td>Yahoo! is not affiliated with the authors of this page or responsible for its content.
African Journal of Science and Technology (AJST) 95
AJST, Vol. 7, No. 2: December, 2006
African Journal of Science and Technology (AJST)
Science and Engineering Series Vol. 7, No. 2, pp. 95 - 98
THE CHARACTERISTICS OF POSITIVE GROUND DISCHARGES
OF TROPICAL THUNDERSTORMS AT IBADAN, NIGERIA
Oladiran, E.O.
1
, E.F. Nymphas
1
, U. E. Akpan
2
and S. Israelsson
3
1
Physics Department, University of Ibadan, Ibadan, Nigeria
2
Physics Department, University of Uyo, Uyo, Nigeria
3
Department of Meteorology, Uppsala University, Uppsala, Sweden
ABSTRACT:- Positive ground lightning discharges were continuously recorded over a period of
three years at Ibadan to supplement earlier composite records from 1987 by adapting the earlier
design of Oladiran, et al (1988a) using the shape of the radiation field signatures and the frequency
components of positive discharges (Oladiran and Israelsson, 1990). We recorded 92% discrimination
between negative and positive ground flashes. The positive ground flashes were recorded at VLF
and 17.3kHz, 3dB attenuation and yielded an average 13 and 11.6% ratios for (+CG)/(-CG)
respectively. Compared to our results of 1988, the flash-rate characteristics for positive ground
flashes is not significantly different from those of negative ground flashes and it does not show any
seasonal preference. This leads to the conclusion that the occurrence of positive and negative
ground flashes depends only on the cloud charge structure, its dynamics and the ground conditions.
INTRODUCTION
The ground discharges, (CG), have been classified into
two groups- (a) those that bring negative charges to
ground (-CG), and (b) those bringing positive charges to
the ground (+CG). The physical and electrical
manifestations of these two types of flashes are very
distinct vide: (i) The radiation fields contain maximum
energies at different peak frequencies (9.7kHz and 17.3kHz
for (-CG) and (+CG) respectively at 3dB attenuation
(Cooray, 1984; Oladiran et al, 1988a)); (ii) When the leader
is followed by return strokes, the number of return strokes
accompanying (-CG) is usually more than those
accompanying (+CG), (Aina, 1971a,b; Bruce and Golde,
1942); (iii) The shape of the radiation fields for the two
types of discharges are of different forms in shape, rise
time, duration and frequency component (Kitagawa and
Brook, 1960); (iv) The (+CG) carry a larger magnitude of
charge to earth than the (-CG) (Rust et al, 1981).
Consistent studies of the two types of CG started at
Ibadan in 1986 and consist of:
(a) (-CG) flashes as earlier reported in 1988 (Oladiran et
al, 1988b) using a new lightning flash counter and
calibration circuit with improved discrimination of
cloud and ground discharges to within 92% (Oladiran,
et al, 1988a), and which showed that there were no
contradicting evidences to the characteristics of (-
CG) as earlier reported, and that finer determinations
of empirical constants were therefore possible. Weibull
distribution was further confirmed as a strong
mathematical tool for characterizing thunderstorm
parameters.
(b) The spectral characterization as reported by Oladiran
and Israelsson (1990), the highlights of the results
being:
(1) The weather characterization of the tropical
thunderstorms as shown in Table 1.
(2) The ratio of negative ground discharges to cloud
discharges was 1:14.2
(3) For all signals monitored, only 2.3% of the flashes
from thunderstorms were (+CG). 96
AJST, Vol. 7, No. 2: December, 2006
E. O. OLADIRAN
Table 1: Weather Characterization of Lightning Signals
Storm duration
(hr)
Max. & Mean rain rate
(mm/hr)
Peak freq. Range
and Mean (kHz)
Av. Discharge/
storm(ground)
No. of signals
studied
% distribution
of duration
Peak field
(kV/m)
0 - 0.5
maximum = 186
8.4 - 14.2
36
46
8
16
Average = 92
Average = 12.4
0.5 - 1.0
maximum = 138
6.8 - 20.6
148
183
26
108
Average = 67
Average = 8.9
1.0 - 1.5
maximum = 105
10 - 15.3
302
198
31
63
Average = 52
Average = 13.6
1.5 - 2.0
maximum = 156
9.3 - 21. 8
185
104
22
38
Average = 43
Average = 10.3
> 2.0
maximum = 122
6.5 - 22.7
87
83
13
13
Average = 22
Average 7.8
In this report, further characterization of the (+CG) by
direct measurement of ground flash density and the
waveform characteristics of the accompanying spherics
is made. A simple electrical circuit implementation for (+CG)
counts is presented. The results are based on three years
of continuous comparative recordings of counts and
signal form of (+CG), (-CG) and the spheric signals.
EXPERIMENTAL ARRANGEMENT
The instrument designed and constructed by Pisler and
Oladiran (1985) for time registration of flashes, and which
stores 2000 flash events by date and time to the nearest
10ms interval was used. The flash count was also
simultaneously made mechanically via a modified flash
counter (Oladiran et al, 1988a). Spheric signals from a
horizontal wire antenna elevated at 10m above the ground
and 26m in length were simultaneously monitored and
recorded on a transient storage oscilloscope. A coupling
network as described by Oladiran and Israelsson (1990)
was used to divide the signal from a plate antenna into
two components to simultaneously activate the flash
counter for mechanical counting and electronic occurrence
time storage and be recorded on the other channel of the
transient storage oscilloscope, so that visual signal
comparison of the various signals could be made.
Frequency decomposition of the electronic and radiation
signals from lightning discharges have shown that the
peak energies of (-CG) and (+CG) are centered at 9.0 kHz
and 17.3 kHz respectively (Oladiran and Israelsson, 1990;
Oladiran et al, 1988b). The lightning flash counters used
in this investigation were calibrated with these properties.
The digitized data is being prepared for rigorous
computer-aided data decomposition.
RESULTS AND DISCUSSIONS
Table 2 shows the percentage distributions of (-CG) and
(+CG) obtained from spheric signals at various frequencies
(Nymphas, 1995). This Table is a low frequency
characterization of discharges to ground and the results
can be summarized as follows:
(i) The percentage distribution are strongly dependent
on the frequency at which one is monitoring the
signal;
(ii) There is no significant weather dependence observed
when flashes are monitored at low frequencies. An
attempt to find a dependence of recorded flash count
on local temperature, maximum wind speed, total
precipitation and maximum and minimum rate of
rainfall; duration of rainfall, and the relative humidity
yielded a maximum regression coefficient of 0.24 ±
0.06.
(iii) The average (+CG) to (-CG) ratio was found to be
0.13.
Table 2: Ground flash Characterization from spheric
s i g n a l s
Frequency (Hz) Number of Flashes
(-CG)
(+CG)
1000
160
146
14
500
190
179
11
200
185
145
40
100
107
90
17
Type of Flashes AJST, Vol. 7, No. 2: December, 2006
The Characteristics of Positive Ground Discharges of Tropical Thunderstorms at Ibadan, Nigeria
97
For measurements with lightning flash counters using the
plate antenna and with a circuit peak frequency at 17.3
kHz and 3dB attenuation, the following results were
obtained:
a.
The rise time of the monitored radiation field lies
within the range 0.4-20µs with a modal frequency
response of 110-10.3kHz and mean values of 2.4ns
rise time and 17.3 kHz peak frequency response at
3dB attenuation respectively.
b. No multiple recording was observed from the counter
as verified by the transient signal recording at a circuit
response time of 0.26s.
c.
The diurnal variation of (+CG) showed a maximum at
between 2100LT and 2200LT, while the minimum was
Jan
Feb
March
April
May
June
July
Aug
Sept
Oct
Nov
Dec
No. of
Thunderstorm
days
1
1
2
3
2
5
8
0
6
7
1
0
No. of
(+CG)
1
2
14
10
21
58
79
0
108
165
3
0
% distribution 0.2
0.4
3.1
2.2
4.6
12.7
17.3
0
23.6
35.4
0.7
0
situated between the local time range of 1400LT and
1500LT and a fractional distribution of 36% during
the day and 64% during the night. The highest count
rate of (+CG) was 105/hour between the hours of 2000
and 2200LT.
d. The number of thunderstorm days on which (+CG)
occurred is 36 days on an annual average over a period
of three years (1996-1998) with an annual count rate
of 458/year as compared to an annual average of
24,850/year for (-CG) resulting in (+CG)/(-CG) ratio of
0.018 ( 2%). The annual average number of (-CG)
flashes is an updated average from 24,226/year
reported in 1988 by Oladiran, et al (1988). The annual
(+CG) monthly distribution by percentage is shown
in Table 3.
Table 3: Monthly Distribution of Annual Positive Ground Flashes
The distribution shows a seasonal variation, hitherto
unknown for the tropics. For each storm in which (+CG)
occurred, the ratio, (+CG)/(-CG) averages 11.6%. We note
that the annual average of the updated (-CG) is not
significantly different from our earlier report (Oladiran et
al, 1988b) and that the effect of the August break as earlier
observed by many workers (Oshodi, 1971; Osaghaede,
1987) is a consistent phenomenon of this region. It is
probable that the short rise times monitored for (+CG)
(0.4ns to 6ns) would be directly related to the rise time of
the accompanying currents. If this were so, then the
degree of protection of properties offered by lightning
protector may not be adequate for (+