|
Items in cart [1] |
Questions? Call 888-624-8373 |
| Copyright © 2009. National Academy of Sciences. All rights reserved. Terms of Use and Privacy Statement |
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 206
The Global Atmo spheric- Electrical
Circuit
RAYMOND G. ROBLE arid ISRAEL TZUR
National Center for Atmospheric Research
Lightning was recognized as a grand manifestation of
static electricity within thunderstorm clouds in the
eighteenth century. It was also recognized that electri-
cal phenomena are not confined to thunderclouds and
that a weak electrification exists as a permanent prop-
erty of the atmosphere even during fair weather. Fur-
ther research established that the Earth's surface is
charged negatively and the air is charged positively,
with a vertical electric field of about 100 V/m existing in
the atmosphere near the Earth's surface. An electro-
static explanation for the phenomena was sought at
first, and one theory suggested that the electric field of
the atmosphere was the result of an intrinsic negative
charge on the Earth, probably collected during the
Earth's formation. With the discovery of cosmic-ray
ionization in the early twentieth century, it was realized
that air possesses an electrical conductivity due to its ion
content. As a result of the finite electrical conductivity,
vertical conduction currents flow from the atmosphere
to the Earth, tending to neutralize the charge on the
Earth. On the basis of actual conductivity values it was
calculated that charge neutralization would take place
in less than an hour, and the continued existence of an
electric field suggested some generation mechanism to
oppose the leakage currents flowing to the Earth. The
search for this generation mechanism soon became the
main object of research on global atmospheric electric-
ity.
206
In the early twentieth century the concept of a global
circuit of atmospheric electricity slowly began to evolve
(Israel, 1973; Pierce, 1977~. The net positive space
charge in the air between the ground and a height of
about 10 km is nearly equal to the.negative charge on the
surface of the Earth. The electrical conductivity of the
air increases rapidly with altitude, and the product of
the local vertical electric field and local conductivity at
any altitude within an atmospheric column gives a con-
stant air-earth current flowing downward. This con-
stant air-earth current with respect to altitude implies
that the current flow is mainly driven by a constant dif-
ference in potential between the surface of the Earth
and some higher altitude in the atmosphere. The discov-
ery of the highly conducting ionosphere in the 1920s ex-
plained the long-range propagation of radio waves and
was important for the evolution of the concept of the
global electric circuit. The ionosphere, with its large
electrical conductivity, provided a means of closing the
global circuit. It, however, is not a perfect conductor
parallel to the Earth's surface, but it possesses a finite
conductivity, and the electric currents and fields within
it are driven by the combined action of the ionospheric
and magnetospheric dynamo systems as well as by cur-
rent generation from the lower atmosphere.
Wilson (1920) first demonstrated that a thunderstorm
supplies a negative charge to the Earth. In the 1920s, it
was also known that over the oceans and in polar areas
OCR for page 207
THE GLOBAL ATMOSPHERIC-ELECTRICAL CIRCUIT
the diurnal maximum of the fair-weather potential gra-
dient at the Earth's surface occurred at the same univer-
sal time (about 1900 UT). Furthermore, radio measure-
ments of atmospherics showed that global thunderstorm
activity also peaked near 1900 UT, with the main thun-
derstorm centers being in Africa and South America.
Scientists studying meteorological statistics of thunder-
storm activity found similar diurnal variations. The di-
urnal UT variation of potential gradient over the oceans
was similar to the diurnal UT variation of thunderstorm
occurrence frequency with no phase delay. These exper-
imental facts all contributed to the concept of the
Earth's global electrical circuit and furthermore sug-
gested that thunderstorms were the generators within
the circuit. This concept of the global electrical circuit
persists today, although there are still basic problems
and details that need to be resolved (Dolezalek, 1971,
1972; Kasemir, 1979~.
THUNDERSTORMS AS GENERATORS IN THE
GLOBAL CIRCUIT
The vast majority of clouds in the atmosphere form
and dissipate without ever producing precipitation or
lightning. A cloud, however, interacts with atmo-
spheric ions and becomes electrified to a certain degree.
The small, fast ions in the atmosphere almost exclusively
provide the electrical conductivity. The intermediate
and large ions do not greatly contribute because of their
lower mobility. The fast ions within clouds become at-
tached to more massive cloud particles and thereby de-
crease the electrical conductivity within the cloud rela-
tive to the surrounding clear air. As a result of this
electrical conductivity change alone, clouds act as an
electrical obstacle; space charge develops on the surface
of the cloud, and the distribution of fair-weather con-
duction currents and fields flowing in the vicinity of the
cloud are altered. As convective activity intensifies,
electrification increases. Strong electrification generally
begins with the rapid vertical and horizontal develop-
ment of a fair-weather cumulus cloud into a cumulo-
nimbus type. Most of the lightning on Earth is produced
by strongly convective cumulonimbus clouds with a vig-
orous system of updrafts and downdrafts.
The updrafts and downdrafts associated with convec-
tion and the interactions between cloud and precipita-
tion particles (Tzur and Levin, 1981; Rawlins, 1982) act
in some manner that eventually separates charges
within the thundercloud. Charge-separation processes
usually fill the upper portion of the thundercloud with
the main positive charge and the lower portion with the
main negative charge. The measured charge structure
within individual clouds is complex (Krehbiel et al.,
207
1979), and a simplifying assumption that is usually
made for modeling purposes considers the charge to be
distributed in a spherical fashion within some finite vol-
ume. A finite spherical distribution of charge has the
same distant electric-field structure in space as an equiv-
alent point charge. Therefore, for simplicity in studies
of the electrical interaction of a thunderstorm with its
immediate environment, the lightning charges, which
are lowered to ground during a lightning stroke, and the
main thunderstorm charges are usually represented as a
series of point charges.
The main negative charge in the lower part of a thun-
derstorm (Krehbiel et al., 1979) occurs at a height where
the atmospheric temperature is between -10°C and
- 20°C. This temperature range is typically between 6
and 8 km for summer thunderstorms and around 2 km
for winter thunderstorms. The positive charge at the top
of the storm does not have so clear a relationship with
temperature as the negative charge but can typically oc-
cur between - 25°C and - 60°C depending on the size
of the storm. This temperature range usually lies be-
tween 8 and 16 km in altitude. The ensemble of thun-
derstorms occurring over the globe at any given time
will have positive and negative charge centers located
over a large altitude range depending on the atmo-
spheric structure, the size of the thunderstorm, its type
(e.g., air mass, frontal), its location (e.g., ocean, plains,
or mountains), its latitude, its stage of development,
and other factors.
There is no generally accepted model of thunder-
storm electrification that can be used to calculate the
current that storms release into the global electrical cir-
cuit. Measurements by Gish and Wait (1950), Stergis et
al. (1957a), Vonnegut et al. (1966, 1973), and Kasemir
(1979) showed that the total current flowing upward
from thunderstorms areas ranges from 0.1 to 6 A with an
average of about 0.7 A per thunderstorm cell. It is, thus,
possible to use this value in a theoretical model without
referring to the details of the charge-separation mecha-
nisms. The simplest model used to investigate the elec-
trical interactions of a thunderstorm with its immediate
environment assumes a quasi-static dipolar charge dis-
tribution embedded within the thundercloud, which is
immersed in a conducting atmosphere whose electrical
conductivity increases exponentially from the surface of
the Earth to a highly conducting region somewhere
within the ionosphere above about 60 km.
In the quasi-static state these charges are maintained
in equilibrium against discharge currents by assuming
that a steady convection current acts between the two
charge centers in the updraft and downdraft regions of
the storm. In earlier studies only conduction currents
were assumed to flow in the environment, and although
OCR for page 208
208
this assumption is probably valid in the highly conduct-
ing region above a storm it is not valid within and below
the storm, where corona, lightning, precipitation, con-
vection, and displacement currents all contribute to the
charge exchange between charge centers and between
the storm and the surface of the Earth. Holzer and
Saxon (1952) and Kasemir (1959) presented analytic so-
lutions for the potential distribution around thunder-
storms whose charge distributions are represented as
point dipole current sources. A calculated normalized
potential distribution around a quasi-static dipolar
charge structure within the atmosphere (Kasemir, 1959)
is shown in Figure 15. liar . The lines represent stream-
lines of current flow between the two charge centers and
between the charge centers and the ionosphere and the
ground. The highly conducting Earth's surface is con-
sidered by including image point sources within the
Earth for the solution of the electrostatic potential dis-
tribution around the thunderstorm. The currents flow
upward from the positive charge center toward the ion-
osphere and from the ground toward the negative
charge. There is also a current flow between the
charges. All currents are calculated as conduction cur
(a)
\~
oo 01 02
/'
~ . . .l, I/,/, / I/,,/, ,//, //////.//////,,,, //,,l,,,,,/////,li,,,i,,/,i,/,,
(b)
,03
[]Wo
(:Wj
bwu
we [ ]
,,,, ,l,,,,7,,,,,,,,,,,,,,,,,,,,,,/////////,////////~///////
FIGURE 15.1 (a) Schematic giving the current streamlines of dipo-
lar point current sources embedded within an atmosphere with expo-
nentially increasing conductivity over a perfectly conducting Earth,
(b) schematic of a lumped parameter representation of the global at-
mospheric electrical circuit where the thunderstorm is represented as a
current generator with internal resistance Wi, WO represents the resis-
tance between cloud top and the ionosphere, Wa represents the fair-
weather load resistance, and W., represents the resistance between the
thunderstorm and the ground (Kasemir, 1959).
RAYMOND G. ROBLE and ISRAEL TZUR
rents in this simple model and have considerably more
complexity in actual storms.
Holzer and Saxon (1952) showed that the current out-
put from such a thunderstorm model is sensitive to the
charge-separation distance, with a greater current out-
put from larger storms having intense vertical velocities
and charge-separation capability. They also showed
that storms with frequent cloud-to-ground lightning
strokes move negative charge to the Earth's surface and
leave the cloud with a net positive charge that has a
greater current output than a dipole. Thus, according to
their calculations, even a weak storm with charge sepa-
ration could supply a current to the ionosphere and the
global circuit.
Kasemir (1959) showed that the exponential conduc-
tivity increase with altitude in the atmosphere is impor-
tant for the current flow from thunderstorms toward
the ionosphere. He pointed out that if the atmosphere
had a constant conductivity at all altitudes the primary
thunderstorm source current would flow downward to
the image source, with no current flow toward the iono-
sphere and into the global circuit. Anderson and Freier
(1969) considered both the limiting fast and slow time
variations within thunderstorms. Their dipole quasi-
static solutions for slow time variation show that the po-
tential distribution about thunderstorms is affected by
both the conductivity and the charge-separation dis-
tance within storms. Their calculations also suggest that
the thunderstorm, in a quasi-static sense, may be electri-
cally closed and only the pumping action of lightning
discharges can provide the current necessary to main-
tain the ionospheric potential.
Park and Dejnakarintra (1973, 1977a) considered a
dipolar thunderstorm model and analytically solved for
the current output and the mapping of thundercloud
electric fields into the ionosphere. This model consid-
ered the anisotropy of the electrical conductivity above
about 60 km and assumed that the Earth's geomagnetic-
field lines were vertical. The largest ionospheric fields
were calculated to occur at night over giant thunder-
storms with values on the order of 10 ~ 4 V/m at 100 km.
During the day the calculated do electric fields are 1 to 2
orders of magnitude smaller because of the increased
ionospheric electrical conductivity caused by solar ex-
treme ultraviolet (EUV) radiation ionization. Dejna-
karintra and Park (1974) examined the penetration of
lightning-induced ac fields into the ionosphere and
found that the lightning electric-field signal recovery
time decreases rapidly with increasing altitude until at
100 km the electric-field wave form appears as a sharp
pulse. The ac fields are also larger at night than during
the day, when ionospheric conductivities are larger.
Wait (1960) pointed out that there is a constant radial
OCR for page 209
THE GLOBAL ATMOSPHERIC-ELECTRICAL CIRCUIT
potential associated with the lower-order (n = 0) mode
in a concentric spherical cavity excited by a radial cur-
rent-moment element. He estimated that the potential
across the earth-ionosphere gap caused by a single 3-km-
length lightning discharge of 1000 A would be on the
order of 1 V, and he suggested that the omnipresent con-
stant voltage during fair weather might result from the
accumulated action of a number of lightning strokes.
Hill (1971) further examined the resonant hypothesis
and showed that the atmospheric field is generated
through electrostatic induction by equivalent charge di-
poles from thunderclouds. Using parameters of light-
ning frequency and the magnitude of the current mo-
ment within storms available at that time, he found an
ionospheric potential only about one third of the ob-
served intensity.
Schumann resonances in the earth-ionosphere cavity
are also excited by worldwide lightning activity, and
this subject has been reviewed by Polk (1982~.
Willett (1979) pointed out that the total current sup-
plied to the global electric circuit from a thunderstorm
differs depending on whether the storm is considered to
be a current or a voltage generator. For a current thun-
derstorm generator, where the current supplied to the
global circuit is independent of the load, his results
showed that current output is proportional to the as-
sumed source current strength and to a proportionality
factor that depends on the ratio of the height of the iono-
sphere to the electrical conductivity scale height. For a
voltage-source thunderstorm generator, where the
charging current increases with time until limited by
some voltage-sensitive dissipative mechanism such as
lightning or corona, the current output may be some-
what smaller depending on assumptions made concern-
ing the storm's internal resistance. Not enough is known
about the thunderstorm generator to distinguish be-
tween the current and voltage source models.
Freier (1979) presented a thunderstorm model with
more details than point-charge sources and conduction
currents. He considered the atmosphere between the
Earth and the ionosphere divided into three separate re-
gions as shown in Figure 15.2. In region 1, below the
negative layer of charge in a thunderstorm, he consid-
ered conduction, displacement, and precipitation cur-
rent densities, allowing all to vary with altitude. In re-
gion 2, between the bottom and top of the thunderstorm
cloud, he considered charging, conduction, displace-
ment, and precipitation current densities that also vary
in space and time. In region 3, above the storm, only
conduction and displacement are considered. All light-
ning currents are considered to be discontinuous charge
transfers, and in the fair-weather regions far to the side
of the thunderstorm only conduction currents flow.
209
~ T ~ Conduction, JE
~ ~ ~ _
Con \ ~ /
12 km \ + + +
,\ + + +
~/
i<'
(-40 to-60°C) ~
Charge
Separation
Current
7
- - 1- 10° to -20°C)
/. . ...
Lightning. JL\3! pO.t I I I I I I I ~ Precipitation, UP
Discharge, ~ ~ I ~
Wf /// ////f~f,Wf~fif/~,7
J M = JE + JO + JL + JP + At Below T. S.
JM = JE + ) ~Above T. S.
FIGURE 15.2 Schematic illustrating the various currents that flow
within and in the vicinity of thunderstorms: JE is the conduction cur-
rent, Jr is a convection current, Jo is the lightning current, JP is the
precipitation current, ~D/8t is the displacement current, and JM is the
total Maxwell current.
This model is an improvement over previous thunder-
storm models, which considered lumped circuit param-
eters of columnar resistance as shown in Figure 15.1(b)
for the dipolar conduction current generator. Freier
(1979) used it to describe a charge-transfer mechanism
that may occur during large severe storms when precipi-
tation is heavy and when the storms are, in certain in-
stances, accompanied by tornadoes. The precipitation
current during such storms may be so large that it re-
moves much of the negative charge in the lower portion
of the storm and allows the generator to operate be-
tween the positive charge and the ground. During such
storms much more electrical energy is produced, with a
possibly greater current output that must be considered
in any global electric model.
Krider and Musser (1982) pointed out that the time
variations in thunderstorm electric fields, both aloft and
at the ground, can be interpreted as a total Maxwell cur-
rent density that varies slowly in intervals between
lightning discharges. The total Maxwell current, ~M,
consists of field-dependent currents, both linear and
nonlinear (corona), ~E; convection currents, ~c; light-
ning currents, If; and the displacement current density,
aDIdt.
OCR for page 210
210
J\I = JE + JC + JL + Edict. (15.1)
The field-dependent currents and convection currents
are distributed over large areas and, in the absence of
lightning, tend to vary rather slowly with time. The
lightning current occurs impulsively and represents a
discontinuous transfer of charge in both space and time.
Corona and lightning usually transfer negative charge
to the ground, and precipitation and turbulence trans-
fer positive charge.
Krider and Musser (1982) showed that the average
Maxwell current density is usually not affected by light-
ning discharges and varies slowly throughout the evolu-
tion of the storm. Since the Maxwell current is steady at
times when the electric field both at the ground and
aloft undergoes large changes in amplitude, and some-
times even polarity, they inferred that the cloud electri-
fication processes are substantially independent of the
electric field. Also, since the Maxwell current varies
slowly throughout the storm, it probably represents an
electrical quantity that is coupled directly to the mete-
orological structure of the storm. Thus, the thunder-
storm appears to be a current source that produces a
quasi-static Maxwell current density.
Grenet (1947, 1959) and Vonnegut (1953, 1965) pro-
posed a convective thunderstorm generator in which the
electrification process is related to the updrafts and
downdrafts acting within the thunderstorm. In this gen-
eration process positive space charge is convected to the
top of the thunderstorm cloud by updrafts within it, and
the negative charge attracted to the top of the cloud is
swept downward by convective downdrafts at the edges
of the cloud. For this generator, the magnitude of the
cloud electrification is related to the strength of thun-
derstorm convective activity. The convection mecha-
nism also involves conduction currents flowing from the
upper atmosphere to the top of the cloud. The magni-
tude of the electrification process can, therefore, in-
crease with an increase in the electrical conductivity of
the atmosphere over the cloud.
The above discussion reveals the complexity and diffi-
culty involved in modeling the thunderstorm as a gener-
ator in the global circuit. Significant progress has been
made in recent years, but there is still a long way to go
before a generally accepted thunderstorm model will be
available for use in a global model of atmospheric elec-
tricity. Yet such a model is of prime importance, and
research in this area should be pursued vigorously.
SOME PROPERTIES OF THE GLOBAL
CIRCUIT
According to the classical picture of atmospheric elec-
tricity (Dolezalek, 1972; Israel, 1973), the totality of
RAYMOND G. ROBLE and ISRAEL TZUR
thunderstorms acting together at any time charges the
ionosphere to a potential of several hundred thousand
volts with respect to the Earth's surface. This potential
difference drives a vertical electric conduction current
downward from the ionosphere to the ground in all fair-
weather regions on the globe. The fair-weather electric
conduction current varies according to the ionospheric
potential difference and the columnar resistance be-
tween the ionosphere and the ground. Horizontal cur-
rents flow freely along the highly conducting Earth's
surface and in the ionosphere. A current flows upward
from a thunderstorm cloud top toward the ionosphere
and also from the ground into the thunderstorm genera-
tor, closing the circuit. A lumped parameter schematic
of the global circuit is shown in Figure 15.1(b). This
schematic does not represent the real circuit but rather
illustrates its basic concepts. The global fair-weather
load resistance is given as Wa and is about 250 Q. The
thunderstorm generator source is shown along with its
equivalent internal resistance, Wi, which is not well
known. The total resistance between the thunderstorm
and ionosphere is represented as WO and is about 105-106
Q. and the total resistance between the thunderstorm
and ground is represented by W7, and is also not well
known. Markson (1978), however, suggests that W7t is
small because of corona discharge beneath the storm,
having a value of about 104-lOs Q. Some of the overall
properties of the global circuit are summarized in Table
15.1. There are many additional elements that compli-
cate this simplified classical picture, and these are dis-
cussed in the following subsections.
Global Thunderstorms
The thunderstorm generator hypothesis proposed by
Wilson (1920) was based on his observations that be-
neath the thundercloud negative charge is transferred to
the Earth and above the thundercloud positive charge is
transferred to the conductive upper atmosphere. A sub-
sequent discovery was the close correlation between the
diurnal universal time variation of the thunderstorm
generator current (represented by the frequency of
thunderstorm occurrence) and the load current (repre-
sented by the fair-weather ground electric field or air-
earth current density), integrated over the surface of the
Earth.
In about the 1920s the electric field over the oceans
was found to vary diurnally in accordance with univer-
sal time (Parkinson and Torrenson, 1931), as shown in
the upper frame of Figure 15.3. The diurnal change of
electrical conductivity over the oceans is relatively
small, and therefore Ohm's law requires that the air-
earth current density also follow the diurnal variation of
electric field. The maximum value of both the average
OCR for page 211
THE GLOBAL ATMOSPHERIC-ELECTRICAL CIRCUIT
TABLE 15.1 Some Properties of the Global Circuit
Number of Thunderstorms Acting at One Time
Currents above Thunderstorms (A)
(a) Range
(b) Average
Global Current (A)
Ionospheric Potential (kV)
(a) Range
(b) Mean
Columnar Resistance at Sea Level (Q/m0)
(a) Lowlatitude
(b) High latitude
(e) Tibet and Antarctic plateau
Total Resistance (Q)
(including resistance decrease by mountains)
Current Density (A/m2)
(a) Inhabited and industrialized areas
(b) Vegetated ground and deserts
(c) South Pole Station
Potential Gradient (V/mJ
(a) Equator
(b) 60° latitude
(c) South Pole
(d) Industrial areas
Average Charge Transfer over the Entire World
(Ckm~0yr~~)
Total Charge on the Earth (C)
Electrical Relaxation Times
(a) 70 km
(b) 18km
(c) 0.01 km
(d) Earth's surface
Electrical Conductivity (mho/m)
Sea level
Tropopause
Stratopause
Ionosphere
(a) Pedersen conductivity
(b) Parallel conductivity
1500-2000
0.1to6
0.5tol
750-2000
150-600
280
1.3 X 101'
3 X 101'
2 X 10
230
200
1 x 10-12
2.4 X 10-12
2.5 X 10-12
120
155
71
300-400
+90C
500,000
10 ~ 4 see
4 see
5-40 min
10 ~ 5 see
10 - 14
10-13
10- lo
10-4-10-5
10
electric field and the current over the oceans occurs near
1900 UT, and minimum values near 0400 UT.
Whipple and Scrase (1936) obtained the average
thunderstorm probability as a function of local time at
Kew, England, from corona current records. Assuming
that the same thunderstorm probability curve also exists
on other continents as a function of local time, they com-
bined this curve with the world thunderstorm day statis-
tics of Brooks (1925) and obtained the diurnal variation
of worldwide thunderstorm activity as a function of uni-
versal time, shown in the bottom frame of Figure 15.3.
The three major component curves with their maxima
at 0800, 1400, and 2000 UT represent the contributions
of the major thunderstorm regions of Asia and Austra-
lia, Africa and Europe, and America, respectively. The
summation of the component curves represents the diur-
nal variation of worldwide thunderstorm activity. The
similarity of the diurnal variation of electric field over
211
the ocean and the diurnal variation of worldwide thun-
derstorm activity supports the hypothesis that thunder-
storms are the electrical generator in the global circuit.
The maxima and minima of both curves occur at about
the same universal times. The amplitudes of modulation
for the two curves, however, are different. The ampli-
tude of the electric-field curve is about 20 percent, and
the amplitude of the thunderstorm curve is about 45
percent. Whipple and Scrase (1936) suggested that this
difference might be resolved if a steady supply current
from worldwide ocean thunderstorms is added to the
worldwide continental distribution of storms. This sug-
gestion, however, is not supported by current data. The
difference in amplitudes probably arises from the enor-
mous variability in thunderstorm electrification.
Although the similarity between the diurnal UT
worldwide thunderstorm frequency curve and the diur-
nal UT electric-field curve suggests that thunderstorms
OCR for page 212
212
120
110
100
on
12OI
100
8O
60j~
4OI
20
o
Carnegie _'
_ r
~7N
12 16 20
HOUR
The World
t' _ ~_ _ _ `
~America
New Zealand ~| ~ ~
0 48 12 16 20 _
HOUR
FIGURE 15.3 (a) Annual curve of the diurnal variation of the atmo-
spheric electric field on the oceans (volts per meter) as measured by the
Carnegie and Maud expeditions (Parkinson and Torrenson, 1931) and
(b) annual curve of the diurnal variations of global thunderstorm ac-
tivity according to Whipple and Scrase (1936~.
are the generators in the global electrical circuit, there is
still considerable uncertainty concerning the details, as
discussed by Dolezalek (1972) and Kasemir (1979~.
Moreover, the data on which the thunderstorm activity
is based are only of a qualitative nature: "a thunder-
storm day is a day when thunder has been heard." The
commonly quoted UT diurnal patterns, shown in Fig-
ure 15.3, are averages over a long time that were made
to reduce the influence of various disturbing factors.
When shorter time averages are used and even single
diurnal variations the correlations show great depar-
tures from the average curves. The variability of thun-
derstorm frequency can be large, with significant de-
partures from the average; for example, whole
continents may be cloudless for a long time. The mea-
sured electric field on the ground is also highly variable
as a result of local influences, and it generally takes a
week's worth of averaging or more to bring out the diur-
nal UT pattern. Most measurements are made on conti-
nents, where the electric field displays variations with
local time, and these measurements do not fit into a
daily worldwide pattern but must be averaged to deter-
mine worldwide characteristics. Paramanov (1950) sug-
gested that local influences might cancel out if the elec-
tric fields measured on many continents were
synchronized to universal time and averaged.
RAYMOND G. ROBLE and ISRAEL TZUR
Dolezalek (1972), in examining the data available at
the time, concluded that a globally controlled current
does flow vertically through the atmosphere but that its
connection to thunderstorm activity is tenuous and, in
fact, is often contradicted by the proper interpretation
I of available measurements.
24 More recently, Orville and Spencer (1979) examined
lightning flashes recorded in photographs by two satel
lites in the Defense Meteorological Satellite Program
(DMSP) and found that most of the lightning is confined
to land areas and that the ratio of global lightning fre
quency during northern summer to that of southern
summer is about 1.4 for both the dusk and midnight sat
ellite data. They pointed out that this summer-winter
difference in global lightning frequency is opposite to
the electric-field measurements. The hypothesized rela
tion of the global atmospheric electric current to thun
24 derstorms is still an unsettled question and clearly needs
to be resolved to make further progress in understanding
the Earth's global atmospheric electric circuit.
Current above Thunderstorms
A few measurements have been made that give the
magnitude of the current flowing upward over the
whole area of a thundercloud (Gish and Wait, 1950;
Stergis et al., 1957a; Vonnegut et al., 1966; Imyanitov
et al., 1969; Kasemir, 1979~. The currents range from
0.1 up to 6 A, with an average between about 0.5 and 1
A per thunderstorm cell. Gish and Wait (1950) flew an
aircraft over a thunderstorm in the central United States
and found an average upward current of 0.8 A at an
altitude of about 12 km. They measured electric fields of
up to 70 kV/m. Stergis et al. (1957b), in a series of bal
loon flights at altitudes of about 25 km in central Flor
ida, measured an average upward current of 1.3 A. The
electric fields that they measured at this altitude were on
the order of a few hundred volts per meter. Holzworth
(1981), with a balloon near 20 km over a large thunder
storm at Fort Simpson, NWT, Canada, on August 15,
1977, measured a vertical upward electric field of more
than 6.7 V/m (the instrumentation threshold) for longer
than 2 hours. These few data indicate that a positive
current flows toward the ionosphere above thunder
storm regions that is of sufficient magnitude to account
for fair-weather conduction current. Is it possible to re
late current output to such factors as frequency of cloud
to-ground strokes, charge structure and separation dis
tances, and cloud-top height? Both Pierce (1970) and
Prentice and Macherras (1977) presented relationships
for a latitudinal variation in the ratio between cloud-to
ground and cloud-to-cloud flashes. This ratio is about
0.1 in the equatorial region, increasing to about 0.4 near
OCR for page 213
THE GLOBAL ATMOSPHERIC-ELECTRICAL CIRCUIT
50° latitude. Do lightning-intensive tropical thunder-
storms deliver more or less current than thunderstorms
at higher latitudes, even though the ratio of cloud-to-
ground to cloud-to-cloud flashes is smaller? A consider-
able amount of research is necessary to determine the
processes responsible for regulating the current flow
from thunderstorms into the global circuit. This re-
search is important for understanding the role of world-
wide thunderstorm activity as the generator for the
global circuit.
Electrical Conductivity, Columnar Resistance, and
Global Resistance
Galactic cosmic rays are the main source of ionization
that maintains the electrical conductivity of the atmo-
sphere from the ground to about 60 km in altitude. Near
the ground, however, there is additional ionization due
to release of radioactive gases from the soil, and above
about 60 km solar ultraviolet radiation becomes impor-
tant. During geomagnetic storms, ionization due to en-
ergetic auroral electron precipitation and to auroral x-
ray bremsstrahlung radiation and proton bombardment
during solar proton events can all be significant sources
for the high-latitude middle atmosphere.
The galactic cosmic rays that bombard the Earth's at-
mosphere are influenced by the Earth's geomagnetic
field, which produces a magnetic latitudinal effect in
the incoming cosmic-ray flux. The full cosmic-ray spec-
trum is only capable of reaching the Earth at geo-
magnetic latitudes higher than about 60°. At lower geo-
magnetic latitudes, the lower-energy particles are
successively excluded by the Earth's geomagnetic field,
and only particles with energies greater than about 15
GeV reach the equator. The cosmic-ray spectrum thus
hardens with a decrease in geomagnetic latitude, with
the height of the maximum ion production rate decreas-
ing from about 20 km at high latitudes to about 10 km
near the equator. The cosmic-ray ion production rate
profile for various geomagnetic latitudes during solar-
cycle minimum using the data from Neher (1967) is
shown in Figure 15.4. The ionization rate increases with
geomagnetic latitude, and both the height of the peak
and the slope of the ionization rate above the peak also
increase. Near the ground there is about a 20 percent
variation between the equatorial region and higher lati-
tudes (Israel, 1973~.
The cosmic-ray ionization undergoes a regular solar-
cycle variation with maximum values near solar mini-
mum. At high latitudes the ionization rate may vary by
about 50 percent at 15 km and about 75 percent at 20
km. At 20 km the ionization-rate variation through the
solar cycle is about 40 percent in mid-latitudes and
40
10~
'~\''"~\'\1111l
- 23° 38° 53°
0~463°
213
O
101 1o2 1 o3
QO (ion pairs/cm3 sec atm)
FIGURE 15.4 Cosmic-ray ion-production rate vertical profiles at
various geomagnetic latitudes (Neher, 1967; R. Williamson, Stanford
University, personal communications, 1981~.
about 20 percent near the equator. In addition to the
solar-cycle variation of galactic cosmic-ray flux, there
are shorter-term variations that are associated with
some magnetic storms, a 27-day quasi-periodic varia-
tion, variations due to solar flares, a diurnal variation,
and a variation in the amplitude of the diurnal variation
with time. All these variations have been reviewed by
Forbush (1966~. The diurnal variations are generally
small. The solar-flare and magnetic-storm variations of
cosmic-ray fluxes are larger (about 2 to 20 percent) and
are more important for understanding solar-terrestrial
electrical coupling mechanisms (Roble and Hays, 1982).
The full impact of all these variations on the electrical
conductivity and the properties of the global circuit has
not yet been evaluated.
Over land, the natural radioactivity of the solid
ground adds to the cosmic-ray ion-production rate, not
by direct radiation from the solid surface but in the re-
lease of gaseous intermediaries from rocks and soil on
the surface and from soil capillaries. These gaseous in-
termediaries are then carried upward by vertical mass
transfer, and radiation from them can affect the ion-
production rate within the first kilometer above the
Earth's surface (Israel, 1973~.
Above about 60 km the solar ultraviolet radiation ion-
ization exceeds the ion production owing to galactic cos-
mic rays. The major source of ionization within the
mesosphere is nitric oxide, which can be ionized to NO +
by solar Lyman-alpha radiation at 121.6 nm. Above
about 80 km, the EUV and soft x-ray radiation from the
Sun produces ionization in the ionospheric E and F re
OCR for page 214
;;
214
, ..
"ions. These EUV radiations all have a large d~y-to-day
variability as well as periodic 27-day solar rotation and
solar-cycle variations. During solar flares the solar EUV
and x-ray radiation can be greatly enhanced, thereby
increasing the ionization source and electrical conduc-
tivity of the ionosphere.
The ionization sources mentioned above all contrib-
ute to the electrical conductivity of the atmosphere
through the production of positive and negative ions
and electrons. The electrical conductivity is governed
by the number of positive and negative ions and elec-
trons as well as by their respective mobility. The num-
ber densities of various ionic species~are controlled by
complicated chemical reactions between ions and neu-
tral species. Many of the neutral constituents important
for ion chemistry, such as water vapor and NO, are fur-
thermore transported by atmospheric motions, thereby
increasing the complexity of determining the global dis-
tribution of electrical conductivity. All these various
processes are discussed by Reid (Chapter 14, this vol-
ume).
Above about 80 km, the electrical conductivity is gov-
erned by collisions between neutrals, ions, and electrons
and by the influence of the geomagnetic field on the mo-
bility of the plasma components. The electron gyrofre-
quency is greater than the electron-neutral collision fre-
quency above about 80 km, and therefore the electrons
are restricted by the geomagnetic-field line. For the ions
this occurs above about 140 km. The differential motion
of the ions and the electrons in the dynamo region be-
tween about 80 and 200 km in altitude gives rise to an
anisotropic behavior of the conductivity. The conduc-
tivity parallel to the geomagnetic-field lines is not af-
fected by the field, and it increases rapidly with alti-
tude, limited only by collisions between electrons and
neutrals and ions. The Pedersen conductivity is parallel
to an applied electric field and orthogonal to the mag-
netic field. It is smaller than the parallel conductivity,
and it has a maximum near 140 km, where the ion-neu-
tral collision rate equals the gyrofrequency of ions. The
conductivity orthogonal to both an applied electric field
and the magnetic field is the Hall conductivity, which
maximizes near 105 km. The Pedersen conductivity is
carried by electrons below 105 km and by ions above
that altitude, whereas the Hall current is mainly due to
electrons.
A typical profile of electrical conductivity through
the daytime atmosphere is shown in Figure 15.5. Near
the Earth's surface the electrical conductivity is about
10- ii mho/m. It increases exponentially with altitude,
having a scale height of about 6 km until about 60 km,
where the effects of free electrons become important
and there is an abrupt increase in the conductivity.
RAYMOND G. ROBLE and ISRAEL TZUR
\ ~ M chemosphere
\
7L rm
\ \~
ah\ \
\ ~ Dynerno R - 'on /
K)C
/a,
E
y
_ ~
:r
Dower Atmosphere
l
-1 2 -10 -8 ~ -4
Earth ~ Interior
-low _
I I _
O 2
log a (A m )
FIGURE 15.5 Altitude variation of the electrical conductivity in the
Earth's atmosphere and ionosphere from the ground to 200 km. The
electrical conductivity within the Earth is shown for comparison (Vol-
land, 1982~.
Above about 80 km the geomagnetic field introduces an-
isotropic conductivity components. The parallel con-
ductivity continues to increase with altitude, whereas
the Hall and Pedersen conductivities peak near 105 and
140 km, respectively, before decreasing with altitude.
The electrical conductivity of the Earth is about 10-3
mho/m, and therefore the Earth's atmosphere can be
considered a leaky dielectric sandwiched between two
highly conducting regions the Earth's surface and the
ionosphere.
A calculation of the latitudinal variation of the verti-
cal and horizontal electrical conductivity components
in the daytime atmosphere during equinox by Tzur and
Roble (1983) is shown in Figure 15.6. The electron den-
sities above about 10 km are calculated using the model
of Reid (1976, 1977) and the properties of the neutral
atmosphere are specified from the model of Solomon et
al. (1982a, 1982b). Below 10 km, the electrical conduc-
tivity is represented by the Gish formula, with a latitu-
dinal variation as determined from Israel (1973~. The
electrical conductivity increases abruptly above about
60 km because free electrons are present in the daytime
ionosphere, and the effect of the geomagnetic-field line
becomes apparent above 80 km. The electrical conduc-
tivity parallel to the geomagnetic-field line is larger than
either the Hall or the Pedersen conductivity, so the verti-
cal component of the electrical conductivity is small in
the equatorial region, where the magnetic-field line is
horizontal; and the horizontal component of the electri-
cal conductivity is small in the polar region, where the
magnetic-field line is vertical.
This calculation for the global distribution of electri-
cal conductivity represents an idealized case considering
fair-weather conditions. In reality, the conductivity is
OCR for page 215
THE GLOBAL ATMOSPHERIC-ELECTRICAL CIRCUIT
90
80
70
-
Y 60
50
5 40
30
o
80 60 40 20 0 -20 -40 -60 -80
LATITUDE
probably quite variable depending on variations of ion-
ization rates, ion and neutral chemical reactions, aero-
sol and cloud interactions, and a host of other meteoro-
logical factors. The electrical columnar resistance that
determines the local air-earth current flow and electric
field is determined by the height integral of the recipro-
cal of the electrical conductivity distribution. The bulk
of the columnar resistance resides in the troposphere,
which can be strongly controlled by electrical conduc-
tivity variation owing to factors such as clouds, fog,
aerosols, and pollution. Boeck (1976) showed that
changes in the global conductivity may result in from
85Kr being released into the atmosphere. The low-fre-
quency radiation from power lines (Vampola, 1977)
may also affect the precipitation of electrons and conse-
quently affect ionization in the stratosphere. The co-
lumnar resistance derived from the Gish formula is
about 1.3 X 10~7 Q/m2. This fair-weather value proba-
bly varies considerably in place and time as determined
by aerosol and weather conditions. The global variation
of the columnar resistance is not well known, yet it is an
important property of the global electrical circuit. The
global resistance is the parallel circuit resistance ob-
tained by adding the various columnar resistance val-
ues. Muhleisen (1977) estimated a global resistance of
230 Q without mountains and 200 Q when the Earth's
orography is considered. The magnitude of the variabil-
ity of global resistance is poorly known, and it is an im-
portant parameter that needs to be determined in order
to improve our understanding of the Earth's global elec-
trical circuit.
Fair-Weather Vertical Current Density and Total
Current
The total current flowing in the global circuit is not
well known, and only crude estimates have been made.
215
80 60 40 20 0 -20 -40 -60 -80
LATITUDE
FIGURE 15.6 Latitudinal distribution of
the (a) vertical and (b) horizontal components
of logo ~ (mhos/m), where ~ is the electrical
conductivity (Tzur and Roble, 1983).
Its value is generally estimated by integrating the mea-
surable fair-weather vertical current density over the
fair-weather area of the globe. A recent estimate by
Muhleisen (1977) used 10-12 A/m2 for inhabited and in-
dustrialized areas, 2-4 X 10-12 A/m2 for vegetated
ground and for deserts, 2.5 X 10-12 A/m2 over the At-
lantic Ocean to derive a total global current of about
1000 A. About 750 A is derived for the current flow over
oceans and 250 A for the current flow over the conti-
nents. Muhleisen (1977) pointed out that the columnar
resistance over mountains is much smaller than over flat
land near sea level, and he estimated that as much as 20
percent of the global vertical current streams toward
mountains. Gathman and Anderson (1977) showed that
the air-earth current also has a latitudinal variation due
to the effect of the Earth's geomagnetic field on the cos-
mic-ray ionization rate throughout the troposphere.
Another means of estimating the total current flow in
the circuit is to estimate the total number of thunder-
storms working simultaneously and multiply that value
by the average current output determined from mea-
surements over thunderstorms, as was discussed previ-
ously. Estimates of the number of global thunderstorms
range from 1500 to 2000. If these numbers are multi-
plied by the average current output of thunderstorms
(0.5 to 1.0 A), the total current is about 750 to 2000 A,
which is nearly the same magnitude as the total current
derived by air-earth current estimates, about 1000 A.
Blanchard (1963) indicated that currents as large as
hundreds of amperes flow from the ocean surface into
the atmosphere as a result of electrified droplets that are
ejected from the bursting of small bubbles. It also should
be mentioned that intense electrification is associated
with some volcanic eruptions. When such phenomena
occur, they may have a significant input into the global
electrical circuit. There is considerable uncertainty
with these estimates that needs to be resolved. It is un
OCR for page 216
216
likely that the global current can be accurately deter-
mined from only a few localized measurements, and
some other means must be used to determine its value.
Localized measurements, however, are needed to cleter-
mine the range of variability, especially from globally
representative stations such as on a mountaintop (Rei-
ter, 1977a).
Ionosphere Potential
An important parameter for determining the electri-
cal state of the global circuit is the ionospheric potential,
which specifies the potential difference between the
ground and the ionosphere assuming that the ground is
arbitrarily referenced to zero. This quantity can be esti-
mated by using the values of the global resistance, about
200 Q. and the values of the total current flowing in the
circuit, 750 to 2000 A, to get 150 to 400 kV.
The ionospheric potential is of fundamental impor-
tance for the global circuit because it is one of the few
measurable global parameters. It can be determined by
integrating the altitude profile of the electric field mea-
sured from ascending or descending balloons or aircraft.
This technique uses the facts that the electrical conduc-
tivity of the ionosphere is so large that any horizontal
potential difference is small and the entire ionosphere
has a uniform potential difference with respect to the
Earth. At high latitudes, however, it is necessary to con-
sider the horizontal potential differences that are gener
RAYMOND G. ROBLE and ISRAEL TZUR
ated by the solar-wind/magnetospheric generator dis-
cussed later. Another factor is that the electric field
decreases exponentially with altitude in such a manner
that the product of the electrical conductivity and elec-
tric fielc] remains constant with altitu~le. Therefore, the
bulk of the ionospheric potential drop occurs within the
first few kilometers above the Earth's surface, a region
that is easily attainable by balloons and aircraft.
The ionosphere potential can also be determined
from measurements of the air-earth current density and
electrical-conductivity profiles. Muhleisen (1977) sum-
marizec] the distribution of his ionospheric potential
measurements made from balloons during the period
1959 to 1970. The minimum measured potential was
145 kV, the mean 278 kV, and the maximum 608 kV.
The measurements also showed that the diurnal UT var-
iation of ionospheric potential is similar to the Carnegie
curve, shown in Figure 15.3(top), and that there is an
11-year variation in ionospheric potential that is out of
phase with the solar sunspot cycle. Markson (1976,
1977) made aircraft measurements of the ionospheric
potential that are shown in Figure 15.7. These measure-
ments have a diurnal UT variation similar to the Carne-
gie curve and also show the magnitude of day-to-day
variability that is associated with the inherent variabil-
ity of the worldwide thunderstorm generator and, to a
smaller extent, the global resistance. Measurements of
the ionospheric potential are important for understand-
ing the global electric circuit and should be continued.
FIGURE 15.7 Summary of ionospheric po- ~. · ~ . . . . . .
tential measurements as a function of time 330 _ A
(UT) of aircraft soundings made by Markson 320 _ / \
_ /:
160 ~\
1 SO 7 ~2 4 3 ~O ~O ~2 4 ~'4 34 ·' 27 27 30 27 24~°
not2r~r ~nterwo.
i I ~ I l I I I I I I I I I I I ~ I I I I ~ I
0 1 2 3 4 5 6 7 8 9 10 it 12 13 14 1S 16 17 18 19 20 21 22 23 0
HOUR (UT)
