Content uploaded by Boryana Markova
Author content
All content in this area was uploaded by Boryana Markova on Jan 14, 2020
Content may be subject to copyright.
XIV International Conference on Atmospheric Electricity, August 08-12, 2011, Rio de Janeiro, Brazil
1
Analyses of Summer Lightning Activity over
Bulgaria and Black Sea – Impact of
Environmental Conditions
Rumjana Mitzeva1, Boryana Markova2,3, Savka Petrova4
1 Faculty of Physics, University of Sofia, Sofia 1164, Bulgaria,
2 PhD student, Faculty of Physics, University of Sofia, Sofia 1164, Bulgaria,
3National Institute of Meteorology and Hydrology – Varna, Varna 9002, Bulgaria,
4 Faculty of Physics, University of Sofia, Sofia 1164, Bulgaria
ABSTRACT: The lightning activity over Bulgaria (land) and Black sea (maritime area) during the summer
period is analyzed in the 3-hour time intervals. The environmental conditions in the grid boxes where
lightning is detected are also examined. The observed maximum flash density over Bulgaria is around 1200
UTC, while over the Black sea - around 0600 UTC. The results show that during the afternoon the flash
density is greater over land than over the sea, while during the night and early morning hours - vice versa.
The obtained classification function that is combination of the environmental conditions at different levels
over land and sea has very high discrimination ability - more than 96% of cases were correctly classified.
This indicates that the atmospheric conditions at thunderstorms development over land and sea are
significantly different.
1. INTRODUCTION
It is now established (Christian et al., 2003) that the global distribution of the annual flash rate over
land is about an order of magnitude greater than this over the oceans. The spatial distribution of lightning
presented in Christian et al, 2003 shows that lightning is detected (with different frequency) all over the
continental area, while lightning over ocean - only in limited areas (spots). One possible reason for this is
the well known difference in surface conditions over land and sea.
The aim of the present work is to examine i) if the established predominance of global annual lightning
activity over land versus over maritime areas is “valid” for a local area and a limited time interval; ii) are
the atmospheric conditions significantly different at the places of lightning detection over land in
comparison with these over sea?
The answer of first question is based on the analyses of spatial and diurnal distribution of lightning
activity over Black sea and over Bulgaria during summer. For the answer of the second question it is
assumed that if the environmental conditions at the detection of lightning over land and sea are
significantly different it should be possible to derive a classification function which would be able to
discriminate the atmospheric conditions (at surface and at higher levels) at the development of
thunderstorms over sea and over land.
Correspondence to:
Rumjana Mitzeva, Faculty of Physics, University of Sofia, Sofia 1164, Bulgaria. E-mail rumypm@phys.uni-sofia.bg
XIV International Conference on Atmospheric Electricity, August 08-12, 2011, Rio de Janeiro, Brazil
2
2. DATA AND RESULTS
The territory of investigation includes Black sea (maritime area) and Bulgaria (land). Lightning data
are provided from the ZEUS network operated by the National Observatory of Athens (NOA). Details on
ZEUS system, its detection efficiency and location errors are given in Kotroni and Lagouvardos (2008) and
Lagouvardos et al. (2009). The surface data and temperature and humidity profiles are taken from
proximity sounding obtained by the numerical model GFS (http://www.arl.noaa.gov/ready/cmet.html).
The available lightning data for 74 days during June, July, August (JJA) 2006 are used to calculate the
flash density over Bulgaria and Black sea. For this purpose, the lightning data from ZEUS network have
been organized into 0.25x0.25 degree grid boxes in 3-hour time intervals. The middles of the intervals
correspond to synoptic observations times (0000 UTC, 0300 UTC, …, 2100 UTC). For example 1200 UTC
corresponds to the time interval 1030 - 1330 UTC. Taking into account that the local time is UTC+3 over
the area of analysis, the 1200 UTC time interval in our work is referred to as the afternoon interval.
The analyses show that the flash density (number of detected flashes during the studied period, JJA
2006 divided by the corresponding surface area: [fl/km2]) is almost 3 times higher over Bulgaria than over
the Black sea (see thick columns in the upper right corner in Fig.1). Similar to the results in Christian et al,
2003, the analysis of spatial distribution of lightning activity indicates that during 2006 summer (JJA) great
number of flashes are detected over scanty parts of the Black sea territory in comparison to lightning
detected over Bulgaria.
The analysis also reveals (Fig. 1) that during the night and early morning hours flash density is greater
over the Black sea than over the land, while during the afternoon there is a dominant presence of lightning
over land. Maximum flash density over Bulgaria is observed around 1200 UTC time interval (1030-1320
UTC); while over the Black sea its maximum is around
0600 UTC time interval (0430-0730 UTC). On Fig.1 it is
visible that the difference between flash density over sea
and over land is the largest at 1200 UTC. For this reason
the detailed analyses of lightning activity and
environmental conditions are carried out for the
afternoon time interval centred at 1200 UTC.
To answer the question if there is a difference in
environmental conditions at lightning detection over
Bulgaria and Black sea a detailed analysis of surface and
at higher level atmospheric characteristic for the grid
boxes with lightning detection is carried out.
Fig.2 reveals that the surface conditions over Black
sea and over continental area of Bulgaria are different at
1200 UTC. The results demonstrate that the
thunderstorms over Black sea developed at significantly higher relative humidity and mixing ratio and
lower surface temperature in comparison with the corresponding surface data over continental area. There
is also a visible difference between dew point depression at 850 hPa, 700 hPa and at 500 hPa (Fig. 3), as
well between several calculated thermodynamical indices (an example is given in Fig. 4).
JJA 2006
0
0,2
0,4
0,6
0,8
1
0 3 6 9 12 15 18 21
hour (UTC)
flash density
Black sea
Bulgaria
0,8
2,2
Fig.1 Diurnal distribution of flash density
(flashes/km2) at 3-h time interval over
Black sea (dark blue column) and over
Bulgaria (white column). In the upper right
corner - flash density for JJA 2006
XIV International Conference on Atmospheric Electricity, August 08-12, 2011, Rio de Janeiro, Brazil
3
Temperature
Mean
Mean±SE
Mean±1.96*SE
land sea
20.0
20.4
20.8
21.2
21.6
22.0
22.4
Relativ Humidity
Mean
Mean±SE
Mean±1.96*SE
land sea
56
60
64
68
72
76
80
Mixing Ratio
Mean
Mean±SE
Mean±1.96*SE
land sea
9.0
9.6
10.2
10.8
11.4
12.0
12.6
Fig.2 Box and Whiskers plot of surface characteristics in the sample over land and sea at 1200 UTC
(T-Td) 850
Mean
Mean±SE
Mean±1.96*SE
land sea
3.0
3.5
4.0
4.5
5.0
5.5
6.0
(T-Td) 700
Mean
Mean±SE
Mean±1.96*SE
land sea
2.4
2.6
2.8
3.0
3.2
(T-Td) 500
Mean
Mean±SE
Mean±1.96*SE
land sea
7.6
8.0
8.4
8.8
9.2
9.6
10.0
Fig.3 Box and Whiskers plot of dew point depression (T-Td) at 850 hPa, 700 hPa and 500 hPa
in the sample over land and sea at 1200 UTC
Li
Mean
Mean ±SE
Mean ±1.9 6*SE
land sea
-2.8
-2.6
-2.4
-2.2
-2.0
TT
Mean
Mean ±SE
Mean ±1 .96*SE
land sea
47.2
47.6
48.0
48.4
48.8
49.2
49.6
K
Mean
Mean ±SE
Mean ±1.9 6*SE
land sea
28.5
29.0
29.5
30.0
30.5
31.0
31.5
32.0
32.5
Fig.4 Box and Whiskers plot of Li, TT and K index in the sample over land and sea at 1200 UTC
Based on general and stepwise discriminant analyses different classification functions F(L,S),
combination of environmental conditions at different levels over land and sea, are obtained. An example is
given in Table 1. At F(L,S) > 0 the case is classified as lightning over Black sea; at F(L,S) ≤ 0 the case is
classified as lightning over Bulgaria.
The results show that 86 % of the cases are correctly classified (see row 1 in Table 1) using only
information for surface data – temperature T, dew-point depression (T-Td) and mixing ratio MR. The
supplement of thermodynamical indices (row 2 in Table 1) or characteristics of the atmosphere at higher
(850, 700 and 500 hPa) levels (row 3 in Table 1) leads to the “creation” of classification function with
better ability to discriminate cases with lightning over land from cases over sea.
XIV International Conference on Atmospheric Electricity, August 08-12, 2011, Rio de Janeiro, Brazil
4
Table1. Classification function F(L,S) for the detection of lightning over sea and over land and the
percentage of correctly classified cases.
row
F(L,S)
Correct classification
%
whole
area
land
sea
1
F(L,S) = 1.5092 T – 1.7392 (T-Td) – 1.6564 MR – 2.1600
86
84.5
95.7
2
F(L,S) = - 0.00005 CAPE + 0.8017 Li – 1.0190 K + 1.0267 TT +
1.5467KO – 0.0204 Sweat + 0.1360 T + 0.2222 (T-Td) + 5.1534 MR
– 66.5461
96.2
95.9
98
3
F(L,S) = 2.2899 T – 2.3511 (T-Td) – 3.2619 T850 + 2.5023 (T-Td)850 +
1.0493 T700 – 0.2807 (T-Td)700 + 0.3626 T500 + 0.1831 (T-Td)500 –
3.1031
97.4
97.8
95.1
3. CONCLUSIONS
The present study shows that the flash density (calculated for JJA 2006) over Bulgaria is significantly
higher than over the Black sea. This result is in accordance with the established on the globe (Christian et
al., 2003) prevalence of mean annual flashes over continental area in comparison with over the maritime
areas. The study however reveals that this is valid for the noon and afternoon hours. During the nigh and
early morning hours flash density is higher over the Black sea than over Bulgaria. The similar diurnal
evolution of the differences in flash density is established for the Mediterranean during summer period
(Petrova et al, 2009).
Our study also shows that the mean values of surface and at higher level atmospheric data, as well as
some thermodynamical indices are significantly different at the development of thunderstorms over
maritime and over continental area. The obtained function, combination of environmental conditions at
different levels over land and sea, has very high discrimination ability - more than 96% of cases are
correctly classified. These results demonstrate that “special” combinations of atmospheric characteristics
are required for the development of thunderstorms over land and sea, indicating that thunderstorms over
land and over sea form at significantly different atmospheric conditions.
ACKNOWLEDGMENTS
The present work is partially supported by the Science Foundation of Sofia University (grant 166/2010).
REFERENCES
Christian, H.J., et al., Global frequency and distribution of lightning as observed from space by the optical
transient detector.J. Geophys. Res. 108 (D1), 4005,doi:10.1029/2002JD002347, 2003
Kotroni, V., K. Lagouvardos. Lightning occurrence in relation with elevation, terrain slope and vegetation
cover over the Mediterranean. J. Geophys. Res. [Atmos], 113, D21118, doi:10.1029/2008JD010605,
2008
Lagouvardos, K., V. Kotroni, H.-D.Betz and K.Schmidt. A comparison of lightning data provided by ZEUS
and LINET networks over Western Europe, Nat. Hazards Earth Syst. Sci., 9, 1713-1717, 2009.
Petrova S., R. Mitzeva, V. Kotroni, J. Latham and E. Peneva. Analyses of summer lightning activity and
precipitation in the Central and Eastern Mediterranean, Atmospheric Research 91, 453-458, 2009