Content uploaded by Montserrat Aran
Author content
All content in this area was uploaded by Montserrat Aran on Jan 25, 2021
Content may be subject to copyright.
Remote sensing analysis of a Mediterranean thundersnow and low-altitude
heavy snowfall event
Joan Bech
a,
⁎, Nicolau Pineda
b
, Tomeu Rigo
b
, Montserrat Aran
b
a
Dep. Astronomy and Meteorology, University of Barcelona, Av. Marti i Franques 1, Barcelona E-08028, Spain
b
Meteorological Service of Catalonia, Berlín 38, Barcelona E-08029, Spain
article info abstract
Article history:
Received 14 January 2012
Received in revised form 18 May 2012
Accepted 19 June 2012
On the 8 of March 2010 a heavy snowfall accompanied by lightning occurred over Catalonia
(NE Spain), in the Western Mediterranean. Total lightning observations included 101
cloud-to-ground flashes and 169 intra-cloud flashes. Precipitation amounts in 24 h exceeded
100 mm and snow depths over low altitude terrain, where snow is rare, surpassed 30 cm.
Snow accumulations collapsed the regional communication transport network and the border
with France was closed several hours. Occurrence of wet snow combined with increasingly
strong winds caused widespread damage over large forest areas estimated in more than
20 MEur and affected dramatically the high voltage power line distribution grid due to ice
accretion, particularly in NE Catalonia where 33 high power electrical towers were knocked
down.
The meteorological framework at synoptic scale was dominated at low levels by a northern
flow over Iberia due to a blocking high pressure system on the British Isles, and an upper level
cold trough, which favoured a rapid cyclogenesis over the Mediterranean (9.2 hPa drop in
12 h). Weather radar observations indicated predominance of stratiform precipitation and
some low-topped convection, with maximum reflectivities and tops mostly below 40 dBZ and
4 km respectively. The presence of mesoscale gravity waves, caused by wind-shear instability,
is suggested as a triggering element for convection and subsequent lightning. Comparison of
accumulated precipitation and lightning maps indicated clusters of lightning data unrelated to
precipitation maxima. Further investigation of total lightning characteristics and co-located
radar observations suggested a triggering effect by tall telecommunication towers inducing
cloud‐to‐ground flashes and subsequent intra-cloud lightning.
© 2012 Elsevier B.V. All rights reserved.
Keywords:
Thundersnow
Low altitude snowfall
Heavy precipitation
Tower triggered lightning
1. Introduction
The occurrence of a thunderstorm accompanied by snow
(i.e., thundersnow) is a relatively uncommon phenomenon,
compared to the frequencies of ordinary snowfalls and thun-
derstorms without snow (Schultz and Vavrek, 2009). In fact
the term thundersnow does not appear explicitly in special-
ized glossaries such as the World Meteorological Organization
terminology database METEOTERM (WMO, 2011), or the
American Meteorological Society Glossary (Glickman, 2000).
However, the concept is mentioned in other texts such as the
UK Met. Office Meteorological Glossary (HMSO, 1991), or the
Standard Dictionary of Meteorological Sciences by Proulx
(1971) where appears as “thundery snowfall”. Because of the
combination of lightning and snow, thundersnows may reduce
substantially visibility and thus affecting surface transport
or aviation activities and also increasing the risk of lightning
accident to individuals staying outdoors (Cherington, 2001). It
should be noted that not all winter thunderstorms are nec-
essarily thundersnows and also that not all thundersnows
produce heavy snowfalls. On the other hand, general charac-
teristics and originating mechanisms associated to cold season
thunderstorms are of primary importance in thundersnows so
they are also considered here.
Atmospheric Research 123 (2013) 305–322
⁎Corresponding author. Tel.: +34 93 402 11 23; fax: +34 93 402 11 35.
E-mail address: joan.bech@ub.edu (J. Bech).
0169-8095/$ –see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.atmosres.2012.06.021
Contents lists available at SciVerse ScienceDirect
Atmospheric Research
journal homepage: www.elsevier.com/locate/atmos
The number of articles in the scientific literature related to
thundersnows, either climatologies or case studies, is rela-
tively scarce and mostly focussed in North America –USA and,
to a lesser extent, Canada –and Japan. For example Michimoto
(1991, 1993),Takeuchi et al. (1978) and Kitagawa and
Michimoto (1994) describe meteorological and electrical
characteristics associated to winter thunderstorms in Japan.
Thundersnow case studies in the USA are described by Sanders
and Bosart (1986),Martin (1998),Market et al. (2007),and
Pettegrew et al. (2009). Other studies are provided by Henson
et al. (2007), who examine an event affecting Montreal,
Canada, and by Dolif Neto et al. (2009) who describe a cou-
ple of low latitude thundersnow events in the US (Texas Coast)
and southern Brazil. Climatologies or statistical analysis con-
sidering thundersnow events in the US can be found in Hunter
et al. (2001),Market et al. (2002, 2006),Crowe et al. (2006) or
Schultz (2009). On the other hand, specific winter thunder-
storm characteristics are given by Munzar and Franc (2003),
dealing with cold season precipitation events in the Czech
Republic in Europe, or by Altaratz et al. (2001) examining
winter thunderstorm characteristics in Israel, in the Eastern
Mediterranean. There are other works specifically related with
lightning characteristics of winter thunderstorms from the
point of view of the microphysical electrification processes or
the engineering electrical protection. For example Miyake et al.
(1990),orDiendorfer et al. (2006) present results from studies
of lightning from winter thunderstorms affecting high towers
or tall structures and López et al. (2013) report in detail a case
study of winter thunderstorm lightning which damaged a
weather radar installed on a high tower. Necessary factors for
the occurrence of thundersnows are summarized by Schultz
and Vavrek (2009) and include abundant moisture, lifting
mechanism, unstable temperature profile plus low tempera-
tures to have snow reaching the ground. General characteris-
tics of winter thunderstorms are a relatively low number of
lightning strokes, compared to summer season events, while in
thundersnows, cloud-to-ground lightning is typically associat-
ed to precipitation maxima, which is sometimes organized as
heavy snow bands.
The objective of this study is to provide an analysis of
meteorological conditions and remote sensing observations of
a heavy snowfall and thundersnow event that took place on
the 8 March of 2010 over Catalonia (NE Spain), a mid-latitude
region in the Western Mediterranean. The snowfall had an
unusual large extension and local intensity (in some places
24 h accumulations exceeded 100 mm of equivalent liquid
precipitation), causing substantial disruption to socioeconomic
activities and transport as it affected low altitude areas where
snow is relatively rare. Therefore, the purpose of documenting
this event is twofold. Firstly, to describe the synoptic and
mesoscale conditions leading to the heavy snowfall, contrib-
uting to improve our understanding of this type of events in
the region and helping to identify those features in future
forecasts. Secondly, to provide a specific analysis of the
thundersnow characteristics, with special emphasis on total
lightning (i.e. cloud to ground and intracloud strokes) and
weather radar observations, so that they can be identified in
the surveillance process, i.e. very short range forecast process, in
similar future events.
The organization of the rest of the paper is as follows.
Section 2 describes briefly the data used and the region study,
to present the main geographical features explained later and
also an overview of low altitude events in the region. Section 3
is focussed on general characteristics of the event, including an
overview of the damage, observed precipitation, and synoptic
and mesoscale meteorological features. In Section 4, a detailed
discussion of total lightning and weather radar observations
is provided highlighting substantial differences with warm
season thunderstorms. Finally, Section 5, gives a discussion and
concluding remarks of the study presented.
2. Region of study and observational data
2.1. Region of study
The thundersnow and heavy snowfall event considered in
this study affected Catalonia, located in the NE of the Iberian
Peninsula (Fig. 1). Catalonia, with an area about 32,000 km
2
,
is limited by the Mediterranean Sea to the east and south, and
by the Pyrenees mountain range to the north, with average
heights of 2000 m and maximum tops exceeding 3000 m.
There are also other lower mountain ranges such as the
Pre-Pyrenees range (parallel and south to the Pyrenees), or
the Coastal range –maximum heights about 700 m –and
Pre-coastal range, with the highest hills between 1000 m and
1700 m approximately. Most population of the region (about
5 out of 7 million inhabitants) live in Barcelona city and the
surrounding metropolitan area (ca. 50 km from the city),
which concentrates a large number of industry, communica-
tion and transport infrastructure, being one of the largest
conurbations in Europe.
While snowfall is usual every winter in the Catalonia
Pyrenees and Pre-Pyrenees, where there are about a dozen of
mountain ski-resorts, it is not that frequent in lower areas such
as the Barcelona metropolitan region. Due to their high impact
to socio-economical activity several meteorological episodes
producing heavy low altitude snowfalls in Catalonia have been
documented in the past —see, for example, Pages et al. (2003),
Pascual et al. (2003) or Mateo et al. (2006).Aran et al. (2010)
examined the climatological frequency of snowfalls through
snow on ground observations affecting Barcelona City from
1947 to 2009 and found 16 cases (about once every four years).
However, high impact cases as this one, characterized by large
extensions and persistence of the snow on ground for several
days, occur much less frequently.
2.2. Observational data and products
Observational data used in this study were provided by the
Meteorological Service of Catalonia, including weather radar
observations, total lightning data, satellite data, radiosonde
data, and surface observations recorded with the automated
surface observation network (Prohom and Herrero, 2008),
covering Catalonia (see Fig. 1a).
Weather radar observations were obtained with a net-
work of four C-band Doppler radar systems (Fig. 1b), designed
to provide quantitative precipitation estimates in a hilly re-
gion where topographical beam blockage is a common prob-
lem (Bech et al., 2003). The weather radars use Travelling
Wave Tube transmitters which allow pulse-compression to
achieve sensitivities and radial resolution similar to those of
higher power systems (O'Hora and Bech, 2007). Weather
306 J. Bech et al. / Atmospheric Research 123 (2013) 305–322
radar products considered in the analysis include, among
others, radar reflectivity composites, velocity volume pro-
cessing displays (Siggia and Holmes, 1991), or quantitative
precipitation estimates derived from quality controlled radar
corrected data (Sánchez-Diezma et al., 2002). Temporal
resolution of radar data is 6′.
Total lightning (cloud-to-cloud and cloud-to-ground)
data was obtained with the Lightning Mapping System
(hereafter, XDDE), composed of four Total Lightning Sensors
(Fig. 1b). Intracloud (IC) and cloud-to-ground (CG) lightning
discharges are detected and processed separately by the
system. IC lightning is detected with a VHF sensor, that
operates in the range from 110 to 118 MHz, and located with
an interferometric technique (Lojou and Cummins, 2006). On
the other hand, cloud-to-ground return strokes are mainly
observed by the low frequency (LF) sensor. The algorithm to
locate the CG strokes is a combined Time-of-Arrival/Magnetic
Direction Finding technique (Cummins et al., 1998). This
network has been experimentally evaluated by means of
electromagnetic field measurements and video recordings
resulting in CG flash detection efficiency above 90% with an
average location accuracy between 0.5 and 1 km (Montanya
et al., 2006). More information on the XDDE can be found in
Pineda and Montanya (2009).
Moreover, the operational SMC object-oriented precipita-
tion tracking system (Rigo et al., 2010), which was employed
Fig. 1. Area of study with relevant geographical features. a). Iberian Peninsula, showing the location of Catalonia (rectangle) in the NE, limited by the Pyrenees to
the north (2000 to 3000 m high) and the Mediterranean Sea to south and east. b) Geographical distribution of the C-band Doppler radar systems (circles) and the
total lightning sensors (squares) considered in the study.
Fig. 2. Monthly averaged cloud-to-ground hourly distributions during the period 2004 to 2008 in Catalonia.
Adapted from Pineda et al., 2011b.
307J. Bech et al. / Atmospheric Research 123 (2013) 305–322
in the past to characterize warm season severe convective
events (Aran et al., 2009; Bech et al., 2007, 2011; Pineda et al.,
2011a) has also been used in this analysis. This system is
similar in some aspects to the TITAN (Dixon and Wiener,
1993) or SCIT (Johnson et al., 1998) systems, and considers
both radar and total lightning data to track and characterize
precipitating storms.
3. The 8 of March 2010 event
Several aspects of the 8 of March 2010 snowfall in Catalonia
have been described recently. Amaro et al. (2010) focussed on
the social impact of the event, Vilaclara et al. (2010) on the
operational warnings issued, and Aran et al.(2010) and Pascual
(2011), provided an overview of the meteorological conditions.
The snow depth recorded at the Fabra observatory in Barcelona
city (16 to 20 cm) was comparable to previous important
snowfalls asthe 8 March 1964 (18 cm), 1 March 1993 (20 cm)
but inferior to the 30 January 1986 (20 to 30 cm) and 15
December 2001 (25 cm) (SMC, 2010). Moreover, during the
event lightning was present (101 CG and 169 IC flashes) and
thunderstorm occurrence in March in this area is per se a rather
anomalous event as most lightning is concentrated in the
warm season. Fig. 2 shows the monthly distribution of hourly
lightning averaged during the period 2004 to 2008 (Pineda et
al., 2011b), indicating thatduring March a thunderstorm event
Fig. 3. Comparison of 850 hPa temperatures (in Celsius) in March 2010 over Barcelona obtained with SMC radiosonde data (blue dotted line) with SMC
radiosonde average (1998–2011) (red line), and ERA-Interim database (1989–2009) (thick black line).
Fig. 4. Examples of the 8 March 2010 snowfall effects. a). One of the 33 high power line towers collapsed due to icing. b). Widespread fallen trees affected not only
vast forest areas but also damaged locally houses and cars. c). Montcau (1053 m asl) and other hills near the coast, were covered by enough snow to allow ski
touring, a very rare circumstance.
308 J. Bech et al. / Atmospheric Research 123 (2013) 305–322
like this is rather uncommon. Several SMC mountain observa-
tories in the Pyrenees broke absolute low temperature records
such as −20.8 °C in Boí (2540 m) and −20.2 °C in Salòria
(2445 m). On the Girona coast, the observatory of L'Estartit
also reached a remarkable minimum low March value with
−2.1 °C, equal to the previous March record in the time
series started in 1968. The below normal temperatures of the
event are illustrated in Fig. 3, which shows March temperatures
over Barcelona at 850 hPa. Climatologicalvalues, obtained both
from the SMC radiosonde data (1998–2011), and the ERA-
Interim database (1989–2009), indicate temperatures above
2.5 °C. In contrast, during the cold advection which caused
this event, temperatures sank below −7.5 °C on the 10 March
2010.
Our focus here is to provide an in-depth analysis the syn-
optic and mesoscale framework and later examine in detail
the thundersnow characteristics through the use of weather
radar and total lightning data.
3.1. Damage overview
As outlined in Amaro et al. (2010) and Aran et al. (2010)
this event caused a high social impact not only in the
Barcelona metropolitan region but also in other large areas
of Catalonia, particularly on the NE. The N part of Catalonia
was affected before 12 Z and due to the heavy snowfall road
authorities closed the highway border between Spain and
France at La Jonquera, leaving about 300 lorries stranded. The
Barcelona airport service had several delays and the effects
were more important in the Girona airport, with 26 out of 31
flights cancelled that day. Regional communication net-
works (highways and railways) were completely affected by
snow accumulations combined with strong winds, especially
during the period from 12Z to 18Z, when most part of the
regional transport system was either stopped or substantially
delayed —many secondary road where completely impass-
able for several hours.
Fig. 5. Top panels: satellite (MODIS) derived snow cover over Catalonia before (a) and after (b) the 8 March 2010 snowfall. c) Precipitation analysis (mm)
performed with automatic surface observations. d). Snow depth analysis (cm) performed with manual and automatic observations.
309J. Bech et al. / Atmospheric Research 123 (2013) 305–322
During some periods of the event the concurrence of
snowfall with humid and light to moderate winds favoured
substantial accumulations of wet snow, which are a relatively
rare phenomenon at low altitudes in this region. Large areas in
relatively low altitude terrain, including hills of the Pre-Coastal
and Coastal range mountains where snow is rare, were affected
by these deep snow accumulations. Wet snow was partic-
ularly important in forests where it broke branches and
downed trees, some of which damaged houses or parked cars
(see Fig. 4)—an area of 35,000 ha was affected according to
regional forest services and mitigation financial support
(21 MEuro) was provided by the regional government. Dam-
age in forests was not onlyan economical problem due to wood
losses, but alsoa serious hazard as favouring factor for wildfires
for the next summer season. Another high impact effect,
particularly in NE Catalonia, was the collapse of a large number
of electrical power lines and subsequent blackouts, some of
which lasted several days. A total of 33 high power tower lines
were knocked down leaving 220,000 users without power. A
preliminary analysis of the most affected areas indicated the
existence of favourable conditions for wet snowicing accretion
on power lines,i.e. combination of near zerotemperatures with
high moisture and moderate to low wind —for more details
see, for example, Makkonen (1998) or Savadjiev and Farzaneh
(2004).
3.2. Precipitation, snow extension and depth
The large extension of the snowfall is illustrated in Fig. 5,
where the upper panels show, before and after the snowfall,
the snow cover variation. The images are derived from
MODIS sensor observations of the Terra satellite, processed
to highlight snow covered surfaces (Hall et al., 2002). Before
the event (Fig. 5a), snow is present only over the Pyrenees
and Pre-Pyrenees mountains, at heights about 2000 m, as
usual at the beginning of March. The change in the snow cover
after the snowfall (Fig. 5b) is evident, and affects the whole of
Catalonia except the SW. The lower panels show accumula-
tions in liquid equivalent, in mm (Fig. 5c), and snow depth
(Fig. 5d), derived from automatic, and manual and automatic
Fig. 7. a) Trajectory of the low pressure centres from the 7 March 2010 12 UTC to 9 March 2010 00 UTC. The observatory locations of Santander (S), Zaragoza (Z),
Barcelona (B), Girona (G), and Palma de Mallorca (P) are indicated with a cross. b). Sea level pressure fall geographical distribution over Catalonia between 00 and
15 UTC on 8 March 2010 obtained from the analysis of surface automatic observations. c). Sea level pressure fall during 8 March 2010 at Barcel ona (thick line),
Girona (dotted line), and Palma de Mallorca (semi-dashed line) airports (AEMET data); see panel a) for location of these observatories.
Fig. 6. MM5 meteorological analysis at 00Z (left column) and 12 Z (right column) showing 300 hPa winds (panels a and b), 500 and 850 hPa geopotential height
contours (gpm) and temperature (shaded, °C) (panels c and d, and e and f, respectively), and higher resolution mean see level pressure field over the region of
study (panels g and h).
311J. Bech et al. / Atmospheric Research 123 (2013) 305–322
observations, respectively. The highest accumulations ex-
ceeded 100 mm in 24 h: Torroella de Montgrí 101.7 mm and
Roses 100.1 mm, in NE Catalonia, while snow depths reached
50 cm above 1000 m in the Pyrenees and Pre-Pyrenees
mountains and valleys, and 30 cm in lower areas. In Barcelona
City, the snow depth was 2 to 3 cm on the beach (at sea level),
about 15 cm in the centre and reached 20 cm at the Fabra
Observatory, located near the Tibidabo hill (SMC, 2010).
3.3. Synoptic and mesoscale framework
The synoptic framework was dominated by several ele-
ments present in other heavy snowfall events previously
documented in this region (Esteban et al., 2005; Pagès et al.,
2003). Previous days were characterized by a northern flow
and cold advection over N Spain due to the presence of a
blocking high over the British Isles (1035 hPa), impinging
cold air from central to southern Europe. This brought lower
than normal temperatures at low levels. The 8 March 2010 a
mid and upper level trough associated to a deep and cold
depression in Central Europe extended over N Spain and the
presence of a strong jet provided a marked cyclonic curvature
at mid-tropospheric levels. Fig. 6 illustrates this setting with
300 hPa winds, 500 and 850 hPa geopotential height and
temperature fields, and sea level pressure at 00 and 12 UTC,
obtained with the operational analysis. Between 00 and 12, at
500 hPa UTC a slow movement of the trough to the SE is
noted (the low centre migrates from SW France to NE
Spain), while changes in the sea level pressure field are
much more dramatic. At 12 UTC over Catalonia a cold air
pool about −30 °C was present and at 850 hPa temperature
was −5 °C. A sudden cyclogenesis over the Mediterranean
generated a surface low pressure area (998 hPa) located
between the Balearic Islands and Catalonia. This changed
completely the low-level (see sea level and 850 hPa maps in
Fig. 6) wind regime in this region leading to increasingly strong
winds, surface convergence, and subsequent cloud development
and precipitation, which was mainly stratiform, but with some
convective parts, and occasional thunder as discussed later.
The movement of the minimum sea-level pressure centres
during the event is depicted in Fig. 7a. The sea level pressure
field evolved in such a way that two lows of similar value on
the 7 March 2010 12 UTC (1012 hPa) moved from N, and
central Spain to the E coast, on the 8 March 2010 00 UTC
(1006 hPa). During the next 12 h the low moved E and then
NE, over the Balearic Islands, reaching a minimum value of
998 hPa located N of the Islands. A secondary low, less intense
(1002 hPa), appeared on the N, nearer to the Catalan coast.
Over Catalonia there was a relatively fast pressure fall il-
lustrated in Fig. 7b which shows a sea level surface pressure
variation map obtained from objective analysis of automatic
observations from 00 to 15 UTC —the pressure drop on the NE
Fig. 8. Meteosat Second Generation infra-red (10.8 μm) and water vapour (6.2 μm) channel images at 12 and 18 UTC 8 March 2010. Note the spiral cloud band
structure associated to the rapidly developing surface low (see inset 18 UTC infra-red image).
312 J. Bech et al. / Atmospheric Research 123 (2013) 305–322
is particularly remarkable, with a variation higher than 9 hPa
in 15 h. Fig. 7c shows more specifically the temporal evo-
lution of mean sea level pressure over three airport observa-
tories: Barcelona, Girona and Palma de Mallorca (locations
shown in Fig. 7a). Barcelona and Girona 12 h pressure drops
were 9.2 and 7.2 hPa respectively. Meteosat Second Genera-
tion images show the cloud shield associated with the rapid
developing surface low (Fig. 8), where a distinct spiral cloud
band structure, an even limited cloudless small area, similar
to a hurricane eye may be observed (see 18 UTC infrared
channel inset). Moreover, the water vapour images indicate
clearly the high level dry intrusion, corresponding to 300 hPa
strong winds according to model data (Fig. 6).
Low altitude snowfalls in Catalonia are usually character-
ized by the combination of cold and dry air advection over the
region and a warmer and more humid flow justabove typically
from the Mediterranean Sea. This configuration was well
identified in this event with the aid of an equivalent potential
temperature cross section analysis. Fig. 9, built with upper air
observations from the observatories indicated in Fig. 7a, i.e.
Santander, Zaragoza, Barcelona, and Palma de Mallorca, shows
cross sections of equivalent potential temperature at 00 and 12
UTC. The cold air advection at low levels is very clear, below a
relatively warmer and more humid air mass. In this case, there
is also an increase in potential instability over the region of
study, particularly at low levels as can be seen comparing
Fig. 9a and b where the stability of each atmospheric layer is
indicated by colour codes (see figure caption for details); as
will be discussed later the absence of deep unstable layers is
also noteworthy during the event considering that lightning
was observed.
During the event temperatures were near 0 °C —see for
example Fig. 10 showing the evolution of wind, precipita-
tion and temperature at La Bisbal d'Empordà (near Girona).
Precipitation type was highly variable, including rain, sleet,
and snow, supported by a constant and relatively weak warm
maritime advection over a much colder and drier air mass
on-shore. As mentioned earlier, the combination of precipi-
tation near 0 °C and winds about 5 m/s favoured ice growth
accretion by riming over surfaces, in particular over power
lines. A few hours later, as the low pressure system developed
near the coast, winds intensified (12 m/s with gusts higher
than 20 m/s) and acted effectively over important ice loads on
the main power lines in Girona which collapsed.
Fig. 9. Upper panels: Barcelona radiosonde plots at 00UTC (a) and 12 UTC (b). On the right of each panel the stability character for each layer is depicted: yellow
(unstable), green (conditionally unstable), and blue (stable). Lower panels: equivalent potential temperature cross sections at 00 UTC (c) and 12 UTC (d). Data
are interpolated approximately along a segment connecting the Bay of Biscay and the Balearic Islands from radiosonde observations recorded at Santander
(08023), Zaragoza (08160), Barcelona (08190), and Palma de Mallorca (08032) (Fig. 7a shows their locations).
313J. Bech et al. / Atmospheric Research 123 (2013) 305–322
Further analysis of visible channel images befor e the onset of
thunderstorm development reveals a cloud structure in bands,
consistent with radar echo top products as seen in Fig. 11 —
note that the flow was from the S, not from the N as it is usual
for mountain gravity waves caused by the Pyrenees in this
region. This banded structure is very similar to that found in
Bosart and Sanders (1986) or Bosart et al. (1998) where they
identified gravity waves as the main triggering and organizing
mechanism of convection in heavy snowfall events. Unlike in
those studies, no evident sign of pressure oscillations was found
on the barometric surface automated records examined —
however, it should be noted that our analysis considered 30′
observations and higher temporal resolution data might
provide more information. The presence of mesoscale gravity
waves in this event seems consistent with the important wind
shear at mid and high levels of the troposphere commented
above. In the absence of airflow over mountains, or penetration
of stable layers by convection from previous deep convection
activity –neither of them existent in this case –vertical wind
shear instability seems the most plausible origin of the gravity
waves observed.
4. Thundersnow characteristics
4.1. Radar data
Weather radar observations allow describing in detail the
spatial distribution and evolution of the precipitation field
Fig. 10. Automatic surface observations from the SMC station at La Bisbal d'Empordà (near Girona) during 8 March 2010 (UTC hours). Variables shown are
temperature (red thick line), mean wind velocity (yellow circle symbols), wind gusts (yellow square symbols), and hourly precipitation amounts (bars).
Fig. 11. a) Radar echo top composite and b) Meteosat Second Generation RGB image composed with HVIS:HVIS:MIR (3.9 μm) channels. Both images correspond
to 8 March 2010 10 UTC.
314 J. Bech et al. / Atmospheric Research 123 (2013) 305–322
during the event, in particular the characteristics of individual
storms. Storm characterization was performed with the SMC
object-oriented storm tracking system (Rigo et al., 2010)based
on the operational analysis of weather radar and lightning
observations. The system distinguishes between non-vertically
developed, or so-called 2-D storm structures, present only in
the lowest radar data volume grid (made up of approximately
2×2 × 1 cubic km bins), and structures with vertical develop-
ment or 3-D structures, whichindicate moreintense convection
and possibly thunderstorms. For each type of structure a cen-
troid is determined. Fig. 12 indicates the position of centroids
of vertically developed storms (3D structures), maximum
reflectivity values (ZMAX) and the volume of each structure
during the event. It may be noted that the number of storms
was relatively low and that were not intense (reflectivitiesb
40 dBZ), with echo top heights around 5 km (not shown). To
examine in more detail these characteristics Fig. 13 illustrates
the evolution of the largest of the precipitating structures in
Fig. 12. Weather radar based precipitation structures (echo tops higher than 2 km) identified during 8 March 2010 over Catalonia with the SMC object-oriented
storm tracking system (Rigo et al., 2010).
315J. Bech et al. / Atmospheric Research 123 (2013) 305–322
terms of lightning and radar data obtained from the storm
tracking system. Maximum reflectivities and 12 dBZ echo tops
were mostly below 40 dBZ and 6 km, respectively, very modest,
compared to typical values found in warm season convective
structures of this area, i.e. maximum reflectivities above 55 dBZ
and echo tops higher than 10 km. Cloud-to-ground flash rates
where about 1 or 2 every 6 min, very low compared to typical
summer thunderstorms with average sustained rates above
5, and peaks above 35 flashes every 6 min (Pineda et al., 2007,
Rigo et al., 2010)—as discussed in more detail in the following
paragraphs.
4.2. Lightning activity
The overall lightning activity recorded by the XDDE during
the thundersnow episode was of 101 CG flashes (231 strokes)
Fig. 13. Example of temporal evolution of selected parameters of a convective precipitation structure during the event derived with the SMC thunderstorm
tracking system (Rigo et al., 2010). The y-axis provides reference units for the variables (radar echo tops TOPS, maximum radar reflectivity ZMAX, areal extension
AREA, and cloud-to-ground flashes CG). Radar echo tops are directly expressed in km; area in reference units/E5 square km, ZMAX in reference units/5 dBZ, and
CG in flash rate.
0
10
20
30
40
50
60
9:00
9:20
9:40
10:00
10:20
10:40
11:00
11:20
11:40
12:00
12:20
12:40
13:00
13:20
13:40
14:00
14:20
14:40
15:00
15:20
15:40
16:00
16:20
16:40
17:00
17:20
17:40
18:00
18:20
18:40
19:00
19:20
19:40
20:00
Time (UTC)
Lightning counts
IC Flash
Stroke 2 Tower"
CG strokes
Fig. 14. Temporal evolution of intra-cloud (IC) and cloud-to-ground (CG) flashes on 8 March 2010 over Catalonia. CG strokes associated to towers are also marked.
316 J. Bech et al. / Atmospheric Research 123 (2013) 305–322
and 169 IC flashes (see Fig. 14). The lightning activity lasted
for almost 8 h, but 90% of the strokes were recorded only in
a five hour period, from 1220 to 1740 UTC. Even during this
period, the intensity was very weak, with a CG flash rate
below 1 flash per minute. Only around 1700 UTC the CG flash
rate reached a 1.7 CG flash per minute.
The lightning data, compared to a typical late spring
or summer thunderstorm day presents some remarkable
differences. Maybe the most noticeable is related to the
intracloud to cloud-to-ground flash ratio, which in summer,
ranges from5 to 10, a higher value sometimes related to severe
weather or heavy precipitation (Boccippio et al., 2001; Cheze
and Sauvageot, 1997). In contrast, in the average value found in
the present case study this ratio is about 1.7, very low (similar
to the specific convective structure examined above).
In comparison to average lightning statistics for the whole
year 2010, the negative CG average first stroke peak current
(−17.4 kA) is similar to the 2010 average (−16.8 kA) while
for the positive CG the thundersnow presented higher values
(34.6 kA) compared to the average (16.6 kA).
Fig. 15. Radar-derived 24 h quantitative precipitation estimates overlaid with cloud-to-ground strokes on different events. Top panel: 28 June 2010 (typical warm
season episode). Bottom panel: 8 March 2010 thundersnow.
317J. Bech et al. / Atmospheric Research 123 (2013) 305–322
4.2.1. Temporal evolution
The first lightning evidence during the thundersnow was
at 0954 UTC with two isolated CG, in the NE of Catalonia, in
the border with France (Fig. 14). No more lightning was
recorded until 1128 UTC, when other CGs began to appear in
the same region. Activity became more regular, with CG flash
Fig. 16. Lightning stroke clusters in the vicinity of telecommunication towers recorded the 8 March 2010 overlaid on Google maps underlays. For each site an inset
with the nearby strokes and another with the tower is shown. a). Puig Neulós (1261 m), on the France–Spain border; b) Rocacorba (991 m), near Girona;
c) Collserola telecommunication tower (447 m, reaching 713 m asl), in Barcelona.
318 J. Bech et al. / Atmospheric Research 123 (2013) 305–322
rates between 0.1 and 0.6 CG min
−1
. The activity at that time
was still in the NE, affecting mainly the coastal area. Only after
1330 UTC some intracloud flashes were observed, but the few
IC flashes of the episode presented big extensions, sometimes
covering distances between 20 and 30 km long. The lightning
activity remained in the same coastal area, in between
Barcelona and the French border until 1500 UTC. Intracloud
activity was larger from 1400 to 1500 compared to the pre-
vious hours, but still with IC:CG ratios around 1. Lightning is
located in regions where radar reflectivity was between 20
and 30 dBZ, with echotops between 5 and 6 km. It is worth
mentioning that at the same moment there were cells located
above the sea but far from the coast with 30–35 dBZ and
echotops between 7 and 8 km but with no lightning at all.
Around 1600 UTC some lightning activity began to be
recorded in the area of the city of Barcelona. As the previous
lightning, the activity is related to radar reflectivity around 25
and 30 and echotops between 5 and 6 km. The higher flash
rates were recorded around 1700 UTC, with a maximum of 17
CG flashes in 10 min. After that, the activity lowered until
1740 UTC, when the last CG flash was recorded. Through the
remainder of this event only intracloud activity was observed.
4.2.2. Spatial distribution
A striking feature, already noticed by SMC weather fore-
casters on duty that day, is that lightning is not collocated
within convective cells and higher reflectivity areas; neither
is coincident with regions with the highest accumulations of
snow. This was evident as the event progressed and the total
precipitation map overlaid with lightning data was available.
During the warm season the maxima of precipitation and
lightning are approximately coincident. To illustrate this dis-
crepancy Fig. 15 shows two examples of radar-derived 24 h
precipitation and cloud-to-ground lightning, one correspond-
ing to an ordinary warm season event, and the other to this
case study. The differences are obvious; in a typical warm
season day only a marginal part of the lightning is not related
to precipitation maxima, while here both locations seem
almost independent. Lightning is not collocated with the
higher storm totals, but it is sparsely disseminated around the
coastal area between Barcelona and the French border. On the
other hand, the distance between a large part of the lightning
and precipitation maxima cannot be explained only by snow
drift below the weather radar beam (Lauri et al., 2012), given
the relative proximity of precipitation observations to radars
and their height.
A more detailed analysis was performed considering a
higher spatial resolution map which showed several spots of
high flash density or clusters. Overlaying the data to high
resolution geographical maps (such as Google Earth TM with
local cartography provided by the Cartographical Institute of
Catalonia) we found that those clusters corresponded to sites
with three high telecommunication towers (see Fig. 16). The
first two towers are on hills in NE Catalonia, on the Pyrenees
and Pre-Pyrenees (at 1260 and 991 m asl respectively), and
the third one is the Barcelona telecommunication tower, built
for the Barcelona 1992 olympic games by the British architect
Sir Norman Foster. It is located over the Collserola mountain
range at 447 m with a height of 266 m, reaching 713 m asl.
4.2.3. Flash clusters and towers
To examine in more detail the characteristics of the
lightning clusters associated to towers, the lightning data set
was split into two groups. The first group was formed by
lightning in radii of 1 km around the towers, assumed to be
direct strikes to the metallic structure, based on the average
error location parameter given by the XDDE system. The rest
of the data was the other group.
Table 1 shows selected characteristics of the two lightning
groups classified as described above. It can be seen that a third
of CG flashes (34 out of 101) are related to one of the three
identified telecommunication towers. Those CGs, almost all of
negative polarity, show higher multiplicity than the other
data set, and also an average peak current about half of the
other group. Taking into account only the first stroke of each
CG flash, the difference in the average peak current is higher,
being almost one third of the average compared to that of
the other group. These results suggest that the characteristics
of the lightning data group related to towers are distinct to
the other one, and are consistent with previous studies of
Diendorfer et al. (2006) or López et al. (2012). Therefore it
seems very plausible that lightning related to towers in this
event was actually triggered by those high structures. This
hypothesis would explain the apparent disconnection of light-
ning clusters to convective activity and subsequent precipita-
tion maxima.
Fig. 17 provides additional insight about convective
precipitation and lightning during the 8 March 2010 event.
It shows a plan view of maximum radar reflectivity of the
radar polar volume, i.e. the so-called MAX product (Vaisala,
2008), with lightning data overlaid, and two complementary
cross-sections. The MAX product was obtained at 14:54 UTC
Table 1
Summary of lightning observations including intra-cloud (IC) and cloud-to-ground (CG) data. Statistics are shown for the whole set considering all lightning data,
and also separately for tower related and non-tower related lightning.
All lightning Lightning to towers Other lightning
Total Neg. Pos. Total Neg. Pos. Total Neg. Pos.
IC flashes 169
CG flases 101 82 19 34 33 1 67 49 18
CG strokes 231 212 19 97 96 1 134 116 18
CG multiplicity 2.6 1 2.9 1 2.4 1
Average peak current (kA) −17.4 34.6 −11.3 8.9 −22.5 36.0
First stroke ave. Peak cur. −23.4 −12.4 −31.8
Subsequent stroke ave. PC −13.7 −10.7 −16.4
Maximum peak current −116.0 100.1 −33.4 8.9 −116.0 100.1
319J. Bech et al. / Atmospheric Research 123 (2013) 305–322
Fig. 17. Radar MAX product and two cross sections recorded with the Puig d'Arques radar (near Girona) at 14:54 UTC on 8th March 2010. The cross sections
correspond to the segments indicated by the letters A and B for the central panel, and C and D for the bottom panel. Reflectivities are expressed in dBZ. Intracloud
VHF sources (black crosses) and cloud-to-ground strokes (black dots) are overlaid on the upper panel.
320 J. Bech et al. / Atmospheric Research 123 (2013) 305–322
with the PDA radar (near Girona) and the lightning data was
observed between 14:50 and 15:00. Two areas with relative
high reflectivity, with values above 30 dBZ, are present
(Fig. 17 upper panel). The first one is over the sea, marked
with the segment AB, associated to a low pressure centre
system and without lightning strokes nearby. The second one
is on the coast of Catalonia, near a tall tower, at about 42°N
latitude, marked with the segment CD and a large number of
lightning flashes. The cross sections reveal that the AB core
over the sea (without lightning) is much more developed
vertically (25 dBZ above 6 km) than the CD region, much
shallower (25 dBZ below 4 km) —with abundant lightning.
This apparent contradiction maybe interpreted by checking
the precise time of lightning data. A cloud-to-ground CG
stroke, shown as a black dot, near label C at (3°W, 42°N), was
observed at 14:58:33.363 UTC followed by a series of
cloud-to-cloud activity initiated at 14:58:33.424 moving to
the E, shown as black crosses (VHF sources). This chronology
of lightning types (first CG and then ICs) seems again con-
tradictory with previous observations of total lightning,
where in a given region ICs occur before CGs.
A possible hypothesis to explain this process is that the IC
activity near the region CD, present with very limited convective
development, is activated by a non-standard mechanism —the
fact that there is no lightning in the CD core would suggest
that the environment was not supportive for electrical dis-
charges. However, a self-triggered CG, i.e. induced by the
presence of a tall tower, may have contributed to charge
redistribution, triggering subsequent IC activity. This would
explain why during this event CG precedes ICs, and IC and CG
are more frequent near tall towers, and are not necessarily
associated with precipitation maxima.
5. Discussion and concluding remarks
The 8 of March 2010 thundersnow event over Catalonia
was remarkable for a number of reasons, including both me-
teorological and social impact aspects due to the abundance
of wet snow combined with wind. Daily precipitation was
over 100 mm and snow accumulations, particularly over low
altitude terrain, were important exceeding 50 cm in some
areas, and reaching 2 to 3 cm at sea level, covering large
extensions of beaches along the coast and collapsing the
transport communication system.
The synoptic framework was dominated at low levels by a
northern flow over Spain due to a blocking high pressure
system on the British Isles, and an upper level cold trough,
which favoured a rapid cyclogenesis over the Mediterranean.
This setting provides necessary elements for heavy precipi-
tation which, given the low temperatures explain the intense
snowfall. However, the mechanisms that lead to thunder-
storm occurrence were not so clear. Radiosonde data in-
dicated that no deep unstable layer was present but there
was an increase of potential instability at low levels. Weather
radar observations indicated predominance of stratiform
precipitation and some low-topped convection, with maxi-
mum reflectivities and tops mostly below 40 dBZ and 4 km
respectively –in agreement with similar winter thunder-
storm studies in the Mediterranean such as Altaratz et al.
(2001). The presence of mesoscale gravity waves, revealed
here by satellite and radar observations, could be a triggering
element for convection, as described in the past regarding
several thundersnow events explaining local intensification
of precipitation. However, in this case lightning was relative-
ly unrelated to total precipitation maxima.
Therefore a second, possibly complementary factor, lead-
ing to lightning occurrence was the triggering effect by tall
telecommunication towers of cloud‐to‐ground flashes. Ac-
cording to Rakov and Uman (2003) the proximity of the cloud
charge to ground and the relatively large horizontal extent of
that charge observed in winter thunderstorms favours the tall
tower lightning triggering effect. Under these favourable
conditions, the simple presence of a tall structure may lead to
the formation of upward leaders and an increase in the num-
ber of lightning strikes (of upward lightning) to the structure
(Mazur et al., 2010). This hypothesis is supported on the
particular characteristics of total lightning observations near
and far from telecommunication towers and comparisons
with weather radar data in precipitation areas with and with-
out convection.
Acknowledgements
Pictures shown in Fig. 4a and b were adapted from Aran et
al. (2010) and the picture in Fig. 4c was provided by the first
author. The satellite MODIS images were retrieved from the
MODIS NASA web site. We appreciate valuable discussions
with Ramon Pascual (AEMET) about this event, data and
comments provided by the Voluntary Spotter Team and the
Forecasting Team of the Meteorological Service of Catalonia,
and also suggestions by two anonymous reviewers which
helped to improve the final form of this paper.
References
Altaratz, O., Levin, Z., Yair, Y., 2001. Winter thunderstorms in Israel: a study
with lightning location systems and weather radar. Mon. Weather Rev.
129, 1259–1266.
Amaro, J., Llasat, M.C., Aran, M., 2010. The social impact of the snowfall of
8 March 2010 in Catalonia. Plinius Conference Abstracts 12, Plinius12-78,
12th Plinius Conference on Mediterranean Storms, Corfu, Greece.
Aran, M., Amaro, J., Arús, J., Bech, J., Figuerola, F., Gayà, M., Vilaclara, E., 2009.
Synoptic and mesoscale diagnosis of a tornado event in Castellcir,
Catalonia, on 18th October 2006. Atmos. Res. 93 (1–3), 147–160.
Aran, M.,Rigo, T., Bech, J.,Brucet, C., Vilaclara, E., 2010.Analysis of the hazardous
low-altitude snowfall, 8th March 2010, in Catalonia. Plinius Conference
Abstracts 2, Plinius12-77, 12th Plinius Conference on Mediterranean
Storms, Corfu, Greece.
Bech, J., Codina, B., Lorente, J., Bebbi ngton, D., 2003. The sensitivity of single
polarization weather radar beam blockage correction to variability in
the vertical refractivity gradient. J. Atmos. Oceanic Technol. 20,
845–855.
Bech, J., Pascual, R., Rigo, T., Pineda, N., López, J.M., Arús, J., Gayà, M., 2007. An
observational study of the 7 September 2005 Barcelona tornado
outbreak. Nat. Hazards Earth Syst. Sci. 7, 129–139.
Bech, J., Pineda, N., Rigo, T., Aran, M., Amaro, J., Gayà, M., Arús, J., Montanyà, J.,
van der Velde, O., 2011. A Mediterranean nocturnal heavy rainfall and
tornadic event. Part I: overview, damage survey and radar analysis.
Atmos. Res. 100, 621–637.
Boccippio, D.J., Cummins, K.L., Christian, H.J., Goodman, S.J., 2001. Combined
satellite- and surface-based estimation of the intracloud-cloud-to-
ground lightning ratio over the Continental United States. Mon. Weather
Rev. 129, 108–122.
Bosart, L.F., Sanders, F., 1986. Mesoscale structure in the megalopolitan
snowstorm of 11–12 February 1983. Part III: a large amplitude gravity
wave and coastal frontogenesis. J. Atmos. Sci. 43, 924–939.
Bosart, L.F., Edward Bracken, W., Seimon, Anton, 1998. A study of cyclone
mesoscale structure with emphasis on a large-amplitude inertia–gravity
wave. Mon. Weather Rev. 126, 1497–1527.
321J. Bech et al. / Atmospheric Research 123 (2013) 305–322
Cherington, M., 2001. Lightning injuries in sports: situations to avoid. Sports
Med. 31 (4), 301–308.
Cheze, J.L., Sauvageot, H., 1997. Area-average rainfall and lightning activity.
J. Geophys. Res. 102, 1707–1716.
Crowe, C.C., Market, P.S., Pettegrew, B.P., Melick, C.J., Podzimek, J., 2006. An
investigation of thundersnow and deep snow accumulations. Geophys.
Res. Lett. 33, L24812.
Cummins, K.L., Murphy, M.J., Bardo, E.A., Hiscox, W.L., Pyle, R.B., Pifer, A.E.,
1998. NLDN'95, a combined TOA/MDF technology upgrade of the US
National Lightning Detection Network. J. Geophys. Res. 103, 9035–9044.
Diendorfer, G., Kaltenböck, R., Mair, M., Pichler, H., 2006. Characteristics of
tower lightning flashes in a winter thunderstorm and related meteoro-
logical observations. 2006 19th International Lightning Detection
Conference (ILDC 2006), 24–25 April •Tucson, Arizona, USA.
Dixon, M., Wiener, G., 1993. TITAN: thunderstorm identification, tracking,
analysis, and nowcasting—a radar-based methodology. J. Atmos. Oceanic
Technol. 10, 785–797.
Dolif Neto, G., Market, P.S., Becker, A.E., Pettegrew, B., Melick, C., Schultz, C.,
Buckley, P.I., Clark, J.V., Lupo, A.R., Holle, R., Demetriades, N., Barbieri,
C.E., 2009. A comparison of two cases of low-latitude thundersnow.
Atmosfera 22 (3), 315–330.
Esteban, P., Jones, P.D., Martín-Vide, J., Mases, M., 2005. Atmospheric
circulation patterns related to heavy snowfall days in Andorra, Pyrenees.
Int. J. Climatol. 25, 319–329.
Glickman, T. (Ed.), 2000. Glossary of Meteorology, 2nd ed. Amer. Meteor.
Soc., Boston, US. 855 pp.
Hall, D.K., Riggs, G.A., Salomonson, V.V., DiGiromamo, N., Bayr, K.J., 2002.
MODIS snow-cover products. Remote. Sens. Environ. 83, 181–194.
Henson, W., Stewart, R., Kochtubajda, B., 2007. On the precipitation and
related features of the 1998 Ice Storm in the Montréal area. Atmos. Res.
83, 36–54.
HMSO, 1991. Meteorological Glossary, Sixth edition. HMSO, London, UK. 335 pp.
Hunter, S.M., Underwood, S.J., Holle, R.L., Mote, T.L., 2001. Winter lightning
and heavy frozen precipitation in the southeast United States. Weather
Forecast. 16, 478–490.
Johnson, J.T., MacKeen, P.L., Witt, A., Mitchell, E.D., Stumpf, G.J., Eilts, M.D.,
Thomas, K.W., 1998. The storm cell identification and tracking
algorithm: an enhanced WSR-88D algorithm. Weather Forecast. 13 (2),
263–276.
Kitagawa, N., Michimoto, K., 1994. Meteorological and electrical aspects of
winter thunderclouds. J. Geophys. Res. 99, 10713–10721.
Lauri, T., Koistinen, J., Moisseev, D., 2012. Advection based Adjustment of
Radar Measurements Mon. Wea. Rev. (in print).
Lojou, J.-Y., Cummins, K.L., 2006. Total lightning mapping using both VHF
interferometry and timeof-arrival techniques. Preprints, International
Lightning Detection Conference, April 24–25, Tucson, Arizona, Vaisala.
9 pp.
López, J., Montanyà, J., Maruri, M., De la Vega, D., Aranda, J.A., Gaztelumendi,
S., 2012. Lightning initiation from a tall structure in the Basque Country.
Atmos. Res. 117, 28–36.
Makkonen, L., 1998. Modeling power line icing in freezing precipitation.
Atmos. Res. 46 (1–2), 131–142.
Market, P.S., Halcomb, C.E., Ebert, R.L., 2002. A climatology of thundersnow
events over the contiguous United States. Weather Forecast. 17,
1290–1295.
Market, P.S., Oravetz, A.M., Gaede, D., Bookbinder, E., Lupo, A.R., Melick, C.J.,
Smith, L.L., Thomas, R., Redburn, R., Pettegrew, B.P., Becker, A.E., 2006.
Proximity soundings of thundersnow in the central United States.
J. Geophys. Res. 111, D19208.
Market, P.S., Ebert-Cripe, R.L., Bodner, M., 2007. Case study of a long-lived
thundersnow event. Natl. Weather Dig. 31, 103–119.
Martin, J.E., 1998. The structure and evolution of a continental winter
cyclone. Part II: frontal forcing of an extreme snow event. Mon. Weather
Rev. 126, 329–347.
Mateo, J., Andrés, A., Ballart, D., Figuerola, F., Miró, A., Aran, M., Pla, M., 2006.
Different precipitation-type in a snow event in Catalonia. A mesoscale
analysis. 6th Annual Meeting of the European Meteorological Society,
6th European Conference on Applied Climatology, Ljubljana, Slovenia,
4–8 September 2006.
Mazur, V., Ruhnke, L., Warner, T., 2010. Sustainable upward leaders from
tall structures: their development and effects. 30th International
Conference on Lightnin g Protectio n, 13th–17th September 2010 —
Cagliari, Italy.
Michimoto, K., 1991. A study of radar echoes and their relation to lightning
discharge of thunderclouds in the Hokuriku district, Part 1: observation
and analysis of thunderclouds in summer and winter. J. Meteorol. Soc.
Jpn. 69, 327–335.
Michimoto, K., 1993. A study of radar echoes and their relation to lightning
discharge of thunderclouds in the Hokuriku district, Part 2: observation
and analysis of “single flash”thunderclouds in midwinter. J. Meteorol.
Soc. Jpn. 71, 195–203.
Miyake, K., Suzuki, T., Takashima, M., Takuma, M., Tada, T., 1990. Winter
lightning on Japan Sea coast-lightning striking frequency to tall structures.
IEEE Trans. Power Deliv. 5, 1370–1376.
Montanya, J., Pineda, N., March, V., Illa, A., Romero, D., Sola, G., 2006. Experimental
evaluation of the Catalan Lightning Detection Network. 19th International
Lightning Detection Conference (ILDC), 24–25 April 2006, Tucson, AZ.
Munzar, J., Franc, M., 2003. Winter thunderstorms in central Europe in the
past and the present. Atmos. Res. 67–68, 501–515.
O'Hora, F., Bech, J., 2007. Improving weather radar observations using
pulsecompression techniques. Meteorol. Appl. 14 (4), 389–401.
Pagès, M., Aran, M., Pineda, N., Andrés, A., Bech, J., 2003. Low altitude
snowfall events in Catalonia. European Geosciences Union. 5th Plinius
Conference on Mediterranean Storms. Ajaccio, Corsica, France.
Pascual, R., 2011. Temporal de nieve con tormenta del 8 de marzo de 2010 en
Cataluña. Revista del Aficionado a la Meteorología (RAM). http://www.
meteored.com/ram/13534/ (in Spanish).
Pascual, R., Callado, A., Berenguer, M., 2003. Analysis of a strong snowstorm
in Catalonia on December 2001. Proc. 31st Conference on Radar
Meteorology, Seattle, US, 6–12 August 2003, pp. 882–884.
Pettegrew, B.P., Market, P.S., Wolf, R.A., Holle, R.L., Demetriades, N.W.S.,
2009. A case study of severe winter convection in the midwest. Weather
Forecast. 24, 121–139.
Pineda, N., Montanya, J., 2009. Lightning detection in Spain:the particular case
of Catalonia. In: Betz, H.-D., Schumann, U., Laroche, P. (Eds.), Lightning:
Principles, Instruments and Applications. Springer, pp. 161–185.
Pineda, N., Rigo, T., Bech, J., Soler, X., 2007. Lightning and precipitation
relationship in summer thunderstorms: case studies in the North
Western Mediterranean region. Atmos. Res. 85 (2), 159–170.
Pineda, N., Bech, J., Rigo, J., Montanyà, J., 2011a. A Mediterranean nocturnal
heavy rainfall and tornadic event. Part II: total lightning analysis. Atmos.
Res. 100, 638–648.
Pineda, N., Soler, X., Vilaclara, E., 2011b. An approach to a lightning
climatology in Catalonia. Nota d'Estudi, 73. Meteorological Service of
Catalonia, Barcelona, Spain (In Catalan). 71 pp.
Prohom, M., Herrero, M., 2008. Towards the creation of a climatic database
for Catalonia (18th to 21st centuries). Tethys 5, 3–12.
Proulx, G.J., 1971. Standard Dictionary of Meteorologica l Sciences. McGill-Queens
University Press, Montreal, Quebec, Canada. 307 pp.
Rakov, V.A., Uman, M.A., 2003. Lightning: Physics and Effects. Cambridge
University Press, p. 687.
Rigo, T., Pineda, N., Bech, J., 2010. Analysis of warm season thunderstorms
using an object-oriented tracking method based on radar and total
lightning data. Nat. Hazards Earth Syst. Sci. 10, 1881–1893.
Sánchez-Diezma, R., Sempere-Torres, D., Bech, J., Velasco, E., 2002. Development
of a hydrometeorological flood warning system (EHIMI) based on radar data.
2nd European Radar Conference. Copernicus Gesellschat, Delft, Holland.
Sanders, F., Bosart, L.F., 1986. Mesoscale structure in the metropolitan
snowstorm, 11–12 February 1983. Part II: Doppler radar study of the
New England Snowband. J. Atmos. Sci. 42, 1398–1407.
Savadjiev, K., Farzaneh, M., 2004. Modeling of icing and ice shedding on
overhead power lines based on statistical analysis of meteorological
data. IEEE Trans. Power Deliv. 19 (2), 715–721.
Schultz,D.M., 2009. Lake-effect snowstorms in northernUtah and westernNew
York with and without lightning. Weather Forecast. 14, 1023–1031.
Schultz, D.M., Vavrek, R.J., 2009. An overview of thundersnow. Weather 64
(10), 274–277.
Siggia, A.D., Holmes, J.M., 1991. One pass velocity unfolding for VVP analysis.
Preprints 25th Conf. on Radar Meteorology. Amer. Meteor. Soc., Paris,
France, pp. 882–884.
SMC, 2010. Monthly Climatic Bulletin. March 2010 Meteorological Service of
Catalonia (SMC). http://www.mateo.cat,section‘Climatology’, subsection
‘Climatic bulletins’(accessed in November 2011). (In Catalan).
Takeuchi, T., Nakano, M., Brook, M., Raymond, D.J., Krehbiel, P., 1978. The
anomalous winter thunderstorms of the Hokuriku coast. J. Geophys. Res.
83, 2385–2394.
Vaisala, 2008. IRIS Product & Display Manual. Vaisala, Westford, MA, USA.
Vilaclara, E., Segalà, S., Andrés, A., Aran, M., 2010. Operational warnings
issued by the SMC in the 8th March snow event in Catalonia. Plinius
Conference Abstracts 12, Plinius 12–79, 2010, 12th Plinius Conference
on Mediterranean Storms, Corfu, Greece.
WMO, 2011. METEOTERM Web page. World Meteorological Organization
(WMO). http://www.wmo.int/pages/prog/lsp/meteoterm_wmo_en.html
(accessed in November 2011).
322 J. Bech et al. / Atmospheric Research 123 (2013) 305–322