Atmospheric Research 100 (2011) 310317

Contents lists available at ScienceDirect

Atmospheric Research
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / a t m o s

A climatology of cloud-to-ground lightning over Estonia, 20052009

S.E. Enno
Department of Geography, University of Tartu, Vanemuise 46, 51014 Tartu, Estonia

a r t i c l e

i n f o

a b s t r a c t
This paper presents the spatial and temporal distribution of cloud-to-ground lightning over Estonia and the adjacent sea for the period 20052009. Data collected by the NORDLIS (NORDic Lightning Information System) lightning detection network was used. The spatial distribution of lightning was calculated in a 10 10 km grid. A total of 172,613 cloud-to-ground ashes were registered in the area under observation during 20052009. The annual mean cloud-toground ash density over the area was 0.34 ashes km 2 year 1. The lowest values found were less than 0.10 ashes km 2 year 1 and the highest ash densities were 0.801.01 ashes km 2 year 1. The monthly distribution of lightning showed the highest activity in July and August. Of all the registered ashes, 99.4% were reported from May to October. The daily distribution of ashes showed single days on which thunderstorm activity was very high, against a background of much lower everyday activity. The diurnal distribution of lightning showed an evident peak between 15:00 and 17:00 local time over land. Over the sea, a atter maximum lay between 13:00 and 21:00. Minimum lightning activity occurred between 22:00 and 06:00 over land and from 02:00 to 09:00 over the sea. Our work revealed that the spatial and temporal characteristics of cloud-to-ground lightning over Estonia generally resemble the characteristics found at other mid-latitude study sites. 2010 Elsevier B.V. All rights reserved.

Article history: Received 12 January 2010 Received in revised form 20 August 2010 Accepted 27 August 2010 Keywords: Detection Cloud-to-ground lightning Flash density Estonia

1. Introduction Thunderstorms are among the most damaging weather events. Thus, it is useful to know the spatial and temporal characteristics of the occurrence of thunderstorms and lightning. Nowadays, a variety of data sources are available for use in studying the spatial and temporal distribution of thunderstorms. Visual observations at meteorological stations are the oldest available records. At many locations, the data for annual and monthly numbers of thunderstorm days go back further than 100 years, thereby making long-term climatic studies possible. For example, Changnon and Changnon (2001) used data of 86 stations for the period 18961995 to study long-term uctuations in the annual numbers of thunderstorm days in the contiguous United States. However, it has been demonstrated that the intensity of thunderstorms, as well as exterior factors, such as the
Tel.: + 3727375824; fax: + 3727375825. E-mail address: sven-erik.enno@ut.ee. 0169-8095/$ see front matter 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.atmosres.2010.08.024

background noise and daylight, signicantly affect visual observations (Reap and Orville, 1990). Hence, it is wise to use data sets that are more reliable. Lightning detection networks allow the collection of continuous lightning data that has high spatio-temporal accuracy and enable discrimination between cloud-to-ground and cloud ashes. The data is processed and made available in real time. Cummins et al. (1998) describe in detail the different lightning sensors and properties of lightning detection networks. The development of national and international lightning detection systems over the last 20 to 30 years has made it possible to study thunderstorm climates by using ash density as the main measure. Flash density is the average number of (usually cloud-to-ground) lightning ashes per area unit per time unit. The measure that is used most widely is the annual number of ashes per square kilometre. It is more accurate than the traditionally used average annual number of thunderstorm days, because it not only reects the presence of thunderstorms, but also gives an overview of the number and spatial distribution of lightning ashes.

S.E. Enno / Atmospheric Research 100 (2011) 310317

311

Diurnal (hour-to-hour), seasonal (day to day or month-tomonth) and annual (year-to-year) variations in lightning activity are widely used for temporal analyses (Tuomi and Mkel, 2008a). It is also possible to calculate the annual number of thunderstorm days or thunderstorm hours from the lightning detection data. Hufnes and Orville (1999) did so for the USA, while Rivas Soriano and De Pablo (2002b) made similar calculations for the Iberian Peninsula. Many reports of regional studies on the spatial and temporal distribution of lightning have been published. The most widely used spatial grid spacing for calculating ash densities seems to be 0.2 0.2. This spacing has been used for the USA (Orville and Hufnes, 2001), the USA and Canada (Orville et al., 2002), the Iberian Peninsula (Rivas Soriano et al., 2005) and Sweden (Sonnadara et al., 2006). However, the areas of the cells in this kind of grid are unequal; they decrease towards high latitudes. Many authors prefer equalarea grid cells. Burrows et al. (2002) used a 20 20 km grid for Canada, and Antonescu and Burcea (2010) used the same grid spacing for Romania. Tuomi and Mkel (2008a) used a 10 10 km spatial grid to calculate ash densities when analyzing the thunderstorm climate of Finland. Schulz et al. (2005) used 10 10 and 1 1 km grids for geographical plots when studying the cloud-to-ground lightning over Austria from 1992 to 2001. The 1 1 km grids were too small, because individual radio towers and mountain summits create noise that clearly affect the results. Relationships between lightning activity and other atmospheric factors or synoptic situations have also been examined extensively in modern lightning climatology. Tuomi and Mkel (2008a) studied the relationships between synoptic weather types and lightning activity. A similar analysis was made for the Iberian Peninsula by Clemente et al. (2004). Tuomi and Larjavaara (2005) used lightning detection network data to identify and analyze ash cells in thunderstorms. Characteristics of cloud-to-ground ashes over the Iberian Peninsula were compared with geographical latitude and longitude (Rivas Soriano et al., 2002), and with the sea surface temperature (De Pablo and Rivas Soriano, 2002). The effect of urban pollution on cloud-to-ground lightning

activity was analyzed for the Midwestern USA (Westcott, 1995) and central Spain (Rivas Soriano and De Pablo, 2002a). The rst weather and thunderstorm records in the archive of the Estonian Meteorological and Hydrological Institute (EMHI) date back to the 18th century, but regular observations with continuous data rows were started in the second half of the 19th century. These records also include the occurrence of thunderstorms. The annual numbers of thunderstorm days for the entire 20th century are available at some weather stations. However, until 2005, the only type of thunderstorm data that was collected in Estonia was visual records. Many local-scale studies have been written on the basis of visual thunderstorm observations. It is known that, on average, there are 1525 thunderstorm days per year in Estonia. Unfortunately, none of the Estonian thunderstorm studies have so far been published internationally. The study reported herein yielded the rst results for the spatial and temporal distribution of cloud-to-ground lightning over the Estonian landmass and the adjacent sea, including islands. Estonia was incorporated into the NORDLIS lightning detection network at the end of 2004. Thus, in the interests of using the most reliable data, data from 2005 to 2009 were used. The remainder of the paper is organized as follows. In Section 2, the data and methods are described. In Section 3, the results are presented. In Section 4, the results are discussed. Section 5 concludes. 2. Data and methods The study reported herein used data from the NORDLIS, which incorporates lightning sensors in Norway, Sweden, Finland, and Estonia. The central unit of the system operates in Finland and belongs to the Finnish Meteorological Institute (FMI). Fourteen sensors in Norway, nine in Sweden, seven in Finland, and one in Estonia are connected to the central unit (Fig. 1). The system uses mainly IMPACT sensors, with some of its successor model, the LS7000, also deployed. These sensors use low-frequency electromagnetic radiation to detect lightning. NORDLIS is mainly capable of locating

Fig. 1. a) Location of the NORDLIS lightning detection network (black rectangle). b) Locations of the NORDLIS sensors in 2009 (black circles) and our study area (black rectangle).

312

S.E. Enno / Atmospheric Research 100 (2011) 310317

cloud-to-ground lightning, because intracloud lightning emits VLF/LF radiation mainly in the range of smaller amplitudes and with waveforms that do not pass the applied discrimination lter. Tuomi (2005) and Cummins et al. (1998) describe the properties of IMPACT sensors in greater detail. The lightning detector was installed in Estonia in the end of 2004, so from 2005, the efciency of the network was increased over Estonia. The initial dataset that we used was obtained from the EMHI; it consists of 228,997 ashes registered during the ve-year period 20052009. All ashes typed as cloud lightning were removed from the dataset, to ensure that, as far as is possible, our nal dataset contained only records of cloud-toground lightning. However, it should be noted that, especially for stroke currents below 15 kA, the system cannot discriminate reliably between intracloud and cloud-to-ground strokes. A 10 10 km spatial grid was used to calculate ash densities. Although Schultz et al. (2005) demonstrate that larger grids give more accurate results, we decided to use a 10 10 km grid, because the study area is quite small, only 410 250 km. The same grid size worked well for the NORDLIS data analysis in Finland 19982007 (Tuomi and Mkel, 2008a). Yearly lightning maps with this grid spacing were also published by Tuomi and Mkel (2006), Tuomi and Mkel (2007), and Tuomi and Mkel (2008a,b,c). The grid that we used is based on the Estonian coordinate system LEST97, which makes it possible to use equal-sized square cells. The grid covered 102,500 km2 between 57.5 and 59.8 N and 21.0 and 28.5 E. The study area encompasses Estonia, the eastern part of the Baltic Sea, the southern part of the Gulf of Finland, Northern Latvia, and the western edge of Russia (Fig. 1). A total of 172,613 cloud-to-ground ashes were registered in the study area in the period 20052009. All the data was imported to Idrisi 32 software, where the spatial grid was created and all calculations of ash density were done. The results of the calculations were imported to ArcGIS, where the nal maps were designed. The number of ground ashes per square kilometre was used as the main characteristic of ash density. Flash density was analyzed for the whole ve-year period, and year by year and month by month. No corrections for detection efciency were made in this analysis. The rst-stroke detection efciency of the NORDLIS network in Finland is estimated to range from 88% to over 99% for cloud-to-ground lightning (Tuomi and Mkel, 2008c). Lowest efciency was found for Lapland near the northern edge of the NORDLIS network, but highest efciency was found in Southern Finland, and Estonia is only 100400 km from Southern Finland. This would indicate that provided network coverage is similar, we could expect levels of efciency similar to that achieved in Southern Finland. On the other hand, no NORDLIS sensors have been installed in the south and east of Estonia. As a consequence, the efciency with which cloud-to-ground ashes are detected in the study area probably decreases slightly from the north to the south and the west to the east. In 2008, one sensor was removed from Western Finland, because it was broken. Accordingly, it is possible that a small drop in detection efciency has occurred since 2008. However, this possibility is not signicant for our study, because the four sensors on the southern coast of Finland that are closest to Estonia were working throughout the whole study period.

It is possible to estimate approximately the actually relevant detection efciency over a given area by knowing the average peak currents of observed ashes as well as the peak current distribution of ashes (Pinto, 2008). For the study area, during 20052009 the average peak currents were 17.7 and 19.9 kA for negative and positive ashes, respectively. The current distribution of registered ashes is presented in Fig. 2. In addition to the spatial distribution, the temporal distribution of lightning was studied month by month and day by day. Diurnal variations in the occurrence of ashes were analyzed for the whole study area and also separately for land and the sea. All ashes over the islands were added to the sea data and all ashes over inland lakes were counted as land data. As justication for this measure, we determined that the diurnal lightning distribution over the biggest island, Saaremaa (area 2700 km2), was similar to the sea areas, and over the biggest lake, Lake Peipsi (area 3500 km2), it resembled that of the land areas. 3. Results The annual average ash density for the whole study area for the period 20052009 was found to be 0.34 ashes km 2 year 1. The interannual variability is remarkable. In 2006, only 17,000 ashes were registered (ash density 0.17 ashes km 2 year 1). In 2008, 23,500 and in 2005 about 25,000 ashes were registered over the study area (ash densities 0.23 and 0.24 ashes km 2 year 1, respectively). Many more thunderstorms occurred in 2007 and 2009. In 2009, 50,000 ashes with an average density of 0.49 ashes km 2 year 1 were registered. In 2007, more than 57,000 recorded ashes gave an average density of 0.56 ashes km 2 year 1. The spatial distribution of ground ashes over Estonia during the period 20052009 (Fig. 3) is rather complicated. The sea areas show lower densities than the land areas. The lowest lightning frequencies (less than 0.1 ashes km 2 year 1) were observed at the western edge of the study area over the open waters of the Baltic Sea. Most sea areas showed values less than 0.4 ashes km 2 year 1 and no more than 0.5 ashes km 2 year 1 were registered on the open sea and the islands.
12

10 Number of flashes (thousands)

0 -100

-80

-60

-40

-20

20

40

60

80

100

Peak current kA

Fig. 2. Peak current distribution of registered cloud-to-ground ashes in the study area during the period 20052009.

S.E. Enno / Atmospheric Research 100 (2011) 310317

313

Fig. 3. Annual average ash density (ashes km 2 year 1) in the study area during the period 20052009.

However, the eastern part of the largest island Saaremaa showed a peak with 0.61 ashes km 2 year 1. Land areas and also some coastal districts exhibit higher ash densities. On land, 0.30.6 ashes km 2 year 1 is common and the ash density generally increases from the southwest to the northeast. Maxima with 0.60.8, in some places 0.81.01, ashes km 2 year 1 are mostly located in the northeastern part of the Estonian land area. The only exception is a small area with 0.60.9 ashes km 2 year 1 on the southwestern coast. Another interesting feature is the fact that high ash densities on the northeastern land areas continue over the waters of the Gulf of Narva towards the Russian Kurgolovo Peninsula (the northeastern corner of the study area). The monthly distribution of registered ashes (Fig. 4) shows the main thunderstorm season from May to August. The most active months of July and August both give about 28% of all registered ashes, so more than half of all the ashes during the period 20052009 occurred in these 2 months. Roughly 21% of the ashes occurred in May and 14% in June. More than 90% of all the ashes occurred from May to August. About 4% of the annual number of ashes occurred in each of September and October, 0.4% in April, and 0.2% in November. Almost no thunderstorms occurred from
60 Number of flashes (thousands) 50 40 30 20 10 0 I II III IV V VI VII VIII IX X XI XII Month

December to March. In December and January, only eight ashes were registered over the whole study area during the 5-year period. During the period 20052009, nine ashes occurred in March and 43 in February. Given that almost no thunderstorms were recorded between November and March, the average daily numbers of ashes over the study area (Fig. 5) are given from April to October. For each date, the number in the gure denotes a ve-year average. It can be seen that the main thunderstorm season begins in mid-May and lasts until the beginning of September. However, some events occurred outside of the main season. These include some days with fairly high thunderstorm activity at the end of September and at the beginning of October. Nevertheless, it is obvious that thunderstorms are not frequent in April, September, and October. The main season is characterized by an average daily activity of about 100200 ashes over the study area. In addition, there are a lot of individual peaks with 5001500 ashes per date. Two extremely high peaks are visible. 13 June has an average activity of 2700 ashes per day and 19 July has 2300 ashes per day. Every strong peak can be

Average number of flashes (thousands)

3.0 2.5 2.0 1.5 1.0 0.5 0.0 APR MAY JUN JUL Date AUG SEPT OCT

Fig. 4. Monthly distribution of ashes during the period 20052009.

Fig. 5. Daily average numbers of registered ashes (in thousands) from April to October over the study area during the period 20052009.

314

S.E. Enno / Atmospheric Research 100 (2011) 310317

associated with a single heavy storm that occasionally occurred on that date in one of the years studied. This inuence of a single heavy storm to the lightning statistics is discussed further in Section 4, with some examples. The diurnal distribution of lightning (Fig. 6) differs remarkably in the land and sea areas, so both distributions are given. Over land, an evident peak occurred between 15:00 and 17:00 local time. Each of the hours at which the peak occurs contains more than 10% of all registered ashes. It may be concluded that one third of all registered ground ashes over land occurred between 15:00 and 17:00. During the evening hours, there is a rapid decrease in lightning activity. The minimum activity occurred between 22:00 and 06:00, when each hour contained only 12% of all registered ashes. Hence, the lightning activity over land during the night-time minimum period is roughly 10 times lower than during the daytime maximum period. Diurnal variation is much smaller over the sea. The maximum is at and lies between 13:00 and 21:00 with an evident peak at 14:00. It is quite difcult to determine the end of the maximum, because activity continues to be quite high until 01:00. The minimum occurs from 02:00 to 09:00. Each of the hours during which maximum activity occurs contains 6 8% and each of the minimum hours 23% of all registered ashes over the sea. So, lightning activity over the sea during the night and morning minimum is about 23 times lower than during the day and evening maximum. 4. Discussion The average annual ash density over the whole study area was 0.34 ashes km 2 year 1 and varied from less than 0.10 ashes km 2 year 1 to 1.01 ashes km 2 year 1. These values are low on a global scale. Similar lightning detection networks register up to 68 ashes km 2 year 1 in northeastern Italy (Bernardi and Ferrari, 2004) and over 9 ashes km 2 year 1 in Florida, USA (Orville and Hufnes, 2001). So, our ash densities are about 2030 times smaller than the highest measured densities in Europe and the USA. The same average ash density of 0.34 ashes km 2 year 1 was found for Finland during 19982007 by Tuomi and Mkel (2008a). It should be taken into account that the Finnish study area does not include open sea low-activity

areas, whereas our study area does. It is also worth mentioning that the years 1998 to 2002, which probably experienced much greater thunderstorm activity, were included by Tuomi and Mkel (2008a), but they are not represented in our dataset. Years 20052008 exhibited ash densities below 0.2 ashes km 2 year 1 over the Finnish study area, which is well below the 10-year average, as well as below the Estonian values for the same years. So it can be concluded that thunderstorm activity over Estonia is higher than over Finland. It is also obvious that Sweden experiences fewer ashes than Estonia. During the period 19872000, the average annual ash density over Sweden varied from 0.03 ashes km 2 year 1 in the north to 0.4 ashes km 2 year 1 near the southern coast of the country (Sonnadara et al., 2006). So, the most active locations in southern Sweden experience as many ashes as the average in the Estonian study area. The annual numbers of registered ashes vary up to ve times between individual years during the 13-year period studied by Sonnadara et al. (2006). Similar high yearto-year variations are also characteristic for Estonia. Ground ash density comparable with Estonia was also found by Antonescu and Burcea (2010) for eastern and northwestern Romania. In North America, ash densities similar to Estonia can be found in northern and western parts of Canada (Orville et al., 2002) and in western and northeastern areas of the USA (Orville and Hufnes, 2001). The spatial distribution of ashes (Fig. 3) is difcult to explain in terms of the main topographic features, such as uplands. The 5-year study period is probably too short and the uplands are too low. The Finnish 10-year study period does not show a correlation between ash density and topography either (Tuomi and Mkel, 2008a), so an even longer period would be needed to establish one. However, our study shows that open waters of the Baltic Sea in the western edge of the study area have the lowest ash densities. The water in this region is a relatively cool surface during summertime with little or almost no spatial variations in temperature; hence it is not conducive to convection and thunderstorms. In contrast, the Gulf of Finland, which lies between land areas of Estonia and Finland, does not inhibit thunderstorms as much as the open sea. Very high ash densities are observed over the eastern part of the gulf. However, these storms are probably not initiated by the sea,

Fig. 6. Diurnal distribution of ashes over land (left) and over sea (right). 14 h local time is 12 h UTC.

S.E. Enno / Atmospheric Research 100 (2011) 310317

315

but formed over northeastern Estonia. Due to the fact that winds from the south and southwest are common, these storms often move to the gulf and frequently continue to produce many ashes over the water. Northeastern Estonia seems to be conducive to the development of thunderstorms. There are large swamps that can supply warm humid air, as well as uplands that can force updrafts. A previous, visual observations-based local study indicated that the highest annual numbers of thunderstorm days occur in the northeastern part of the country. However, the cause of the prevalence of thunderstorms in this area is not immediately evident. Other upland areas in southern and southeastern Estonia do not show lightning maxima, even though they are higher than the northeastern uplands, which would indicate that the prevalence of thunderstorms in northeastern Estonia is not due to the presence of uplands. In addition, one maximum lies just on the southwest coast, on coastal lowland, which would conrm that the prevalence of thunderstorms is not associated with the presence of uplands. However, this maximum is probably associated with one very powerful storm in that region on 19 July 2009, and thus it is not a long-term climatic feature. It is evident that the available data is not currently sufcient to account for the prevalence of thunderstorms over northeastern Estonia. Further data is needed. It is to be hoped that such further data will minimize the inuence of single strong thunderstorm events and enable us to improve our map of the distribution of ashes according to climate. Longer data rows will also mean more ashes per grid cells and hence a higher statistical accuracy for values for the density of ground ashes (Diendorfer, 2008). The monthly distribution of lightning (Fig. 4) is not very typical, but its main properties are still similar to those found in other mid-latitude studies. July was found to be the most active month. The same was found for Finland (Tuomi and Mkel, 2008a), Sweden (Sonnadara et al., 2006), Austria (Schulz et al., 2005), Canada, (Burrows et al., 2002) and the USA (Orville and Hufnes, 2001). July is the warmest month in mid-latitude continental areas, so the development of a greater number of active thunderstorms in this month is rendered likely by the high temperatures. Other mid-latitude studies found that June and August have clearly lower activities than July, and that in May activity is markedly lower than June and August. These results are, at rst sight, in contrast to those found in our study. We found that August is almost as active as July and that May has a markedly higher number of registered ashes than June. However, these initial discrepancies can be explained away. The high activity that we found in August may be attributed, at least in part, to the inuence of the sea. Sea water attains its highest temperature in August, which has been recorded to be the warmest month at some areas of the coast in Estonia. Hence, the highest lightning activity over the sea is expected to occur not in July, but in August, and the other studies did not cover open sea areas. The high lightning activity that we found in May seems to be occasional; it mainly comprises unusually powerful and wide storms during the last days of May 2007. Hence, it would be unwise to take the results for May as being representative of activity in May in general. In contrast with May, June

showed unusually low lightning activity during the period 20052009. For the same period, unusually low activity in June is also reported for Finland by Tuomi and Mkel (2008a). If the high activity during the last days of May 2007 and the low activity during June is taken to be an anomaly, our results will be in agreement with those of other studies. Of all the ashes that were registered during the study period, 95.6% occurred from May to September and 99.4% from May to October. These results are very similar to results obtained for Austria, Romania, and Canada. Schulz et al. (2005) reported that 96% of ashes in Austria during the period 19922001 occurred from May to September. Antonescu, and Burcea (2010) found that about 98% of all registered lightning in Romania during the period 2003 2005 and in 2007 was detected from May to September. In Western Canada, 98.9% and in Eastern Canada, 93.0% of ashes recorded from 1998 to 2000 occurred in the 6 warm months from May to October (Burrows et al., 2002). From November to March, almost no thunderstorms occurred in Estonia. The mean temperatures for these months are below zero, so vertical air movements do not occur and as a result, thunderclouds do not form. Only occasional ashes produced by powerful cold fronts are possible. The daily distribution of ashes (Fig. 5) shows that the main thunderstorm season runs from the middle of May to early September. During this period, 100200 ashes per day occur over the study area. Such activity is characteristic of local storms. The ashes are often clustered in small parts of the study area, whereas no thunderstorms occur in the rest of the area. Frontal storms produce much more ashes, because they cover much larger areas, sometimes affecting the whole study area. As seen in Fig. 5, they cause individual strong peaks even in the 5-year average for daily activity. Most days that average more than 1000 ashes over the study area are associated with the occurrence of individual heavy frontal storms in one of the years from 2005 to 2009. Dramatic examples are 13 June and 19 July 2009, on which 13,000 and 9500 ashes occurred, respectively. These two days together produced about 45% of the total ground ashes registered in 2009. All the increased activity at the end of May is due to powerful frontal storms that occurred from 26 to 31 May 2007. There was another outbreak of heavy storms from 21 to 24 August 2007. One late heavy storm, which is easily recognizable in Fig. 5, occurred on 1 October 2006. Within a few hours, it produced about one third of the total number of ashes that were registered in 2006. Thus, in Estonia, most of the annual lightning activity seems to be concentrated in a few very active days. Similar results are reported for Canada. In southern Ontario, the 4 most active days in 1998 and 1999 accounted for more than 25% of annual lightning. In some places, about half of the annual lightning in 1998 occurred in a 26-hour period and 25% in just 4 h (Burrows et al., 2002). All strong storms in the Estonian study area that occurred in the period from 2005 to 2009 are associated with fronts. In a usual situation, the advection of warm and humid air from the south is followed by a cold front from the west, producing powerful and widespread thunderstorms. Due to the fact that the study area is quite small, it takes less than 24 h for most fronts to cross it. So on one day, an activity peak can appear in the lightning data, while there is no or very weak activity during the previous and next days. Only stationary or very

316

S.E. Enno / Atmospheric Research 100 (2011) 310317

slow-moving fronts can generate very active periods that last for 25 days, as happened in May and August 2007. The daily distribution of lightning was also published for Finland by Tuomi and Mkel (2008a). The report was based on 10-year data, so the results are more homogenous. Many individual peaks are still visible, but these do not stand out as much as the peaks that we found for Estonia for 2005 to 2009. We expect that over a longer timescale, we would nd a more even daily distribution of lightning over Estonia, in which the effect of individual storms would be lessened. Overall, individual powerful storms do not have a strong effect on the overall 5-year spatial distribution of ash density. They tend to produce more lightning over land and in northeastern Estonia, and less over the coastal and openwater areas. In so acting, they act as local storms. The only exception in our study area during the study period is the heavy storm on 19 July 2009. On that afternoon, an unusually high number of ashes were produced at some locations on the southwest coast by a strong frontal storm. As mentioned earlier, the high activity on the southwestern coastal area (Fig. 3) was caused by that storm. The storm peaked at around 14:00 local time and a lot of ashes also hit the sea. The single peak that is visible at 14:00 in the diurnal distribution of ashes over the sea (Fig. 6) was also caused by that storm. The diurnal distribution of ashes (Fig. 6) is much more even over the sea. This phenomenon is probably associated with the fact that the sea surface temperature varies little or not at all between day and night. The distribution of land ashes correlates well with the diurnal temperature cycle over land. A strong peak occurred between 15:00 and 17:00 local time, just after the usual daily maximum temperatures. This peak time corresponds well with the results of other similar studies conducted in Finland (Tuomi and Mkel, 2008a), Austria (Schulz et al., 2005), the Iberian Peninsula (Rivas Soriano et al., 2005), Romania (Antonescu and Burcea, 2010), and Canada (Burrows et al., 2002). Minimum activity is usually found to occur during the night and in the early morning. This is also the case in the Estonian land area. The diurnal distribution of lightning over the sea in Finland was studied by Tuomi and Mkel (2008a). Their results are quite similar to ours. The maximum is at and elongated towards the evening hours. However, at least the rst half of the maximum over the sea occurred at the same time the land maximum. It is evident that some storms develop over land, but then drift to the sea, thus it is not surprising that during the land maximum hours, the sea areas, especially near the coast, also experience a high number of ashes. As the water maintains its temperature, the dissipation of storms during evening hours is slower. That can explain why the diurnal maximum is elongated towards night over the sea.

5. Conclusions We have provided an overview of the spatio-temporal distribution of cloud-to-ground lightning over Estonia and the adjacent sea during the period 20052009. The total study area covers 102,500 km2 in Estonia, the eastern part of the Baltic Sea, the southern part of the Gulf of Finland, Northern

Latvia, and the western edge of Russia. Flash densities were calculated for a 10 10 km spatial grid. A total of 172,613 cloud-to-ground ashes were registered in the study area during the study period. About three times as many ashes occurred during the high-activity years of 2007 and 2009 than in 2006. The annual mean cloud-to-ground ash density over the area was 0.34 ashes km 2 year 1. The lowest values of less than 0.10 ashes km 2 year 1 were found over the open waters of the Baltic Sea in the western edge of the study area. Flash densities over the sea were generally lower than over land. The only exception was the eastern part of the Gulf of Finland, where remarkably high densities were found near the northeast coast of Estonia. The highest ash densities of 0.801.01 ashes km 2 year 1 were concentrated in northeastern Estonia. All these values are low on a global scale and are comparable with ash densities in Finland, Sweden, western USA, and northern and western Canada. No clear correlation between landforms and lightning activity appeared. Study of data that covers longer periods is required to specify the climatic distribution of ashes in more detail and demonstrate the statistical accuracy of results. The mid-latitude specic annual distribution showed 99.4% of annual ashes occurring from May to October and 95.6% occurring from May to September. July and August were the most active months. Together, about 56% of the total annual number of ashes occurred during these months. May showed high and June unusually low activity. Almost no thunderstorms occurred from November to March. The average daily numbers of ashes over the study area show the main thunderstorm season running from the middle of May to early September. However, some days on which activity was quite high lay outside the main season, at the end of September and at the beginning of October. The main season is characterized by an average daily activity of about 100200 ashes over the study area. During this period, there are high peaks of activity that are associated with strong and wide frontal storms. Up to 13,000 ashes have been registered during one day. Just a few days can yield more than half of the total number of ashes registered annually. In a typical scenario that is characterized by high activity, advection of warm and humid air from the south is followed by a cold front from the west, which produces powerful thunderstorms that cover a large area. The diurnal distribution of lightning differs markedly between the land and the sea areas. Over land, an evident peak occurred between 15:00 and 17:00 local time. One third of all registered ashes over land occurred from 15:00 to 17:00. Minimum lightning activity occurred between 22:00 and 06:00. These times correspond well with the results of other similar studies conducted on mid-latitude countries. Lightning activity over land during the night-time minimum was roughly 10 times lower than during the daytime maximum. Diurnal variations were much smaller over the sea. The maximum is at and lies between 13:00 and 21:00. The minimum occurs from 02:00 to 09:00. Due to the fact that the water maintains its temperature at night, the dissipation of storms during the evening hours is slower over the sea. This dissipation might explain why the diurnal maximum is elongated towards night over the sea.

S.E. Enno / Atmospheric Research 100 (2011) 310317

317

Acknowledgements The author would like to thank Tapio J. Tuomi and Antti Mkel from the Finnish Meteorological Institute for helpful materials and advice. The author also would like to thank Dr. David M. Schultz and two anonymous reviewers for their helpful comments. This research was supported by the Estonian Science Foundation grant No. 7510. References
Antonescu, B., Burcea, S., 2010. A cloud-to-ground lightning climatology for Romania. Mon. Weather Rev. 138, 579591. Bernardi, M., Ferrari, D., 2004. Evaluation of the LLS efciency effects on the ground ash density, using the Italian lightning detection system SIRF. J. Electrostat. 60, 131140. Burrows, W.R., King, P., Lewis, P.J., Kochtubajda, B., Snyder, B., Turcotte, V., 2002. Lightning occurrence patterns over Canada and adjacent United States from lightning detection network observations. Atmos. Ocean 40, 5981. Changnon, S.A., Changnon, D., 2001. Long-term uctuations in thunderstorm activity in the United States. Clim. Change 50, 489503. Clemente, T., De Pablo, F., Rivas Soriano, L., 2004. Circulation weather types and cloud-to-ground ash density over the Iberian Peninsula. Int. J. Climatol. 24, 10912310.1002/joc.917 Cummins, K.L., Krider, E.P., Malone, M.D., 1998. The U.S. National Lightning Detection Network and applications of cloud-to-ground lightning data by electric power utilities. IEEE Trans. Electromagn. Compat. 40, 465480. De Pablo, F., Rivas Soriano, L., 2002. Relationship between cloud-to-ground lightning ashes over the Iberian Peninsula and sea surface temperature. Q. J. R. Meteorol. Soc. 128, 173183. Diendorfer, G., 2008: Some comments on the achievable accuracy of local ground ash density values. Proceedings of the 29th International Conference on Lightning Protection (ICLP 2008), Uppsala, Sweden 2326 June 2008. (available at http://www.aldis.at/pdf/12_ICLP_2008_DIN.pdf). Hufnes, G.R., Orville, R.E., 1999. Lightning ground ash density and thunderstorm duration in the continental United States: 198996. J. Appl. Meteorol. 38, 10131019. Orville, R.E., Hufnes, G.R., 2001. Cloud-to-ground lightning in the United States: NLDN results in the rst decade, 198998. Mon. Weather Rev. 129, 11791193. Orville, R.E., Hufnes, G.R., Burrows, W.R., Holle, R.L., Cummins, K.L., 2002. The North American Lightning Detection Network (NALDN)rst results: 19982000. Mon. Weather Rev. 130, 20982109.

Pinto, O., 2008: An overview of cloud-to-ground lightning research in Brazil in the last two decades. Proceedings of the 20th International Lightning Detection Conference, Tuscon, Arizona, USA, 2123 April 2008. (available at http://www.vaisala.com/ les/An_overview_of_cloud-to-ground_lightning_research_in_Brazil_in_the_last_two_decades.pdf). Reap, R.M., Orville, R.E., 1990. The relationship between network lightning locations and surface hourly observations of thunderstorms. Mon. Weather Rev. 118, 94108. Rivas Soriano, L., De Pablo, F., 2002a. Effect of small urban areas in central Spain on the enhancement of cloud-to-ground lightning activity. Atmos. Environ. 36, 28092816. Rivas Soriano, L., De Pablo, F., 2002b. Study of lightning event duration and ash rate in the Iberian Peninsula using cloud-to-ground lightning data. Atmos. Res. 61, 189201. Rivas Soriano, L., De Pablo, F., Garcia Diez, E., 2002. Relationship between geographical latitude and longitude and cloud-to-ground lightning ash characteristics in the Iberian Peninsula. Atmosfera 15, 139146. Rivas Soriano, L., De Pablo, F., Tomas, C., 2005. Ten-year study of cloud-toground lightning activity in the Iberian Peninsula. J. Atmos. Sol. Terr. Phys. 67, 16321639. Schultz, M.D., Underwood, S.J., Radhakrishnan, P., 2005. A method to identify the optimal areal unit for NLDN cloud-to-ground lightning ash data analysis. J. Appl. Meteorol. 44, 739744. Schulz, W., Cummins, K., Diendorfer, G., Dorninger, M., 2005. Cloud-toground lightning in Austria: A 10-year study using data from a lightning location system. J. Geophys. Res. 110 (D09101)10.1029/2004JD005332 Sonnadara, U., Cooray, V., Gtschl, T., 2006. Characteristics of cloud-toground lightning ashes over Sweden. Phys. Scr. 74, 541548. Tuomi, T.J., 2005. Salamahavainnot 2005 Lightning Observations in Finland, 2005. Finnish Meteorological Institute. 40 pp. Tuomi, T.J., Larjavaara, M., 2005. Identication and analysis of ash cells in thunderstorms. Q. J. R. Meteorol. Soc. 131, 1191121410.1256/qj.04.64 Tuomi, T.J., Mkel, A., 2006. Salamahavainnot 2006 Lightning Observations in Finland, 2006. Finnish Meteorological Institute. 39 pp. Tuomi, T.J., Mkel, A., 2007. Salamahavainnot 2007 Lightning Observations in Finland, 2007. Finnish Meteorological Institute. 47 pp. Tuomi, T.J., Mkel, A., 2008a. Thunderstorm climate of Finland 19982007. Geophysica 44, 6780. Tuomi, T.J., Mkel, A., 2008b. Binomial model of lightning detection efciency. Journal of Lightning Research 1, 18. Tuomi, T.J., Mkel, A., 2008c. Salamahavainnot 2008 Lightning Observations in Finland, 2008. Finnish Meteorological Institute. 49 pp. Westcott, N.E., 1995. Summertime cloud-to-ground lightning activity around major Midwestern urban areas. J. Appl. Meteorol. 34, 16331642.