)
20/58
19
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (5)
Multiple hops allow for very long range communication links (transcontinental). Using asimple flat Earth model, the virtual height (h), incidence angle ( i ), and skip distance
(d) are related by h
di = 2tan . This implies that the wave is launched well above the
horizon. However, if a spherical Earth model is used and the wave is launched on the
horizon then hRd e = 221.
EFFECTIVE SPECULAR
REFLECTION POINT
IONOSPHERE
i
h
d Single ionospheric hop
(flat Earth)
EARTHS
SURFACE
TX
IONOSPHERE
Multiple ionospheric hops
(curved Earth)
1Whether to use Re or Re depends on the amount of path through the troposphere where refraction is significant. In the remaining
notes Re is used.
7/30/2019 Ionosphere(v1.6.2)
21/58
20
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (6)
Approximate virtual heights for layers of the ionosphere
Layer Range for h (km)
2F 250 to 400 (day)
1F 200 to 250 (day)
F 300 (night)E 110
Example: Based on geometry, a rule of thumb for the maximum incidence angle on the
ionosphere is about o74 . The MUF is
cc ff 6.3)74sec(MUF ==o
For
312
max m/10=eN , 9cf MHz and the 4.32MUF = MHz. For reflection from the 2Flayer, 300h km. The maximum skip distance will be about
4516)10300)(108500(222233
max == hRd e km
7/30/2019 Ionosphere(v1.6.2)
22/58
21
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (7)
For a curved Earth, using the law of sines for a trianglei
eRh
tan
1
sin
cos/1=
+
R/2
h
d/ 2
eR
i
LAUNCH ANGLE:
90o
i = 90o
=
R/2
where
eR
d
=
2
and the launch angle (antennapointing angle above the horizon)
is
i ==oo 9090
The great circle path via thereflection point isR, which can be
obtained from
i
eRR
sin
sin2 =
7/30/2019 Ionosphere(v1.6.2)
23/58
22
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (8)
Example: Ohio to Europe skip (4200 miles = 6760 km). Can it be done in one hop?
To estimate the hop, assume that the antenna is pointed on the horizon. The virtual height
required for the total distance is
( ) 3976.02/2/ === ee RdRd rad = 22.8 degrees720/coscos)( ===+ eeee RRhRhR km
This is above the F layer and therefore two skips must be used. Each skip will be half of
the total distance. Repeating the calculation for 16902/ =d km gives( ) 1988.02/ == eRd rad = 11.39 degrees
711/cos == ee RRh km
This value lies somewhere in the F layer. We will use 300 km (a more typical value) in
computing the launch angle. That is, still keep 16902/ =d km and = 11.39 degrees, butpoint the antenna above the horizon to the virtual reflection point at 300 km
ooo 4.74)39.11cos(8500
3001)39.11sin(tan
1
=
+=
ii
7/30/2019 Ionosphere(v1.6.2)
24/58
23
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (9)
The actual launch angle required (the angle that the antenna beam should be pointed abovethe horizon) is
launch angle, = ooooo 21.44.7439.119090 == i
The electron density at this height (see chart, p.3) is 311max m/105eN whichcorresponds to the critical frequency
= max9 ec Nf 6.36 MHzand a MUF of
7.234.74sec36.6MUF = o MHz
Operation in the international short wave 16-m band would work. This example is
oversimplified in that more detailed knowledge of the state of the ionosphere would benecessary: time of day, time of year, time within the solar cycle, etc. These data areavailable from published charts.
7/30/2019 Ionosphere(v1.6.2)
25/58
24
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Ionospheric Radiowave Propagation (10)
Generally, to predict the received signal a modified Friis equation is used:
( )
LL
R
GGPP x
rttr
2/4
=
where the losses, in dB, are negative:
ionoreflpol GLLLx +=
=reflL reflection loss if there are multiple hops
=polL polarization loss due to Faraday rotation and earth reflections=ionoG gain due to focussing by the curvature of the ionosphere=L absorption loss
R = great circle path via the virtual reflection point
Example: For 30=tP dBW,f= 10 MHz, 10== rt GG dB, d= 2000 km, 300=h km,5.9=xL dB and 30=L dB (data obtained from charts).
From geometry compute: o3.70=i ,R = 2117.8 km, and thus 5.108=rP dBw
7/30/2019 Ionosphere(v1.6.2)
26/58
25
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Charges in Magnetic Fields (1)
The motion of charged particles in a magnetic field can be derived from the Lorentzequation. For a charge q of mass m moving at velocity u
m
duF qu B m
dt= =
rr rr
Decomposing ur
into longitudinal and transverse components with respect to Br
gives
( ) ( )L T L Td
q u u B m u udt
+ = +rr r r r
Let the magnetic field be uniform and in thez direction, so thatoB B z=
r, L zu u z=
r, and T x yu u x u y= +
r
and therefore the vector equation can be written as two scalar equations
/x
y o
duqu B m
dt= , /
y
x o
duqu B m
dt=
The classical gyrofrequency is / /H o o oqB m q H m = = . To solve the two equations definea new variable x yu ju = +
0Hd
jdt
+ =
LurT
ur u
r
BrLur
Tur u
r
Br
7/30/2019 Ionosphere(v1.6.2)
27/58
26
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Charges in Magnetic Fields (2)
The solution is
( ) ( )cos sinHT T H H j t
u e u t j t
= = + r
Therefore,( )
( )
cos
sin
x T H
y T H
z L
dxu u t
dt
dyu u t
dt
dzu udt
= =
= =
= =
Integrating with respect to time
( )
( )
sin
cos
To H
H
To H
H
o L
ux x t
uy y t
z z u t
= +
= +
=
7/30/2019 Ionosphere(v1.6.2)
28/58
27
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Charges in Magnetic Fields (3)
The Larmor radius is defined as /T Hcr u = . The previous equations describe a circle ofradius cr with the constants ox , oy , oz and that depend on the initial conditions. Note that
the particle motion has the following characteristics:
1.The particle rotates around the magnetic field line in a circle of radius cr at frequency
H
2.The center of the circle (the particles guiding center) travels along the field line withvelocity Lu
rso that the particle traces out a helix, as shown in the figure
oB B z=r
Lur
Tur
ur
Guiding
Centercrr
rr
O
x
y
BroB B z=
r
Lur
Tur
ur
Guiding
Centercrr
rr
O
x
y
oB B z=r
Lur
Tur
ur
Guiding
Centercrr
rr
O
x
y
Br
Br
N l P d S h l D f El i l & C E i i M C lif i
7/30/2019 Ionosphere(v1.6.2)
29/58
28
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Charges in Magnetic Fields (4)
The Earths magnetic field is inhomogeneous. The magnetic field is strongest at the poleswhere the flux lines converge. The force on the charge varies as the gradient of the field
m
du
F M B m dt= =
rr r r
2
2
cq r BMm
=r
is the magnitude of the magnetic dipole moment. If momentum (energy) is
conserved along the path,1 2 2
2( ) constantL Hm u u+ = . Conservation of momentum implies:
Diverging flux lines
1.As Br increases H increases and Tu increases
2.As Br
increases cr decreases and Tu increases
The reverse is true when Br
decreases. Thus, Tu
increases as the flux lines converge, which requires thatLu decrease so that momentum is conserved. At some
point all energy is in the transverse component, andlongitudinal motion stops. This is a turning point wherethe particle reverses direction.
N l P t d t S h l D t t f El t i l & C t E i i M t C lif i
7/30/2019 Ionosphere(v1.6.2)
30/58
29
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Earth Magnetic Field (1)
The Earths magnetic field is comprised of three main components:1.The main field due to the outer core. The molten core consists of rotating liquid metal
that give rise to a dynamo effect.
2.Fields that arise in the Earths crust and upper mantle.3.Atmospheric magnetic field
due to currents flowing in theatmosphere and in the oceans.
The first component dominatesand is approximately a dipolefield with the magnetic dipolepointing south.
Typical values:
At poles, 40.6 to 0.7 10 TB = r
Near equator: 40.25 10 TB = r
2 4(1 Wb/m = 1 T =10 G)
From: www.ngdc.noaa.gov/seg/WMM/back.shtml
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey California
7/30/2019 Ionosphere(v1.6.2)
31/58
30
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Earth Magnetic Field (2)
The Earths main magnetic field reverses on the time scale of every several hundredthousand years. It is believed that the main field is currently in the process of reversal. Theevidence includes the weakening of the magnetic field, relatively rapid migration of the
poles (below left) and an increase in anomalies. Below is a simulation result showing theEarths magnetic field during the transition period.
Simulation of the Earths magnetic field, G.A. Glatzmaier and P.H. Roberts, "A three-dimensional convective dynamo solution with rotating and finitely conducting innercore and mantle," Phys. Earth Planet. Inter., 91, 63-75 (1995).
(http://science.nasa.gov/headlines/y2003/29dec_magneticfield.htm?list770900)
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey California
7/30/2019 Ionosphere(v1.6.2)
32/58
31
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Magneto-ionic Medium (1)
Thus far the Earths magnetic field has been ignored. An electron with velocity ur
moving
in a medium with a magnetic field ooo HBrr
= experiences a force om BueFrrr
= .
Assume a plane wave that is propagating in thez direction ( zeHE ~,rr
). From
Maxwells equations
[ ]( )
( )( )
=+=
+=
+=
+==
)3(0
)2(/
)1(/
zzzo
yyoyx
xxoxy
o
DjPEj
EPEjH
EPEjH
PEjDjH
rrrr
=
=
=
=
)6(0
)5(
)4(
zo
yox
xoy
o
Hj
HjE
HjE
HjE
rr
These equations show that the plane wave will be transverse only with respect to the B
r
, D
r
and Hr vectors, which is different from the isotropic case. The characteristic impedance ofthe medium is
/// oxyyx jHEHE === .
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
33/58
32
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
Magneto-ionic Medium (2)
Solve (5) for yH and use the result in (1):
43421
2
122
n
E
P
xo
xoo
+=
. A similar process
starting with equation (4) would lead toyo
y
E
Pn
+= 12 . Equating this to the expression for
2n above gives the polarization ratio of the wave,
y
x
y
x
E
E
P
PR =pol .
Rewrite the equation of motion for an average electron when both electric and magneticfields are present
( )oBrEermrmr
&rr
&r&&r +=+
There is also a term in the parenthesis for the magnetic field of the wave, but it turns out tobe negligible. We now define the longitudinal and transverse components, which areparallel and perpendicular to the direction of propagation, respectively.
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
34/58
33
g p f p g g y, f
Magneto-ionic Medium (3)
Definition of the coordinate system:y lies in the plane defined by )( ooo HB = andz
)transverse(sin
)allongitudin(cos
oyT
ozL
HHH
HHH
==
==
Note the direction ofx is chosen to forma right-handed system (RHS). Anycombination of propagation direction
and magnetic field can be handled withthis convention, as long as they are notparallel. If they are parallel anycombination ofx andy that forms aRHS is acceptable.
x
y
Direction of
propagation
oBr
Lies in the plane
defined byz andBoEarths magneticfield vector
LBr
TB
r
Determinedfromz andy
z
In this coordinate system the equation of motion can be written as three scalar equations,which for thex component is
( )ToLox HzHyEexmxm &&&&& +=+
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
35/58
34
g p f p g g y f
Magneto-ionic Medium (4)
For time-harmonic displacements, phasors can be used and the derivatives in time give amultiplicative factor j
( )ToLox HzjHyjEemxjmx +=+2
The polarization is caused by the electron displacements inx,y, andz
e
z
e
y
e
x
eN
Pz
eN
Py
eN
Px
=
=
= ,,
Substituting:e
Tzo
e
Lyox
e
x
e
x
eNHPj
eNHPjeE
eNmPj
eNmP +=
2
Multiply by mo/ and define the longitudinal and transverse gyro-frequencies1:
cos/ BLoL meH =
sin/ BToT meH =
1In general, the gyro-frequency (also called the cyclotron frequency) of an electron in a magnetic field is defined as meB
oB/ or
mHeooH
/ .
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
36/58
35
Magneto-ionic Medium (5)
For convenience we define the frequency ratios2
=
pX ,
LLY = ,
TTY = , and
=Z .
Recall that the plasma frequency iso
ep
m
eN
2
= . With this notation
zTyLxxo PjYPjYjZPXE += )1(
Going back to they andz components of the equation of motion, and treating them thesame, gives two more equations
xTzzo
xLyyo
PjYjZPXE
PjYjZPXE
=
+=
)1(
)1(
These are the equations needed to find n in terms of the components of Er , and Pr .Arbitrarily we choose to solve for yP and yE , eliminating zzxx EPEP ,,, .
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
37/58
36
Magneto-ionic Medium (6)
With several substitutions and a fair amount of algebra, two equations are obtained:
2polpolpol
pol2
polpol
)1(
)1/()1(
RPjYjZRPEXR
jZXPRYPjYjZRPEXR
yLyyo
yTyLyyo
+=
+=
Equate the right hand sides: 2polpol2 )1/( RjYjZXRYjY LTL =+
Solving this quadratic equation gives:
+
=
2/12
2
42
pol)1(4)1(2
LTT
L
YjZX
Y
jZX
Y
Y
jR m
Not all plane waves can pass through the medium. To remain plane waves they must haveone of two polarizations that are characteristic of the medium. If this equation is satisfied,both equations at the top of the page give the same solution:
pol1/
RjYjZ
XEP
L
oyy
=
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
38/58
37
Magneto-ionic Medium (7)
Use this ratio to find the index of refraction: [ ]pol2 1/1 RjYjZXn L= and with polR from the quadratic equation yields the Appleton-Hartree formula
2/1
2
2
42
22
)1(4)1(21
1)(
+
===
LTT
r
YjZX
Y
jZX
YjZ
Xnjnn
Physical interpretation: The Earths magnetic field causes the ionosphere to be anisotropic.
There are two modes of propagation, each with a particular polarization, that dependsentirely on the properties of the medium. The phase velocities of the two modes, from thetwo values ofn above, are different, and when they recombine they have different phaserelationships. The solution with the positive sign is called the ordinary wave; that with thenegative sign is the extraordinary wave. When the frequency is greater than about 1 MHz,
which is always the case for systems that transmit through the ionosphere, the wave can beconsidered solely longitudinal ( TH , T and TY are zero) and therefore,
)(1
22
jn
L
p
is sufficiently accurate for most calculations.
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
39/58
38
Anisotropic Media
In an isotropic medium the phase velocity at any given point is independent of thedirection of propagation. For an isotropic medium, the direction of phase propagation (i.e.,movement of the equiphase planes) is radial with velocity pu . The energy propagation,
which is generally taken as the group velocity,1 gu is also radial. In an anisotropic medium
the directions of the equiphase planes and energy are different.
Isotropic Medium Anisotropic Medium
1This is the velocity of propagation of a packet of frequencies centered about a carrier frequency. For simple amplitude modulated
(AM) waveforms, it is the velocity of propagation of the envelope, whereas the phase velocity is the velocity of propagation of thecarrier.
Noncircular
Wavefront
Pr
Point
Source
Energy
Direction
(Radial)
Wavefront
Normal
oBr
CircularWavefront
P
r
Point
Source
Noncircular
Wavefront
Pr
Point
Source
Energy
Direction
(Radial)
Wavefront
Normal
oBr
oBr
CircularWavefront
P
r
Point
Source
CircularWavefront
P
r
Point
Source
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
40/58
39
Phase and Group Velocities
The phase velocity is/ /pu c n = =
and the phase index of refraction (our normal index of refraction) is related to it by
pn u c = . A group index of refraction, gn , can be defined analogous to the phase index of
refraction:
g gn u c=
where /gu d d = . It is always true that2
cuu gp = . The group index can be expressed
as:
( )
2/
2
2
g g
p
g
d d cn c u c
d d
n ud c
d c fd
nd
dnn n f
df
= = =
=
=
= +
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
41/58
40
Total Electron Content
If we neglect the magnetic field and consider the case where 0Z= and 1X
7/30/2019 Ionosphere(v1.6.2)
42/58
41
Phase Delay and Time Delay (1)
Let the distance between point P1 and P2 be l . As a check, for a homogeneous medium,
r on nc c
= = = = l l l l
A typical value for a vertical path at 100 MHz is 217 melectrons/103 , which gives
( )( )
( )( )( )( )
217 19
31 12 8 6
3 10 1.6 1012526 rad
4 9 10 8.85 10 3 10 100 10
The time delay between a transmitter at P1 and receiver at P2 is:
=2
1)(
11 P
Pd ds
snct
With the same assumptions onZandX, the (group) time delay is
2
1
2
1 1 40.31
2
P
P
Td
Nt X ds
c c cf
+ = +
l
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
43/58
42
Phase Delay and Time Delay (2)
The phase rotation of the ordinary wave is found to be2
o2 4 2
2 4
To
oT T L
N ek
m cY Y Y
= +
+
l
A handy identity is2
51.06 10o
e
m c
= . For the extraordinary wave
++
+=
cm
e
YYY
Nk
oLTT
To
2
242
x
42
l
Faraday rotation is the rotation of the phase angle of a linearly polarized plane wave:
( )4 22 2
4 2o x 2 2
41 1 14
2 4 41
T LT T T L
o oT L
Y Ye eN N Y Y
m c m cY Y
+ = = +
where the approximation assumes that the wave frequency is much larger than the gyrofrequency. With a magnetic field present the equations are more complex.
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
44/58
43
Faraday Rotation Example
Find the Faraday rotation of a 135 MHz plane wave propagating a distance of 200 kmthrough the ionosphere. The electron density is constant over the path and equal to
10105 /m3. The gyro-frequency is 6108 Hz. Assume collisions can be neglected and
that the propagation is longitudinal.
The TEC is ( ) 161030
1010510200)( == == eeT NdssNN ll
/m2. For longitudinal
propagation and no collisions 0==ZYT and /BLY =
( )( )( )( )( )
radians75.3
1013522
1082101006.1
24
1
26
6165
2
BT
o
Ncm
e
x xy y
inEr
outEr
z
200 km
3.75 rad
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
45/58
44
GPS Error Due to Ionospheric Time Delay
An increase in phase velocity is accompanied by a decrease in group velocity:
( )
( )
1
2
12
11
21
p
g
u c X
u c X
u c X
=
== +
Differential relationships
1 1 11 1
2 2ds c X dt dt X ds
c
= =
m
The path length error due to ionospheric time delay error is
{
2 2 2
1 1 1
1 1 1
22 2 2
1
40.31( )
t t P
Tt t P
u dt cX dt X X ds N f
7/30/2019 Ionosphere(v1.6.2)
46/58
45
GPS Example
GPS uses two frequencies (L1 and L2 in the table on the previous page). If one canestimate the TEC then it is possible to correct the position errors due to the ionospherictime delay. Given that the one-way time delay for a path from the ground to the satellite
differs by 15 ns for the two frequencies, estimate the total electron contect of the path.
The change in time delay due to the ionosphere is given by the term
2
40.3 Td
Nt
cf
For the two frequencies:
( )( )
2 1
9
2 22 1
19 917 3
2 2 7 192 1
40.3 1 115 10
15 10 1 1 15 104.2 10 / m
40.3 1.34 10 2.6 10
Td d
T
Nt t
c f f
cN
f f
= = =
Applying this correction in practice is difficult because of other ambiguities that exist.
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
47/58
46
Absorption
Once the index of refraction is determined for the medium, the attenuation constant can beobtained from
n
c
jj
=+=
For simplicity, the effect of the magnetic field is ignored in this discussion ( 0== LT YY )so that
+
+==
22
22
111)(
Z
XZj
Z
Xnjnn
yielding2
12
1
Z
XZ
ncn
c +== . In general, is not constant along a path through the
ionosphere, because the electron density, and henceXand n are changing along the path.The collision frequency, if not zero, can also change with location in the ionosphere.Thus, the loss due to absorption from 1P to 2P should be computed from
=2
1
)(expattn
P
P
dssL
where s is the distance along the path.
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
48/58
47
Example
Find the attenuation (in dB/km) in the ionosphere forf = 30 MHz, 710= /s and810=eN /m
3.
Using the formula from the previous page with 0531.010302
106
7
=
=
=
Z
and 62
6
82
109.810302
1034.56 =
=
=
pX ,
( )11
22
2
+
=
m
eN
o
er . Substituting
gives:
=+
= 72
1048.112
1
Z
XZ
ec r
0013.0)log(20
1000 = e dB/km.
Alternatively, we could use the basic equation for attenuation
+= 11
2
2
ro
roo
where( )22
2
+
=m
eNe .
For 1110=eN the attenuation increases to 1.3 dB/km.
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
49/58
48
Ionospheric Soundings and Ionograms
An ionosonde is an instrument used to measure the variation of critical frequency withheight. The measurements are used to determine ( )eN h . A train of short RF pulses istransmitted vertically and the time delay to reception is measured. The range to the
reflection point can be computed. The frequency of the pulse train vf is swept to obtain aplot ofh versusf, which is called an ionogram. The three segments of the curves betweenspikes represent reflections from the E, F1, and F2 layers.
From: www.ukssdc.ac.uk/ionosondes/ionogram_interpretation.html
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
50/58
49
Transmission Curves (1)
Two paths through the ionosphere are shown below, one for vertical incidence at
frequencyfv and the other for oblique incidence at frequencyfob. We know that
c
f f< , a reflection always occurs;
cf f> , reflection only if cos i cf f < .
From the Secant law, / secob v if f = . For a flat earth, tan /(2 )i d h = , and since
[ ]22 2sec 1 tan 1 /(2 )i i d h = + = + , we can solve for h:
[ ]
[ ]
( )
2
2
2
sec 1 /(2 )
/ 1 /(2 )
2 / 1
i
ob v
ob v
d h
f f d h
dh
f f
= +
= +
=
i
vf f=
obf f=
h
/ 2d / 2d
i
vf f=
obf f=
h
/ 2d / 2d
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
51/58
50
Transmission Curves (2)
Plot a family of curves ofh versus vf with obf treated as a parameter. Then overlaythem on an ionogram (heavy line in the figure). Together they determine what paths asignal radiated at frequency obf may take over distance d(neglecting loss and magnetic
effects). They are called transmission curves.Intersections of transmission curveswith the ionogram represent validreflection heights for the given
frequency.
In this example, at frequenciesabove about 13 MHz there are nopoints common to the transmissioncurves and ionogram. This is the
MUF.
(Curves shown are for d= 800 km,after K. Davies)
8 MHz 10 12 14
E
F1
F2
13
8 MHz 10 12 14
E
F1
F2
8 MHz 10 12 14
E
F1
F2
13
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
7/30/2019 Ionosphere(v1.6.2)
52/58
51
Time Delay Contours
Nightside transionospheric time delay at 30MHz for receiver on a LEO platform at 1000km.
Upper left: Receiver field-of-view showingtime delay for ground-based transmitterswhose signals range in elevation angle from90 (zenith propagation) to 5 (definesboundary of circular area).
Upper right: GAIM vertical TEC, orbit path,and receiver location on nightside.Lower left: Bisecting cut through image inupper left from E to W.Lower right: Corresponding bisecting cut fromS to N.
(http://www.cpi.com/splash/signalprop.html)
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
I h i R B di
7/30/2019 Ionosphere(v1.6.2)
53/58
52
Ionospheric Ray Bending
The paths (red curves) taken by signalsthrough two ionospheres under daytimeconditions with different altitude
distributions of the ionization. Theionospheres are displayed in color and thefrequency and elevation angle consideredfor the illustration are 30 MHz and 5,respectively. The blue curves show thecorresponding straight-line paths that
appear with curvature due to the use of aflat-Earth projection for the displays. Thetotal electron content along the straight-line paths (TECSL) is the same for bothcases (100 TEC units). The displaysillustrate different degrees of ray bending
due to different distributions of theionization along the ray paths butconstrained to the same TECSL values.
(http://www.cpi.com/splash/signalprop.html)
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
S i ill i (1)
7/30/2019 Ionosphere(v1.6.2)
54/58
53
Scintillation (1)
Scintillation refers to rapid fluctuations in amplitude, phase, or angle of arrival of waves. Itis produced by small-scale refractive index variations in the troposphere and ionosphere. Anoptical effect of scintillation is the twinkling of stars.
Tropospheric scintillation is produced in the first few kilomenters of the surface and dependson season, local climate, frequency, and elevation angle. Typical peak-to-peak losses are:
Loss (dB) Conditions
1 Clear, summer0.25 to 1 Winter
2 to 6 Some clouds
The fading rate ranges from 0.5 Hz to 10 Hz. Scintillation increases with frequency and is
worst at low elevations (
7/30/2019 Ionosphere(v1.6.2)
55/58
54
Scintillation (2)
A simple model to explain scintillation is the dielectric layers or ellipsoids model.
Plane input
wavefront
Distorted output
wavefront
Time varying inhomogeneous
dielectric region
Plane input
wavefront
Distorted output
wavefront
Time varying inhomogeneous
dielectric region
Ellipsoids with slightly different relativedielectric constants are changing shape as afunction of time. The output wavefront isdistorted due to atomospheric multipath.
Wavefront distortion gives rise to phase
errors and angle of arrival errors.
Multipath causes signal fading.
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
S i till ti (3)
7/30/2019 Ionosphere(v1.6.2)
56/58
55
Scintillation (3)
Ionospheric scintillation is due to localized changes in the electron density eN , and hence
r . The effect decreases as21/f and usually can be neglected above 4 GHz, except when
severe magnetic storms are occurring. Characteristics are:
3.At 4 GHz causes 0.5 to 10 dB fading 0.1 % of the time
4.Worst at the poles and in an equatorial belt 25 o from the equator5.Worst 1 2 hours after sunset6.Independent of elevation angle
Ionospheric irregularities can take on a wide range of sizes but are typically around 1 km inextent. They tend to align with the Earths magnetic field lines and form striations orplumes.
Br
Plumes
Electron density
irregularities
Earth
Br
Field aligned striations
Br
Plumes
Electron density
irregularities
Earth
Br
Plumes
Electron density
irregularitiesBr
Plumes
Electron density
irregularities
Earth
Br
Field aligned striations
Br
Field aligned striations
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
S i till ti I d
7/30/2019 Ionosphere(v1.6.2)
57/58
56
Scintillation Index
A measure of scintillation isthe scintillation index
22
4 2P PS
P
=
P = signal power
P = average power
(usually over 1 minute)
Significance:
4 0.6S < : weak
4 0.6S > : strong
4 1.0S = : saturation
http://gps.ece.cornell.edu/scintplot.html
Current information available atNational Weather Service Space Weather Prediction Centerhttp://www.swpc.noaa.gov/index.html
Naval Postgraduate School Department of Electrical & Computer Engineering Monterey, California
I h i Di t b
7/30/2019 Ionosphere(v1.6.2)
58/58
57
Ionospheric Disturbances
Sudden disturbances are usually due to solar flares, with the following effects:
1.phase anomolies: change in h over short paths of a few hundred km (D layer)2. frequency deviations: change in frequency of a reflected signal (E and F regions)
3. cosmic noise absorption: sudden decrease in cosmic noise4. increase in TEC (E and F regions)
The time extent of the effects is from several minutes to several hours.
Magnetic storms are disturbances of the geomagnetic field that last from a few hours to
several days.
Factors that influence magnetic stormbehavior include:
polar heatingthermal expansionelectromagnetic driftneutral air dragglobal atmospheric circulationatmospheric composition
Consequences of magnetic storms:
power outages
increase in atmospheric drag on satellitesaffects the discharging of satellitesinduces large voltages in big conductors(e.g., Alaskan pipeline)
adverse effects on radio communications