A 6-yr Climatology of Fronts Affecting Helsinki, Finland, and Their Boundary Layer Structure

Victoria A. Sinclair Division of Atmospheric Sciences, Department of Physics, University of Helsinki, Helsinki, Finland

Search for other papers by Victoria A. Sinclair in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

A 6-yr climatology of the frequency, characteristics, and boundary layer structure of synoptic-scale fronts in Helsinki, Finland, was created using significant weather charts and observations from a 327-m-tall mast and from the Station for Measuring Ecosystem–Atmosphere Relationships III. In total, 855 fronts (332 cold fronts, 236 warm fronts, and 287 occluded fronts) affected Helsinki during the 6-yr period, equating to one front every 2.6 days. Seasonal and diurnal cycles were observed, with frontal frequency peaking during the cold season and during daytime. Composites of warm and cold fronts were developed to provide observationally based conceptual models of the low-level structure of fronts at the end of the North Atlantic Ocean storm track. The composite warm front displays a temperature increase of 4.0°C; a broad, forward-tilting frontal zone; and prolonged, weak-to-moderate precipitation. The composite cold front is characterized by a temperature decrease of 4.4°C, a narrow and slightly rearward-tilting frontal zone, and moderate precipitation collocated with the surface front. Relationships between frontal characteristics and the direction from which fronts approached, the season, time of day, prefrontal boundary layer lapse rate, and the location of the wind shift relative to the thermal gradient were investigated. The prefrontal lapse rate was the single most important variable in determining the temperature change, the height of the maximum temperature change, and the near-surface tilt of both warm and cold fronts. This result demonstrates the interaction between boundary layer and synoptic-scale processes that must be captured by numerical weather prediction models to accurately forecast surface fronts.

Corresponding author address: Victoria Sinclair, Dept. of Physics, P.O. Box 48, FI-00014, University of Helsinki, Helsinki, Finland. E-mail: victoria.sinclair@helsinki.fi

Abstract

A 6-yr climatology of the frequency, characteristics, and boundary layer structure of synoptic-scale fronts in Helsinki, Finland, was created using significant weather charts and observations from a 327-m-tall mast and from the Station for Measuring Ecosystem–Atmosphere Relationships III. In total, 855 fronts (332 cold fronts, 236 warm fronts, and 287 occluded fronts) affected Helsinki during the 6-yr period, equating to one front every 2.6 days. Seasonal and diurnal cycles were observed, with frontal frequency peaking during the cold season and during daytime. Composites of warm and cold fronts were developed to provide observationally based conceptual models of the low-level structure of fronts at the end of the North Atlantic Ocean storm track. The composite warm front displays a temperature increase of 4.0°C; a broad, forward-tilting frontal zone; and prolonged, weak-to-moderate precipitation. The composite cold front is characterized by a temperature decrease of 4.4°C, a narrow and slightly rearward-tilting frontal zone, and moderate precipitation collocated with the surface front. Relationships between frontal characteristics and the direction from which fronts approached, the season, time of day, prefrontal boundary layer lapse rate, and the location of the wind shift relative to the thermal gradient were investigated. The prefrontal lapse rate was the single most important variable in determining the temperature change, the height of the maximum temperature change, and the near-surface tilt of both warm and cold fronts. This result demonstrates the interaction between boundary layer and synoptic-scale processes that must be captured by numerical weather prediction models to accurately forecast surface fronts.

Corresponding author address: Victoria Sinclair, Dept. of Physics, P.O. Box 48, FI-00014, University of Helsinki, Helsinki, Finland. E-mail: victoria.sinclair@helsinki.fi

1. Introduction

The passage of fronts is a critical, and often dominant, factor in determining the weather in the midlatitudes. A front is broadly defined as a zone of transition between two air masses with different origins and densities (e.g., Markowski and Richardson 2010, p. 115) and consequently the passage of a front can lead to changes in temperature, humidity, and wind direction. In addition, fronts can induce hazardous weather such as heavy rain, damaging winds, and disruptive snowfalls and, in the warm season, play a critical role in initiating convection; however, not all fronts cause notable weather. Frontal passages can also impact air quality either by ventilating pollutants out of the boundary layer (e.g., Purvis et al. 2003; Sinclair et al. 2008, 2010) or by transporting pollutants long distances (e.g., Stohl et al. 2002; Mari et al. 2004; Cooper et al. 2004). Hence, knowledge of how often fronts affect specific locations, what weather patterns they typically cause, and their typical structure is beneficial to forecasters, for validating numerical weather prediction and climate models, and for developing and revising conceptual models. Therefore, the purpose of this study is to create a climatology of frontal frequency in Helsinki, Finland, and to identify the typical structure of fronts within the boundary layer. Helsinki (60.1°N, 24.9°E) is an interesting location for studying boundary layer frontal structure as it is located at the end of the North Atlantic storm track, is at a high latitude where stable boundary layers are common and the length of day has a large annual variation, and is situated along the coast.

Few studies have examined the frequency of fronts, most likely due to the difficulties faced in compiling an objective climatology. Operational forecast centers traditionally analyzed surface charts manually and therefore limited digitalized archives of frontal position exist. Thus, climatologies need to be constructed manually and, hence, are time consuming to generate and are subjectively dependent on how the local forecasters originally analyzed the fronts. Consequently, the vast majority of frontal climatologies based on synoptic analysis charts focus on small geographic areas. For example, Chiang (1961) and Morgan et al. (1975) both compiled climatologies of fronts over the state of Illinois. Climatologies for single stations have also been constructed based on observations, which removes the sensitivity to the original analyses but limits the spatial applicability of the climatology. For example, Hoinka (1985) created a single-station, 10-yr climatology for Munich, Germany, and Fraedrich et al. (1986) compiled a 14-yr climatology for Berlin, Germany.

Recently, reanalysis datasets at moderate spatial and temporal resolutions have become widely available and frontal climatologies have been generated from these data. However, to objectively identify fronts from any gridded dataset, a mathematical algorithm must be specified, and hence a specific definition of what constitutes a front is required. Herein lies a problem, as was demonstrated by Uccellini et al. (1992); there is no consensus on what constitutes a front and hence no standard method to objectively identify a front's existence. [A similar problem, which must be addressed when constructing cyclone climatologies, is how to objectively identify (and track) extratropical cyclones; see, e.g., Raible et al. (2008) and Neu et al. (2013).] Renard and Clarke (1965) were the first to propose an algorithm to objectively locate fronts that was based on the thermal front parameter (TFP), which is a scalar quantity that represents “the gradient of the magnitude of the gradient of a thermodynamic variable.” This method has been extended by Hewson (1998) and Hewson and Titley (2010) and applied by numerous authors (e.g., Serreze et al. 2001; Jenkner et al. 2010; Berry et al. 2011), yet it remains unclear which thermodynamic variable (temperature, wet-bulb temperature, etc.) should be used in calculating the TFP, and how to determine the thresholds that eliminate thermally weak fronts. Additional objective frontal identification schemes have been proposed. Shafer and Steenburgh (2008) identified cold fronts using subjectively determined criteria in surface temperature decrease, surface pressure increase, and the 700-hPa temperature gradient. Simmonds et al. (2012) used an Eulerian scheme based on 6-hourly changes in the meridional wind speed and direction to locate fronts, as they found that 6-hourly temperature changes did not reliably identify frontal zones. In contrast, Sanders (1999) proposed that only horizontal thermal gradients should be considered when analyzing fronts and suggested that a horizontal potential temperature gradient of at least 8°C (110 km)−1 constitutes a strong frontal zone, and a gradient of half this value [8°C (220 km)−1] represents a moderate frontal zone.

Despite the challenges involved in constructing objective frontal climatologies from reanalyses, a small number do exist. For example, Payer et al. (2011) used National Centers for Environmental Prediction surface analyses and 2-m potential temperature data extracted from the North American Regional Reanalysis dataset to analyze frontal frequency and spatial distribution in the Great Lakes region. Also in the United States, Shafer and Steenburgh (2008) generated a 25-yr climatology of strong cold fronts in the Intermountain West and created a composite of the mesoscale structure and evolution of the 25 strongest fronts. In Europe, Jenkner et al. (2010) created a 3-yr climatology over the Alps of the frontal frequency and motion at 700 hPa based on high-resolution reanalysis data. However, the results of Shafer and Steenburgh (2008), Jenkner et al. (2010), and Payer et al. (2011) are only directly applicable to specific geographic areas. In contrast, Berry et al. (2011) utilized the 40-yr European Centre for Medium-Range Weather Forecasts Re-Analysis (ERA-40) dataset to generate what is believed to be the first global climatology of fronts. The results from Berry et al. indicate that frontal frequencies (the percent of ERA-40 analyses times that contain a front) peak at 10% in the midlatitude storm tracks, which corresponds to one front every 2.5 days. In southern Finland, the frontal frequency was ~5% (Fig. 2d in Berry et al. 2011), equivalent to one front every 5 days, which will be validated against observations in this study.

Although climatological studies of frontal frequency exist, very few include frontal characteristics; this serves as a motivation to analyze frontal characteristics, in addition to the frontal frequency, in this study. Furthermore, few observations of the vertical structure of fronts have been published, and, to the author's best knowledge, no climatologies exist that include the vertical structure of fronts within the atmospheric boundary layer. The majority of published studies examining the low-level structure of fronts have focused on case studies that are often selected explicitly because they were unusual events. Thus, it is plausible that very few average fronts have been analyzed and hence we potentially lack the knowledge of what an average front consists. Numerical modeling studies (e.g., Keyser and Anthes 1982; Muir and Reeder 2010) have shown that boundary layer processes have a significant effect on the structure of synoptic-scale fronts and consequently the vertical structure of fronts near the surface is worthy of detailed investigation.

Therefore, it is apparent that a climatology of frontal structures would be beneficial, particularly to local forecasters, especially as it is likely that the typical structure of fronts in Helsinki, at the end of the climatological North Atlantic storm track, differs considerably from fronts observed elsewhere in the world. In addition, combined with knowledge of local, seasonal patterns of fronts, the additional knowledge of typical frontal characteristics (e.g., temperature change and amount of precipitation) allows forecasters to identify in advance fronts that are likely to produce exceptional weather that may cause disruption to society and to therefore issue effective warnings in advance.

The main objectives of this study are to (i) create a climatology of frontal frequency in Helsinki, (ii) generate a composite warm and cold front from observations to determine the typical horizontal and vertical structures of fronts that affects Helsinki, and (iii) relate different frontal structures to synoptic-scale factors and boundary layer characteristics, which will allow the variation in frontal structures and characteristics to be better understood. The structure of this paper is as follows: a description of the data and the methodology is given in section 2. A traditional climatology of frontal frequency is presented in section 3, and the composite fronts are discussed in section 4. In section 5, attempts are made to relate the observed structure of fronts to the prefrontal boundary layer structure, season, and large-scale synoptic conditions, and our conclusions are presented in section 6.

2. Methodology and data

A climatology of fronts that passed over Helsinki (60.1°N, 24.9°E; Fig. 1) has been created from 6 yr of data, spanning 1 January 2006–31 December 2011. Significant weather charts (SWCs) that cover Fenno-Scandinavia and the Baltic countries and have approximate dimensions of 1500 km × 1800 km were examined to create the initial climatology. The SWCs are produced every 6 h by aviation forecasters at the Finnish Meteorological Institute (FMI). The positions of the fronts are analyzed by forecasters who locate fronts using a combination of temperature and wind observations, satellite images, and radar data. For a front to be included in the climatology, it was required to be analyzed on two consecutive SWCs and to pass directly over Helsinki. SWCs were selected as a basis for the climatology as the archive was 100% complete, the style of the charts did not change during the 6-yr period, and, in addition to the positions of the fronts, additional information such as the location of cloud, rain, and jet streams, as well as the direction that fronts were traveling in, was also available. The valid time of the SWC when each analyzed cold, warm, and occluded front was closest to Helsinki was identified and was included in the basic climatology dataset as the analysis time of each front. The analysis times were then used as a starting point when searching the observations for fronts. The approximate direction (i.e., south, southwest, etc.) that each front approached Helsinki from was also determined manually from the SWCs by considering consecutive charts and was added to the basic climatology dataset.

Fig. 1.
Fig. 1.

(a) The location of interest, which is shown by the black rectangle near 60°N 25°E, relative to the rest of northern Europe. The dashed line shows the Arctic Circle. (b) Zoomed-in map over the rectangle shown in (a). The H indicates the center of Helsinki city, S indicates the location of the SMEAR III station, and K indicates the location of the Kivenlahti mast. Note that the distance between K and S is 17 km.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

Observations from a 327-m-tall instrumented mast were then analyzed to determine the exact timing of all fronts and their characteristics. The mast is located in Kivenlahti (Fig. 1b), approximately 15 km west of Helsinki's city center and 8 km from the coast. Observations of temperature are available at seven measurement levels between 5 and 296 m, with wind speed and direction available at four levels (26, 93, 218, and 327 m). Both 10-min mean values and maximum and minimum values are available. As the mast is owned by a broadcasting company, access is highly restricted and expensive and thus the instruments are rarely calibrated. Therefore, the observations from the Kivenlahti mast are only suitable for analyzing the mesoscale structure of fronts but the height of the mast does provide useful, and uncommon, vertical profiles of frontal structure and boundary layer stratification.

To identify the exact time that each warm and cold front passed the Kivenlahti mast, time series of the temperature observations at all seven measurement levels were plotted for a 24-h period centered on the analysis time, as determined from the SWCs. In contrast to the method used to locate the fronts on the SWCs, only temperature data were used to subjectively locate the fronts in the observations. Therefore, occluded fronts were not located in the observations, as it proved difficult to identify with confidence the exact locations of occluded fronts using only near-surface temperature observations. A manual analysis of the 24-h-long temperature time series was then conducted to identify the exact time that each warm and cold front occurred, the duration of the frontal zone, and the temperature change across each front. The start of a cold (warm) frontal zone was taken to be the time at which the temperature began to decrease (increase) rapidly. The end of a cold (warm) frontal zone was defined to be where the temperature no longer decreased (increased) as rapidly as before. The frontal zones were independently identified for each of the seven measurement levels, and the start and end time of each frontal zone, at each level, was entered into the database. Additionally, the temperature at both the start and end time of each front was extracted for each level from the Kivenlahti dataset, and the temperature change ΔT across each front was calculated by ΔT = T(t = end time) − T(t = start time). Furthermore, to allow composites of the fronts to be calculated, a reference time t0 for each front was defined to be the time that the front was observed at 5 m at the Kivenlahti mast. Conventionally, fronts are defined to be located on the warm side of the thermal gradient, and therefore the reference time for cold (warm) fronts is the time at which the start (end) of the frontal zone was identified.

It should be noted that 10.5% of the cold fronts and 3% of the warm fronts that were analyzed on the SWCs could not be located in the temperature observations from the Kivenlahti mast. However, as almost all of the “missing” cold fronts exhibited a change in wind direction, this suggests that forecasters analyzed troughs or convergence lines as fronts. This emphasizes the lack of consensus in what constitutes a front and is related to the issue of incorrect frontal analysis raised previously by Sanders and Doswell (1995), Sanders (1999), and Sanders and Hoffman (2002), who argue that fronts should be identified by the presence of a moderate-to-strong thermal gradient and not by the presence of a wind shift or pressure trough.

To analyze micrometeorological aspects of frontal structure and to obtain observations of additional variables, observations from the Station for Measuring Ecosystem–Atmosphere Relationships (SMEAR) III have been included in the climatology. SMEAR III is located on the Kumpula campus of the University of Helsinki, approximately 17 km east of the Kivenlahti mast (Fig. 1b). SMEAR III (see Järvi et al. 2009 for more details) consists of a 31-m instrumented mast and a weather station on top of a nearby building, which is approximately 50 m high. Measurements from the mast include temperature, wind speed and direction, radiation, and turbulent fluxes of sensible heat and momentum, which are measured by the eddy covariance technique. In addition, the automated weather station provides measurements of temperature, dewpoint temperature, wind, radiation, pressure, and precipitation. Data were extracted for all fronts for a 24-h period centered on the reference times defined from the Kivenlahti mast data. This assumes that fronts are observed at SMEAR III and Kivenlahti at the same time, which after examining a comparison of observations could be shown to be a reasonably good approximation.

3. Climatology of frontal frequency

The frequency with which fronts passing over Helsinki were analyzed on SWCs is summarized in Table 1. In total, 855 fronts were analyzed during the 6-yr period, which equates to an average of 142.5 fronts per year, or 1 front per 2.6 days. Cold fronts were the most common type of front to affect Helsinki (1 front per 6.6 days), followed by occluded fronts (1 per 7.6 days), and then warm fronts, which were the least likely (1 front per 9.3 days) type of front to affect Helsinki. These results agree well with previous observationally based frontal climatology studies. Hoinka (1985) and Payer et al. (2011) also found cold fronts to be the most common type of front in Munich and the U.S. Great Lakes region, respectively, and warm fronts were found to be the rarest type of front in Berlin by Fraedrich et al. (1986). Warm fronts are less commonly observed than cold fronts in central and eastern Europe simply because warm fronts tend to be shorter in length than cold fronts. The shorter nature of these warm fronts is a result of the exit regions of jet streams being diffluent and thus having meridionally oriented axes of dilatation, which results in stubby warm fronts and frontolysis of zonally oriented warm fronts. In contrast to studies in central Europe and the United States, occluded fronts were observed in Helsinki notably more often, which can be attributed to the location of Helsinki, which is far from any major cyclogenesis region and also at the end of the North Atlantic climatological storm track.

Table 1.

Summary of the frequency with which warm, cold, and occluded fronts were observed in Helsinki during 1 Jan 2006–31 Dec 2011.

Table 1.

The quantitative frontal frequency found in Helsinki also agrees well with previous observational studies: Hoinka (1985) found that cold fronts occur once every 6.7 days and Jenkner et al. (2010) found that both warm and cold fronts occur once per 7 days in western Europe away from mountainous regions. In contrast, the frontal frequency in Helsinki calculated from SWCs is twice as large as the frontal frequency for Helsinki in the global climatology presented by Berry et al. (2011): one front per 2.6 days compared to one front per 5 days. However, the climatology from Berry et al. identified fronts at 850 hPa using reanalysis data with a 2.5° × 2.5° horizontal grid spacing and required fronts to exceed a specified value of the thermal front parameter. This suggests that half of the fronts observed in Helsinki are either too shallow or, more likely, have thermal gradients that are too weak to be included in the global climatology.

The number of fronts per year (Fig. 2a) varies considerably (>20%); however, the observed variations can be related to the average weather statistics of each year, which are calculated and published by FMI (Finnish Meteorological Institute 2012). The largest numbers of fronts were observed during 2007 and 2008, which were both notably warm years in Helsinki, and 2008 was an abnormally wet year. In addition, both 2007 and 2008 experienced mild winters, with anomalous westerly flow relative to 2010 and 2011, which experienced colder winters with blocking high pressure patterns.

Fig. 2.
Fig. 2.

Number of cold (black), warm (gray), and occluded (white) fronts: (a) total number of fronts per year, (b) mean number of fronts per month, (c) total number of fronts analyzed as a function of SWC analysis times, and (d) total number of fronts observed in 5-m observations from the Kivenlahti mast as a function of diurnal cycle.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

A seasonal cycle of frontal frequency exists for all frontal types (Fig. 2b) but is more pronounced for warm and occluded fronts than for cold fronts. The frequency of cold fronts exhibits a bimodal distribution, with peaks in spring and autumn and minima occurring during January–March and in summer. Warm and occluded fronts both peak in frequency during the cold season and have the lowest frequency in summer. The seasonal cycle for occluded fronts is particularly strong, with 50% fewer occluded fronts observed in summer than in winter. Using a one-sided Student's t test with a 5% confidence level, the mean number of warm fronts in summer (June–August) was determined to be significantly smaller (p value of 0.027) than the mean number of warm fronts in winter (December–February). The mean number of occluded fronts in summer was also found to be significantly smaller than in winter (p value of 0.0026), whereas no similar significant relationship was found for cold fronts. In total, more fronts occur during the cold season than in the warm season, which can be explained by the weaker meridional temperature gradient in summer and the related reduction in cyclones.

The diurnal cycle of frontal frequency was first assessed using the SWCs (Fig. 2c); however, as the charts were only available every 6 h, and no attempt to visually interpolate the timing of the fronts to Helsinki was made, this was inadequate to resolve any diurnal cycle. However, the SWCs suggest that fronts arriving in Helsinki do not display a strong diurnal cycle, with the exception of warm fronts, which appear more commonly at 0600 UTC than at any other analysis time. The weak/nonexistent diurnal cycle was expected as the SWCs depict synoptic-scale dynamics, which are not affected by the diurnal variation in solar radiation. However, Shafer and Steenburgh (2008) showed cold fronts to have diurnal variations in their strength, with the strongest surface cold fronts observed in the afternoon, and the passage of fronts in coastal regions having been shown to display a diurnal cycle (e.g., Bosart 1975; Garratt 1988). Therefore, the diurnal cycle of warm and cold fronts was also analyzed using the reference times defined for each front based on the 5-m temperature data from the Kivenlahti mast. The passage of both surface warm and cold fronts displayed a diurnal cycle, but the diurnal cycle was more pronounced for cold fronts than warm fronts (Fig. 2d). Warm front passages are most likely to be observed in Helsinki between 0600 and 1200 UTC (local time = UTC + 2 h) and cold fronts were most frequently observed between 1200 and 1500 UTC. A similar but weaker diurnal cycle is found when the temperature data from 296 m are used, and no diurnal cycle is found when only observations from December–February are considered (not shown). Therefore, it is hypothesized that the diurnal cycle in solar radiation that drives diurnal cycles in surface heat fluxes, boundary layer structure, and coastal, mesoscale circulation patterns can both modulate the time that synoptic-scale fronts pass over Helsinki and modify the temperature gradient of synoptic-scale fronts near the surface via diabatic frontogenesis or frontolysis. The morning peak in warm front passages in Helsinki likely occurs as warm fronts primarily approach Helsinki from the south and southwest and hence nocturnal land breezes directed off shore can prevent warm fronts from moving onshore until morning, as was shown in a case study by Kemppi and Sinclair (2011). Coastal circulations are not thought to have a large effect on cold fronts, as they rarely approach from the south or southwest. Instead, the afternoon peak in cold front passages is hypothesized to be a consequence of daytime diabatic frontogenesis and nocturnal frontolysis driven by differential cloud shading and hence differential surface heat fluxes (e.g., Koch et al. 1995; Gallus and Segal 1999). During the afternoon, surface cold fronts have large temperature gradients and are easy to identify, whereas at night frontolysis can weaken the surface temperature gradient to the extent that at the surface the cold front is no longer evident. Thus, although there is a pronounced diurnal cycle in the passage of fronts in Helsinki at the surface, similar diurnal cycles are not expected to occur above the boundary layer.

The direction that fronts approach Helsinki from is obtained from SWCs and the resulting distributions agree well with synoptic experience and climatologies of cyclone tracks (e.g., Murray and Simmonds 1995; Wernli and Schwierz 2006; Dacre and Gray 2009); the majority of fronts approach Helsinki from either the southwest or the west (Fig. 3). Warm and occluded fronts have very similar distributions of direction of approach but cold fronts differ considerably. Over 20% of cold fronts approach Helsinki from the northwest and 40% from the west, whereas the corresponding values for warm fronts are 8% and 26%. No cold fronts approached Helsinki from the east or southeast, whereas a small percentage of warm and occluded fronts arrived from this sector. The seasonal variation in the direction that fronts approach Helsinki (not shown) demonstrates that in winter and spring cold fronts are more likely to arrive from the north or northwest than in summer and autumn and that warm fronts only ever approach Helsinki from the east and southeast in spring and summer. This agrees well with synoptic experience, as in winter arctic cold fronts occasionally move southward over Finland while in summer warm and occluded fronts may arrive from the southeast and then quite often lead to severe convection (e.g., Punkka and Bister 2005; Punkka et al. 2006).

Fig. 3.
Fig. 3.

Direction from which fronts approach Helsinki as determined from the SWCs. Percentages are per total number of each type of front: (left) cold, (center) warm, and (right) occluded fronts.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

4. Composite warm and cold fronts

The mean 5-m temperature change due to cold fronts at Kivenlahti was −4.4°C and the change for warm fronts was 4.0°C. However, a large variation in temperature change due to both warm and cold fronts exists (Fig. 4a). The histogram of temperature change is strongly skewed toward lower values and indicates that fronts rarely lead to temperature changes of more than 10°C in Helsinki. The vast majority of fronts modify the 5-m air temperature by less than 6°C; 78% of cold fronts had a temperature decrease of less than 6°C and 82% of warm fronts had a temperature increase of less than 6°C. Furthermore, in the 6-yr study period, only 20 cold fronts led to a temperature decrease of more than 10°C and only 10 warm fronts caused a temperature increase of more than 10°C. Thus, it is apparent that the temperature difference across synoptic-scale fronts in southern Finland is notably smaller than in other regions of the world. For example, cold fronts with large temperature decreases are not infrequently observed in the Great Plains regions of the United States (e.g., Sanders 1955; Shapiro et al. 1985; Schultz 2004) and are relatively common in southeast Australia (e.g., Smith et al. 1982). However, fronts with temperature changes similar to those observed in Helsinki have been found to occur on the Pacific coast (e.g., Hobbs et al. 1980; Shafer and Steenburgh 2008) of the United States, which is also located at the end of a major storm track. However, two recent case studies of fronts in southern Finland (Kemppi and Sinclair 2011; Sinclair et al. 2012) have demonstrated that in Helsinki fronts with small temperature changes can have unconventional structures and can produce notable weather, such as heavy precipitation.

Fig. 4.
Fig. 4.

(a) Histogram of the 5-m temperature change measured at the Kivenlahti mast due to warm (gray) and cold (black) fronts and (b) the 5-m temperature change vs the 296-m temperature change for both warm (positive temperature change) and cold (negative temperature change) fronts. The shading scale indicates the lapse rate (°C km−1) of the boundary layer immediately ahead of the front. The percentages indicate the percentage of warm and cold fronts in each sector.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

The mean temperature change due to both warm and cold fronts was larger at 5 m than at 296 m and hence it can be concluded that the majority of fronts had their strongest thermal gradients near the surface (Fig. 4b). However, approximately one-third of both warm and cold fronts are stronger aloft rather than at the surface. The lapse rate, γ = −∂T/∂z, calculated between the highest and lowest levels of the Kivenlahti mast immediately ahead of warm and cold fronts, is a good indicator of whether the temperature change will be largest at the surface or aloft (Fig. 4b). The temperature change due to cold fronts is larger at the surface than aloft when the lapse rate is large and positive, representative of an unstable prefrontal boundary layer, whereas when the lapse rate is small and positive, or negative and representative of a stable prefrontal boundary layer, the temperature change due to cold fronts is greater aloft than at the surface. For warm fronts, a large lapse rate ahead of the front results in a larger temperature increase aloft than at the surface and small lapse rates lead to the temperature increase near the surface exceeding that aloft. When the prefrontal boundary layer is stably stratified, inversions may be present, which subsequently will be eroded by approaching fronts. Thus, when a cold front erodes a thermal inversion, warm air from aloft will be transported toward the surface, partially offsetting the cooling due to the front, resulting in a smaller temperature decrease near the surface than aloft. When a warm front erodes an inversion, the same process of turbulent mixing of warm air toward the surface occurs, however this combined with warm-air advection due to the front results in larger temperature increases near the surface rather than aloft. This process was previously noted by Doswell and Haugland (2007) for a front in Oklahoma, but the results found in the present study demonstrate that this process occurs for almost one-third of fronts in Helsinki and therefore is not a rare phenomenon.

Composite time series of warm and cold fronts have been constructed based on SMEAR III data and observations from the Kivenlahti mast. A relative time t′ is defined as t′ = tt0, where t is the actual time and t0 is the reference time for each front. Therefore, t′ equals zero at the time of the front, is negative before, and positive after, the front. The mean of all variables was then calculated at each relative time point for both warm and cold fronts. Ahead of the composite cold front the temperature increases by 2°C, potentially due to warm-air advection in the warm sector, before decreasing by more than 4°C after the front (Fig. 5a). The dewpoint temperature ahead of the composite cold front remains relatively constant and then decreased behind the cold front, as drier, postfrontal air masses advance. A pronounced pressure trough, with a 2–3-hPa pressure decrease, occurs for the composite cold front (Fig. 5b), which is consistent with the change in wind direction from 205° to 225° (Fig. 5c). The small wind shift in the composite front (20°) is very likely due to the different wind shift locations relative to the temperature gradient for each individual front. Only 36% of cold fronts had a wind shift collocated with the thermal gradient, and hence it is likely that the wind direction change is “smeared out” when the composites are created. If fronts were considered individually, the mean wind direction change would almost certainly be larger; this point will be revisited later. The mean wind speed did not change significantly across the composite cold front (Fig. 5c). Ahead of the composite cold front the wind speed increased by less than 1 m s−1 and decreases slightly after the surface front. This small signal in mean wind speed arises as the wind speed can either increase or decrease during a frontal passage. Therefore, the mean change in wind speed poorly represents typical cold fronts as a large degree of cancellation occurs. No clear correlations were identified between the change in the wind speed and other variables (e.g., lapse rate, temperature decrease) and, therefore, it is concluded that the synoptic-scale pressure gradient is the dominant factor in determining the change in wind speed across a cold front. The mean precipitation due to cold fronts in Helsinki agrees well with conceptual models (e.g., Browning 1985) as the heaviest rainfall occurs immediately after the surface cold front and is collocated with the thermal gradient (Fig. 5d). However, light rain rates are also observed ahead of the surface cold front and very light rain rates occur behind the surface front, indicating the variability in frontal structures.

Fig. 5.
Fig. 5.

The mean (a) temperature (solid lines) and dewpoint (dashed lines) at SMEAR III and temperature at Kivenlahti (dash–dotted lines), (b) surface pressure, (c) wind speed (solid lines) and wind direction (dashed lines), and (d) rain rate as a function of time relative to the time that the surface front was observed. The gray lines represent the composite warm front, and black lines represent the composite cold front. SMEAR III observations are from 51 m, and Kivenlahti data are at 48 m. All composites are constructed from 10-min observations, except precipitation, which is based on 30-min averages.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

The composite warm front calculated from SMEAR III data displays many features of traditional conceptual models of warm fronts and, in agreement with typical cyclone conceptual models, the thermal gradient due to the warm front is weaker than that of the composite cold front. The temperature and dewpoint temperature both increase due to the front and the smallest dewpoint depression (~2°C) occurs at the time of the surface front (Fig. 5a). No pressure trough is observed within the composite warm front; instead, a continuous decrease in pressure occurs (Fig. 5b). The change in wind direction is less pronounced than for cold fronts, but a weak veering occurs, and, in contrast to the composite cold front, the wind speed experiences a notable increase due to the passage of warm fronts (Fig. 5c). Rain due to the composite warm front is more widespread than that due to cold fronts, with rain occurring both ahead and behind the surface front (Fig. 5d), and, consequently, the total rainfall within a 24-h period of a warm front is larger (5.2 mm) than that for cold fronts (3.4 mm). These values are in general agreement with results of Catto et al. (2012), who related satellite-derived precipitation amounts to the occurrence of synoptic-scale fronts on global scales and found that on average cold fronts produce 2 mm of precipitation while warm fronts yield 3 mm. It should also be noted that no rain was observed to be associated with 74 (27%) cold fronts and 43 (20%) warm fronts. Dry cold fronts are relatively common, especially in continental regions (e.g., Garratt 1988; Smith et al. 1995; Friedrich et al. 2008), whereas dry warm fronts are not as commonly observed. However, as the precipitation data in this study come solely from one observation point (the automated weather station at SMEAR III), it is likely that fronts having scattered or light precipitation associated with them are classified incorrectly as dry fronts.

To provide a comparison between SMEAR III and Kivenlahti, the composite temperature time series for both cold and warm fronts based on Kivenlahti mast data at 48 m is also shown in Fig. 5a. For both warm and cold fronts, the Kivenlahti mast temperature observations are approximately 1°C cooler than those at SMEAR III. The difference may be caused by either poorly calibrated instruments on the Kivenlahti mast or, more likely, be due to the different local environments. SMEAR III is located in a suburban area with buildings nearby, whereas Kivenlahti is a semirural area and the mast is located in a forest. However, despite the differences in the absolute values, the time series of temperature due to the mean cold and warm fronts agree well. This demonstrates that the assumption that fronts are observed simultaneously at SMEAR III and Kivenlahti is a valid approximation.

While the composite analysis provides a good overall perspective on the frontal structures in Helsinki, it does not display the variability in the frontal structures. The variation in the temperature change due to fronts was discussed previously (Fig. 4a), and additional histograms of the change in wind direction and speed across all warm and cold fronts have been created based on SMEAR III data (Fig. 6). The change in both variables was calculated as the difference between the maximum and minimum values observed during a 12-h period centered on the reference time (t0) for each front, with a correction applied for fronts where the wind direction veered from the northwest to northeast (i.e., through 0°). A large amount of variability exists for both warm and cold fronts, and the median change in wind direction and wind speed is much larger than is suggested by the composite analysis: the median wind direction change for cold fronts was 77° and for warm fronts was 78°, whereas the median, absolute change in wind speed was 4.6 m s−1 for cold fronts and 4.3 m s−1 for warm fronts. This confirms that the change in the wind direction is seldom collocated with the thermal gradient and, hence, that the composite analysis smears out the cross-frontal gradient of the wind direction. Furthermore, as the change in wind speed is bimodal, with 48% of cold fronts and 62% of warm fronts leading to an increase in wind speed and the remaining fronts showing a decrease, it is evident that a large amount of cancellation occurs in the wind speed composites.

Fig. 6.
Fig. 6.

Histograms of the (a) change in wind direction and (b) change in wind speed for cold (black) and warm (gray) fronts. The change is calculated over a 12-h period. The median value of the wind direction change and the median of the absolute change in wind speed are displayed in each panel.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

The vertical structures of the composite warm and cold fronts are obtained from the Kivenlahti mast observations (Fig. 7). The mean temperature structure of cold fronts (Fig. 7a) shows that the prefrontal warming is more pronounced at low levels, which suggests that positive (upward) turbulent sensible heat fluxes, as well as synoptic-scale warm-air advection, may contribute to this warming. The rate of change of the temperature is strongest at the surface and weakens with height, while the fastest rate of temperature decrease at all levels occurs immediately after the surface front is observed (i.e., on the warm side of the thermal gradient). The temperature continues decreasing for 12 h after the surface front was observed, but the rate of temperature decrease is slower between 3 and 12 h after the surface front than during the 3 h immediately after the surface front was observed. The lapse rate during the 4 h before the composite cold front is conditionally unstable, and as the front approaches, the boundary layer becomes more unstable. The maximum lapse rate occurs immediately before the front and within the frontal zone the lapse rate rapidly decreases and a stable stratification develops. The wind direction veers with height both ahead and behind the composite cold front and the change in wind direction across the cold front is greatest nearest the surface (not shown). Veering winds with height in a synoptic context relate to warm-air advection, which likely does occur ahead of the mean cold front. However, below 300 m, the veering is very probably mostly due to surface friction, which causes the wind vectors to turn to the right of the geostrophic wind near the surface as the Coriolis force is reduced relative to the pressure gradient force (e.g., the behavior observed in the Ekman spiral).

Fig. 7.
Fig. 7.

Temperature time series for the composite (a) cold and (b) warm fronts based on Kivenlahti mast observations at the heights indicated.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

The composite warm front (Fig. 7b) demonstrates that the rate of change of the temperature is largest near the surface and again that the largest temperature gradient occurs on the warm side of the front. The composite warm front indicates that the temperature increases over a 6-h period, indicating that warm fronts are either wider or slower moving (or both) than cold fronts. The lapse rate ahead of the warm front is very small and is indicative of a stable layer. In the frontal zone (between t′ = −6 h and t′ = 0 h in Fig. 7b), there is a stable stratification; however, the stability decreases toward the surface front (t′ = 0 h). The maximum lapse rate is collocated with the surface composite warm front but is still representative of a stable layer (lapse rate = 6.2°C km−1). The wind direction (not shown) veers with height both ahead of and behind the composite warm front but the veering is greatest during the 2 h ahead of the surface front, which corresponds to the time period when the temperature is increasing the most rapidly.

It is more common to consider fronts in terms of potential temperature rather than temperature. Potential temperature θ cannot be calculated for the Kivenlahti mast data using the standard Poisson equation for potential temperature, as pressure observations are not available, but it can be calculated using θ(z) = T(z) + zΓd, where z is height, T is temperature, and Γd is the dry-adiabatic lapse rate. [This unusual formulation for potential temperature has also been applied to mast-based temperature measurements by Young and Johnson (1984).] For the composite cold front, the potential temperature height–time cross sections (Fig. 8a) show that the frontal surface tilts slightly forward with height immediately above the surface, is vertical between 50 and 150 m, and above 150 m tilts rearward relative to the direction of the frontal movement with height. The potential temperature distribution also shows that ahead of the composite cold front, superadiabatic lapse rates occur near the surface, which supports the presence of positive (upward) sensible heat fluxes, and that during the 2 h after the front is observed, the surface layer is well mixed. The potential temperature distribution of the composite warm front shows that the warm frontal surface is near vertical within the lowest 50 m and then tilts forward with height (Fig. 8b). Surface friction and enhanced convergence are the most likely causes of the vertical isentropes near the surface in both the composite cold and warm fronts.

Fig. 8.
Fig. 8.

Potential temperature (contour interval: 0.5°C) for the composite (a) cold and (b) warm fronts as a function of height and the time relative to the time that the surface front was observed. All data are from the Kivenlahti mast. Note that in this reference frame the fronts appear to be moving from left to right.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

5. Factors that influence frontal characteristics

Predicting the temperature change and amount of precipitation due to a front is a fundamental task for forecasters. However, many factors, acting across a wide range of scales, determine such properties of fronts, creating a complex problem for forecasters to address. This section attempts to identify which factors have the largest influence in determining these characteristics of fronts.

a. Direction of approach

The characteristics of fronts are likely to depend on the origins of the air masses that the fronts separate, which in turn suggests that the direction from which fronts approach Helsinki may influence the observed frontal characteristics. Therefore, the temperature change and the total amount of precipitation per front are considered as a function of the approach direction, which was identified from SWCs (Fig. 9). Cold fronts approaching from the north and northwest (from over land) cause a larger decrease in temperature than those arriving from either the west or southwest (at least partially from over the sea). Thus, it can be concluded that the dependence of cold front strength on the direction of approach is at least partly due to the different underlying surfaces and their characteristics. In addition, cold fronts that arrive from the north and northwest almost always have very little rain associated with them. This indicates the arctic nature of such cold fronts, the dry and cold air masses that are typically found behind such fronts, and the limited latent heat fluxes over a land surface compared to a sea surface.

Fig. 9.
Fig. 9.

(top) Temperature change (°C) and (bottom) total rain (mm) due to (a),(c) cold and (b),(d) warm fronts as a function of the direction from which the fronts approached Helsinki. Percentages are per total number of each type of front.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

The temperature increase due to warm fronts has a less pronounced directional variation, with the strongest warm fronts approaching from the northwest or west, but strong fronts are also observed to arrive from the south, southeast, and east. Compared to the temperature decrease, the total precipitation due to warm fronts has a much stronger dependence on the direction of the approach. Warm fronts with heavy precipitation all arrive from the south or southwest, which reflects both local sources of moisture located to the south (the Gulf of Finland and the Baltic Sea) and the generally more moist air masses of southwest Europe.

b. Seasonal and diurnal effects

A diurnal cycle in the temperature change, measured at Kivenlahti, due to both warm and cold fronts was observed (Fig. 10). Cold fronts that occurred during the afternoon (1300–1700 UTC) had the largest temperature decrease (mean value, −6.1°C; standard deviation, 3.6°C) associated with them whereas cold fronts occurring during early morning (0000–0400 UTC) had the smallest temperature decrease (mean value: −2.5°C; standard deviation: 1.4°C; see Fig. 10a). This supports the argument presented in section 3, where it was hypothesized that diabatic frontogenesis occurs during the day and diabatic frontolysis occurs at night due to differential cloud shading and surface heating. Warm fronts also exhibited a diurnal cycle (Fig. 10b), with the strongest fronts generally being observed during the day and the weakest fronts occurring at night. Interestingly, the temperature change due to both warm and cold fronts is much more variable when the fronts are observed during the day than at night. Weak seasonal variations exist in temperature change for cold fronts (Fig. 10c), with the strongest cold fronts occurring during summer and the weakest fronts occurring in autumn, but no pronounced seasonal variation exists for the intensity of warm fronts (Fig. 10d).

Fig. 10.
Fig. 10.

Box-and-whiskers plots of temperature change (°C) due to (a),(c) cold and (b),(d) warm fronts. Shown are (top) diurnal and (bottom) seasonal variation. The central line is the median, the edges of the box are the 25th and 75th percentiles, the whiskers extend to the most extreme data points that are not considers to be outliers, and the outliers are plotted individually. Data points are considered to be outliers if they are larger than Q3 + 1.5 × (Q3 − Q1) or smaller than Q1 − 1.5 × (Q3 − Q1), where Q1 and Q3 are the 25th and 75th percentiles, respectively. The overplotted solid black lines show the mean temperature change for each hour or month.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

The total amount of precipitation due to either warm or cold fronts did not display a strong diurnal cycle (Figs. 11a and 11b), but the variation in the amount of precipitation due to cold fronts did have a diurnal cycle. Many outliers [total rain greater than Q3 + 1.5 × (Q3 − Q1), where Q1 and Q3 are the 25th and 75th percentiles of the total rain distribution, respectively] occur in the afternoon, which reflects the potential for convection, and associated heavy rain, to be initiated by cold fronts. No diurnal cycle in the amount of variation in the total rain due to warm fronts exists. The seasonal cycle for total rain is similar for both warm and cold fronts (Figs. 11c and 11d), with fronts occurring in autumn causing the largest amount of rainfall and fronts in February and March producing the least precipitation. The seasonal cycle for warm front precipitation has a secondary peak in June and July, which probably is caused by warm fronts that approach Helsinki from the southeast and east and which can cause convection to develop.

Fig. 11.
Fig. 11.

As in Fig. 10, but for the total precipitation due to (a),(c) cold and (b),(d) warm fronts.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

c. Prefrontal boundary layer stability

The stability of the prefrontal boundary layer was shown to affect the height of the maximum temperature change (Fig. 4b) and it is also found that the prefrontal stability, quantified by the lapse rate, is strongly correlated with the absolute temperature change for both warm and cold fronts (Fig. 12). A negative correlation exists between the temperature change due to the front and the prefrontal lapse rate for both warm and cold fronts, but the correlation has a larger R-squared value for warm fronts than for cold fronts (−0.74 vs −0.39). Thus, the stability of the prefrontal boundary layer plays a significant role in determining the structure of synoptic-scale fronts within the boundary layer. This is confirmed when composite cold and warm fronts are constructed for fronts that move into either a stable or unstable prefrontal boundary layer (Fig. 13). Fronts are included in the stable prefrontal boundary layer composite front if the 3-h average prefrontal boundary layer lapse rate is less than 5°C km−1 and in the unstable prefrontal boundary layer composite if the 3-h average prefrontal boundary layer lapse rate is greater than 10°C km−1. Note that for cold fronts the 3-h averaging period is from relative time t′ = −3 to 0 h, whereas for warm fronts the averaging period was from t′ = −9 to −6 h, as it was assumed that the composite warm frontal zone extended from t′ = −6 to t′ = 0 h (e.g., Fig. 7b). Cold fronts that have stable prefrontal boundary layers are weaker and more rearward tilting with height than cold fronts, which have unstable prefrontal boundary layers (Figs. 13a and 13c). Furthermore, the post-cold-frontal boundary layer differs between the unstable and stable composites, indicating that cold fronts almost always change the stability of the boundary layer. Warm front structure also depends strongly on prefrontal stability with the strongest and most vertically orientated warm fronts occurring when the prefrontal boundary layer is stably stratified. Warm fronts are notably weaker when they advance into an unstable boundary than into a stable boundary layer (Figs. 13b and 13d).

Fig. 12.
Fig. 12.

The temperature change due to (a) cold and (b) warm fronts as a function of the 3-h average, prefrontal lapse rate derived between 5 and 296 m.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

Fig. 13.
Fig. 13.

Potential temperature (contour interval: 0.5°C) as a function of height and time relative to the time that the surface front was observed for (a) cold and (b) warm fronts that have a stable prefrontal boundary layer and for (c) cold and (d) warm fronts that have an unstable prefrontal boundary layer. Note that within this reference frame, the fronts appear to be moving from left to right.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

d. Relations among front intensity, precipitation, and wind shifts

Fronts experiencing frontogenesis have forced, ageostrophic circulations and ascent on the warm side of the thermal gradient. It could be assumed that fronts with strong thermal gradients have undergone strong frontogenesis, and thus strong thermal gradients are correlated with strong ascent and precipitation. However, the observed relationship between the temperature change and total precipitation (Figs. 14a and 14b) contradicts the above-stated hypothetical argument; strong warm and cold fronts lead to significantly less rain than weaker fronts while fronts with weak to moderate temperature changes (1° < |ΔT| < 5°C) produce the largest amounts of precipitation. Three explanations exist for this observed relationship. First, forced ascent will not generate clouds and precipitation unless moisture is available, and therefore, for fronts that approach from the north, strong ascent will not produce heavy rain. Second, many frontal systems that affect Helsinki are mature and are reaching the point of occlusion, when the temperature change across fronts is small. Therefore, weak fronts most probably produce more rain as they are approaching occlusion and the part of the front that affects Helsinki is close to the triple point of the cyclone, where heavy rain is known to occur. The third reason is that frontal precipitation can develop due to isentropic uplift and therefore the amount of precipitation will not be correlated to the strength of the forced circulation but instead to the slope of the front in the lower to midtroposphere.

Fig. 14.
Fig. 14.

The total amount of rain measured during a 12-h period centered on the surface front as a function of the temperature change due to (a) cold and (b) warm fronts.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

Finally, it is hypothesized that the temperature decrease due to cold fronts will depend on the location of the wind shift, with the strongest temperature decreases occurring for fronts that have collocated thermal gradients and wind shifts. Figure 15a shows that cold fronts with prefrontal, collocated, and postfrontal wind shifts all have similar median values of temperature decrease and similar distributions, indicating that cold fronts with collocated wind shifts on average do not lead to greater decreases in temperature. However, when the rate of the temperature decrease was considered, cold fronts with collocated wind shifts were found to have much more rapid decreases in temperature than fronts with either prefrontal or postfrontal wind shifts (Fig. 15b). This suggests that fronts with collocated wind shifts also exhibit strong convergence and frontogenesis, which contracts the thermal gradient into a narrow zone.

Fig. 15.
Fig. 15.

Box-and-whiskers plot of (a) temperature change and (b) rate of temperature change for cold fronts with wind direction changes that are located before (prefront), after (postfront), and exactly at the same time (collocated) as the surface front as identified from temperature time series. The central line is the median, the edges of the box are the 25th and 75th percentiles, the whiskers extend to the most extreme data points that are not considered to be outliers, and the outliers are plotted individually. Data points are considered to be outliers if they are larger than Q3 + 1.5 × (Q3 − Q1) or smaller than Q1 − 1.5 × (Q3 − Q1), where Q1 and Q3 are the 25th and 75th percentiles, respectively.

Citation: Journal of Applied Meteorology and Climatology 52, 9; 10.1175/JAMC-D-12-0318.1

6. Conclusions

Atmospheric fronts are dominant features in determining the weather and influencing the climate in the midlatitudes. Therefore, it is important that the frequency and characteristics of such features are documented. A 6-yr climatology of fronts in Helsinki was derived from significant weather charts, and front characteristics were determined by analyzing observations from a 327-m-tall mast and the SMEAR III station.

A front affected Helsinki every 2.6 days, confirming the importance of fronts in determining the weather in southern Finland. Cold fronts were the most common type of front, followed by occluded fronts, while warm fronts were the least common, which, except for the prevalence of occluded fronts, is in broad agreement with previous observational studies. The magnitude of the mean temperature change due to cold fronts was larger than for warm fronts, which, coupled with the fact that cold fronts are more common than warm fronts, suggests an unrealistic situation of ever-decreasing temperatures. Therefore, if it is assumed that diabatic processes are of secondary importance in modifying the temperature, it can be inferred that warm-air advection must occur more frequently than cold-air advection in Helsinki. This hypothesis is further supported by a wind rose of all wind direction observations over the 6-yr period (not shown) that shows that the predominant wind direction in Helsinki is southwesterly. The quantitative frontal frequency found in this study was approximately half of that reported by Berry et al. (2011) in a global study based on the ERA-40 reanalysis dataset. This suggests that 50% of the fronts observed in Helsinki were weaker than the threshold specified by Berry et al. (2011), and that typically fronts are weaker in Helsinki than in many other locations.

Seasonal and diurnal cycles exist for frontal frequency. The seasonal cycle was as expected, with the majority of fronts occurring in winter, the fewest fronts occurring in summer, and a secondary peak in cold fronts occurring in autumn. The diurnal cycle was less expected, with the maximum frequency of warm fronts occurring in the morning and the maximum frequency of cold fronts occurring in the afternoon. The diurnal cycle in warm front passages is hypothesized to be caused by land breezes preventing warm fronts from moving onshore, whereas nocturnal, diabatic frontolysis is thought to be the reason why few cold fronts are observed at night. However, as the reported diurnal cycles in frontal passages are for fronts observed at 5 m, and localized processes tied to the diurnal cycle in solar radiation are thought to be the cause, similar diurnal cycles are unlikely to occur above the boundary layer. However, to rigorously determine the cause of the diurnal cycle in frontal frequency, further investigation using additional observations and numerical model output would be required.

Composite front analysis demonstrated the typical characteristics of warm and cold fronts, most of which at least qualitatively agreed with preexisting conceptual models of fronts. One qualitative feature that is visible in the composite fronts yet is omitted in many conceptual models is the near-surface forward tilt of cold fronts and the vertical warm front surface immediately above the surface. The absence of such features in many conceptual models may be due to the inability of numerical weather prediction models to simulate such features, which are strongly coupled to the surface. Quantitative differences did exist between the composite fronts and schematic diagrams and conceptual models of fronts. Notable examples include the large number of cold fronts without collocated wind shifts and temperature changes, and the bimodal distribution of wind speed change across both warm and cold fronts. Furthermore, the temperature change of the composite warm and cold front is smaller than for many previous frontal studies. For example, the composite cold front has a smaller temperature decrease than all but 1 of the 11 cold fronts that Brundidge (1965) analyzed using mast observations, which further supports the discrepancy between the observed frontal frequency and the results of Berry et al. (2011). Ultimately, the composite warm and cold fronts developed in this study provide an ideal basis to compare frontal case studies with, and can be treated as location-specific, observationally based conceptual models of near-surface frontal structure.

Investigations aimed at identifying which factors determine the resultant structure of both warm and cold fronts showed that the direction that the fronts approach from affects both the temperature change and amount of precipitation, whereas only the temperature change has a diurnal cycle and only the amount of precipitation has a seasonal cycle. However, the single most important variable in determining the thermal structure of both warm and cold fronts in the boundary layer was the lapse rate, or stability, of the prefrontal boundary layer. Larger prefrontal boundary layer lapse rates lead to stronger, more vertically oriented cold fronts and weaker and more forward-tilting warm fronts. This correlation demonstrates that when stable layers and thermal inversions are eroded by fronts, the rate of temperature change is determined by a combination of horizontal thermal advection within the frontal zone and vertical turbulent mixing of heat. This highlights the interaction between boundary layer processes and synoptic-scale processes that numerical weather prediction models must be able to capture to correctly forecast the structure of characteristics fronts.

Acknowledgments

I thank Pasi Aalto, Ari Aaltonen, Leena Järvi, and Eveliina Tuovinen for providing data and the significant weather charts. The author was funded by Grant 126853 from the Academy of Finland.

REFERENCES

  • Berry, G., M. Reeder, and C. Jakob, 2011: A global climatology of atmospheric fronts. Geophys. Res. Lett.,38, L04809, doi:10.1029/2010GL046451.

  • Bosart, L., 1975: New England coastal frontogenesis. Quart. J. Roy. Meteor. Soc., 101, 957978.

  • Browning, K. A., 1985: Conceptual models of precipitating systems. Meteor. Mag., 114, 293319.

  • Brundidge, K. C., 1965: The wind and temperature structure of nocturnal cold fronts in the first 1,420 feet. Mon. Wea. Rev., 93, 587603.

    • Search Google Scholar
    • Export Citation
  • Catto, J. L., C. Jakob, G. Berry, and N. Nicholls, 2012: Relating global precipitation to atmospheric fronts. Geophys. Res. Lett.,39, L10805, doi:10.1029/2012GL051736.

  • Chiang, I., 1961: An analysis of selected synoptic elements of the climatology of Illinois for the period 1945–1959. M.S. thesis, Dept. of Geography, Southern Illinois University, 45 pp.

  • Cooper, O. R., and Coauthors, 2004: A case study of transpacific warm conveyor belt transport: Influences of merging airstream on trace gas import to North America. J. Geophys. Res., 109, D23S08, doi:10.1029/2003JD003624.

    • Search Google Scholar
    • Export Citation
  • Dacre, H. F., and S. L. Gray, 2009: The spatial distribution and evolution characteristics of North Atlantic cyclones. Mon. Wea. Rev., 137, 99115.

    • Search Google Scholar
    • Export Citation
  • Doswell, C. A., III, and M. J. Haugland, 2007: A comparison of two cold fronts—Effects of the planetary boundary layer on the mesoscale. Electron. J. Severe Storms Meteor.,2. [Available online at http://www.ejssm.org/ojs/index.php/ejssm/article/viewArticle/30/24.]

  • Finnish Meteorological Institute, cited 2012: Weather in recent years. [Available online at http://en.ilmatieteenlaitos.fi/weather-in-recent-years.]

  • Fraedrich, K., R. Bach, and G. Naujokat, 1986: Single station climatology of central European fronts: Number, time and precipitation statistics. Beitr. Phys. Atmos., 59, 5465.

    • Search Google Scholar
    • Export Citation
  • Friedrich, K., D. E. Kingsmill, C. Flamant, H. V. Murphey, and R. M. Wakimoto, 2008: Kinematic and moisture characteristics of a nonprecipitating cold front observed during IHOP. Part I: Across-front structures. Mon. Wea. Rev., 136, 147172.

    • Search Google Scholar
    • Export Citation
  • Gallus, W. A., Jr., and M. Segal, 1999: Diabatic effects on late-winter cold front evolution: Conceptual and numerical model evaluations. Mon. Wea. Rev., 127, 15181537.

    • Search Google Scholar
    • Export Citation
  • Garratt, J. R., 1988: Summertime cold fronts in southeast Australia—Behavior and low-level structure of main frontal types. Mon. Wea. Rev., 116, 636649.

    • Search Google Scholar
    • Export Citation
  • Hewson, T. D., 1998: Objective fronts. Meteor. Appl., 5, 3765.

  • Hewson, T. D., and H. A. Titley, 2010: Objective identification, typing and tracking of the complete life-cycles of cyclonic features at high spatial resolution. Meteor. Appl., 17, 355381.

    • Search Google Scholar
    • Export Citation
  • Hobbs, P. V., T. J. Matejka, P. H. Herzegh, J. D. Locatelli, and R. A. Houze Jr., 1980: The mesoscale and microscale structure and organization of clouds and precipitation in midlatitude cyclones. I: A case study of a cold front. J. Atmos. Sci., 37, 568596.

    • Search Google Scholar
    • Export Citation
  • Hoinka, K. P., 1985: On fronts in central Europe. Beitr. Phys. Atmos., 58, 560571.

  • Järvi, L., and Coauthors, 2009: The urban measurement station SMEAR III: Continuous monitoring of air pollution and surface–atmosphere interactions in Helsinki, Finland. Boreal Environ. Res., 14, 86109.

    • Search Google Scholar
    • Export Citation
  • Jenkner, J., M. Sprenger, I. Schwenk, C. Schwierz, S. Dierer, and D. Leuenberger, 2010: Detection and climatology of fronts in a high-resolution model reanalysis over the Alps. Meteor. Appl., 17, 118.

    • Search Google Scholar
    • Export Citation
  • Kemppi, M. L., and V. A. Sinclair, 2011: Structure of a warm front: Helsinki testbed observations and model simulation. Mon. Wea. Rev., 139, 28762900.

    • Search Google Scholar
    • Export Citation
  • Keyser, D., and R. A. Anthes, 1982: The influence of planetary boundary layer physics on frontal structure in the Hoskins–Bretherton horizontal shear model. J. Atmos. Sci., 39, 17831802.

    • Search Google Scholar
    • Export Citation
  • Koch, S. E., J. T. McQueen, and V. Mohan Karyampudi, 1995: A numerical study of the effects of differential cloud cover on cold frontal structure and dynamics. J. Atmos. Sci., 52, 937964.

    • Search Google Scholar
    • Export Citation
  • Mari, C., M. J. Evans, P. I. Palmer, D. J. Jacob, and G. W. Sachse, 2004: Export of Asian pollution during two cold front episodes of the TRACE-P experiment. J. Geophys. Res., 109, D15S17, doi:10.1029/2003JD004307.

    • Search Google Scholar
    • Export Citation
  • Markowski, P., and Y. Richardson, 2010: Mesoscale Meteorology in Midlatitudes. Wiley-Blackwell, 407 pp.

  • Morgan, G., D. Brunkow, and R. Beebe, 1975: Climatology of surface fronts. Illinois State Water Survey Circular ISWS-75-CIR122, 46 pp.

  • Muir, L. C., and M. J. Reeder, 2010: Idealized modelling of landfalling cold fronts. Quart. J. Roy. Meteor. Soc., 136, 21472161.

  • Murray, R. J., and I. Simmonds, 1995: Responses of climate and cyclones to reductions in Arctic winter sea ice. J. Geophys. Res., 100, 47914806.

    • Search Google Scholar
    • Export Citation
  • Neu, U., and Coauthors, 2013: IMILAST—A community effort to intercompare extratropical cyclone detection and tracking algorithms. Bull. Amer. Meteor. Soc., 94, 529–547.

    • Search Google Scholar
    • Export Citation
  • Payer, M., N. F. Laird, R. J. Maliawco, and E. G. Hoffman, 2011: Surface fronts, troughs, and baroclinic zones in the Great Lakes region. Wea. Forecasting, 26, 555563.

    • Search Google Scholar
    • Export Citation
  • Punkka, A. J., and M. Bister, 2005: Occurrence of summertime convective precipitation and mesoscale convective systems in Finland during 2000–01. Mon. Wea. Rev., 133, 362373.

    • Search Google Scholar
    • Export Citation
  • Punkka, A. J., J. Teittinen, and R. H. Johns, 2006: Synoptic and mesoscale analysis of a high-latitude derecho–severe thunderstorm outbreak in Finland on 5 July 2002. Wea. Forecasting, 21, 752763.

    • Search Google Scholar
    • Export Citation
  • Purvis, R. M., and Coauthors, 2003: Rapid uplift of nonmethane hydrocarbons in a cold front over central Europe. J. Geophys. Res., 108, 4224, doi:10.1029/2002JD002521.

    • Search Google Scholar
    • Export Citation
  • Raible, C. C., P. M. Della-Marta, C. Schwierz, H. Wernli, and R. Blender, 2008: Northern Hemisphere extratropical cyclones: A comparison of detection and tracking methods and different reanalyses. Mon. Wea. Rev., 136, 880897.

    • Search Google Scholar
    • Export Citation
  • Renard, R. J., and L. C. Clarke, 1965: Experiments in numerical objective frontal analysis 1. Mon. Wea. Rev., 93, 547556.

  • Sanders, F., 1955: An investigation of the structure and dynamics of an intense surface frontal zone. J. Meteor., 12, 542552.

  • Sanders, F., 1999: A proposed method of surface map analysis. Mon. Wea. Rev., 127, 945955.

  • Sanders, F., and C. A. Doswell III, 1995: A case for detailed surface analysis. Bull. Amer. Meteor. Soc., 76, 505521.

  • Sanders, F., and E. G. Hoffman, 2002: A climatology of surface baroclinic zones. Wea. Forecasting, 17, 774782.

  • Schultz, D. M., 2004: Cold fronts with and without prefrontal wind shifts in the central United States. Mon. Wea. Rev., 132, 20402053.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., A. H. Lynch, and M. P. Clark, 2001: The Arctic frontal zone as seen in the NCEP–NCAR reanalysis. J. Climate, 14, 15501567.

    • Search Google Scholar
    • Export Citation
  • Shafer, J. C., and W. J. Steenburgh, 2008: Climatology of strong intermountain cold fronts. Mon. Wea. Rev., 136, 784807.

  • Shapiro, M. A., T. Hampel, D. Rotzoll, and F. Mosher, 1985: The frontal hydraulic head: A micro-α scale (~1 km) triggering mechanism for mesoconvective weather systems. Mon. Wea. Rev., 113, 11661183.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., K. Keay, and J. A. T. Bye, 2012: Identification and climatology of Southern Hemisphere mobile fronts in a modern reanalysis. J. Climate, 25, 19451962.

    • Search Google Scholar
    • Export Citation
  • Sinclair, V. A., S. L. Gray, and S. E. Belcher, 2008: Boundary-layer ventilation by baroclinic life cycles. Quart. J. Roy. Meteor. Soc., 134, 14091424.

    • Search Google Scholar
    • Export Citation
  • Sinclair, V. A., S. L. Gray, and S. E. Belcher, 2010: Controls on boundary layer ventilation: Boundary layer processes and large scale dynamics. J. Geophys. Res.,115, D11107, doi:10.1029/2009JD012169.

  • Sinclair, V. A., S. Niemelä, and M. Leskinen, 2012: Structure of a narrow cold front in the boundary layer: Observations versus model simulation. Mon. Wea. Rev., 140, 24972519.

    • Search Google Scholar
    • Export Citation
  • Smith, R. K., B. F. Ryan, A. J. Troup, and K. J. Wilson, 1982: Cold fronts research: The Australian summertime cool change. Bull. Amer. Meteor. Soc., 63, 10281028.

    • Search Google Scholar
    • Export Citation
  • Smith, R. K., M. J. Redder, N. J. Tapper, and D. R. Christie, 1995: Central Australian cold fronts. Mon. Wea. Rev., 123, 1638.

  • Stohl, A., S. Eckhardt, C. Forster, P. James, and N. Spichtinger, 2002: On the pathways and timescales of intercontinental air pollution transport. J. Geophys. Res., 107, 4684, doi:10.1029/2001JD001396.

    • Search Google Scholar
    • Export Citation
  • Uccellini, L. W., S. F. Corfidi, N. W. Junker, P. J. Kocin, and D. A. Olson, 1992: Report on the surface analysis workshop at the National Meteorological Center 25–28 March 1991. Bull. Amer. Meteor. Soc., 73, 459471.

    • Search Google Scholar
    • Export Citation
  • Wernli, H., and C. Schwierz, 2006: Surface cyclones in the ERA-40 dataset (1958–2001). Part I: Novel identification method and global climatology. J. Atmos. Sci., 63, 24862507.

    • Search Google Scholar
    • Export Citation
  • Young, G. S., and R. H. Johnson, 1984: Meso- and microscale features of a Colorado cold front. J. Climate Appl. Meteor., 23, 13151325.

    • Search Google Scholar
    • Export Citation
Save
  • Berry, G., M. Reeder, and C. Jakob, 2011: A global climatology of atmospheric fronts. Geophys. Res. Lett.,38, L04809, doi:10.1029/2010GL046451.

  • Bosart, L., 1975: New England coastal frontogenesis. Quart. J. Roy. Meteor. Soc., 101, 957978.

  • Browning, K. A., 1985: Conceptual models of precipitating systems. Meteor. Mag., 114, 293319.

  • Brundidge, K. C., 1965: The wind and temperature structure of nocturnal cold fronts in the first 1,420 feet. Mon. Wea. Rev., 93, 587603.

    • Search Google Scholar
    • Export Citation
  • Catto, J. L., C. Jakob, G. Berry, and N. Nicholls, 2012: Relating global precipitation to atmospheric fronts. Geophys. Res. Lett.,39, L10805, doi:10.1029/2012GL051736.

  • Chiang, I., 1961: An analysis of selected synoptic elements of the climatology of Illinois for the period 1945–1959. M.S. thesis, Dept. of Geography, Southern Illinois University, 45 pp.

  • Cooper, O. R., and Coauthors, 2004: A case study of transpacific warm conveyor belt transport: Influences of merging airstream on trace gas import to North America. J. Geophys. Res., 109, D23S08, doi:10.1029/2003JD003624.

    • Search Google Scholar
    • Export Citation
  • Dacre, H. F., and S. L. Gray, 2009: The spatial distribution and evolution characteristics of North Atlantic cyclones. Mon. Wea. Rev., 137, 99115.

    • Search Google Scholar
    • Export Citation
  • Doswell, C. A., III, and M. J. Haugland, 2007: A comparison of two cold fronts—Effects of the planetary boundary layer on the mesoscale. Electron. J. Severe Storms Meteor.,2. [Available online at http://www.ejssm.org/ojs/index.php/ejssm/article/viewArticle/30/24.]

  • Finnish Meteorological Institute, cited 2012: Weather in recent years. [Available online at http://en.ilmatieteenlaitos.fi/weather-in-recent-years.]

  • Fraedrich, K., R. Bach, and G. Naujokat, 1986: Single station climatology of central European fronts: Number, time and precipitation statistics. Beitr. Phys. Atmos., 59, 5465.

    • Search Google Scholar
    • Export Citation
  • Friedrich, K., D. E. Kingsmill, C. Flamant, H. V. Murphey, and R. M. Wakimoto, 2008: Kinematic and moisture characteristics of a nonprecipitating cold front observed during IHOP. Part I: Across-front structures. Mon. Wea. Rev., 136, 147172.

    • Search Google Scholar
    • Export Citation
  • Gallus, W. A., Jr., and M. Segal, 1999: Diabatic effects on late-winter cold front evolution: Conceptual and numerical model evaluations. Mon. Wea. Rev., 127, 15181537.

    • Search Google Scholar
    • Export Citation
  • Garratt, J. R., 1988: Summertime cold fronts in southeast Australia—Behavior and low-level structure of main frontal types. Mon. Wea. Rev., 116, 636649.

    • Search Google Scholar
    • Export Citation
  • Hewson, T. D., 1998: Objective fronts. Meteor. Appl., 5, 3765.

  • Hewson, T. D., and H. A. Titley, 2010: Objective identification, typing and tracking of the complete life-cycles of cyclonic features at high spatial resolution. Meteor. Appl., 17, 355381.

    • Search Google Scholar
    • Export Citation
  • Hobbs, P. V., T. J. Matejka, P. H. Herzegh, J. D. Locatelli, and R. A. Houze Jr., 1980: The mesoscale and microscale structure and organization of clouds and precipitation in midlatitude cyclones. I: A case study of a cold front. J. Atmos. Sci., 37, 568596.

    • Search Google Scholar
    • Export Citation
  • Hoinka, K. P., 1985: On fronts in central Europe. Beitr. Phys. Atmos., 58, 560571.

  • Järvi, L., and Coauthors, 2009: The urban measurement station SMEAR III: Continuous monitoring of air pollution and surface–atmosphere interactions in Helsinki, Finland. Boreal Environ. Res., 14, 86109.

    • Search Google Scholar
    • Export Citation
  • Jenkner, J., M. Sprenger, I. Schwenk, C. Schwierz, S. Dierer, and D. Leuenberger, 2010: Detection and climatology of fronts in a high-resolution model reanalysis over the Alps. Meteor. Appl., 17, 118.

    • Search Google Scholar
    • Export Citation
  • Kemppi, M. L., and V. A. Sinclair, 2011: Structure of a warm front: Helsinki testbed observations and model simulation. Mon. Wea. Rev., 139, 28762900.

    • Search Google Scholar
    • Export Citation
  • Keyser, D., and R. A. Anthes, 1982: The influence of planetary boundary layer physics on frontal structure in the Hoskins–Bretherton horizontal shear model. J. Atmos. Sci., 39, 17831802.

    • Search Google Scholar
    • Export Citation
  • Koch, S. E., J. T. McQueen, and V. Mohan Karyampudi, 1995: A numerical study of the effects of differential cloud cover on cold frontal structure and dynamics. J. Atmos. Sci., 52, 937964.

    • Search Google Scholar
    • Export Citation
  • Mari, C., M. J. Evans, P. I. Palmer, D. J. Jacob, and G. W. Sachse, 2004: Export of Asian pollution during two cold front episodes of the TRACE-P experiment. J. Geophys. Res., 109, D15S17, doi:10.1029/2003JD004307.

    • Search Google Scholar
    • Export Citation
  • Markowski, P., and Y. Richardson, 2010: Mesoscale Meteorology in Midlatitudes. Wiley-Blackwell, 407 pp.

  • Morgan, G., D. Brunkow, and R. Beebe, 1975: Climatology of surface fronts. Illinois State Water Survey Circular ISWS-75-CIR122, 46 pp.

  • Muir, L. C., and M. J. Reeder, 2010: Idealized modelling of landfalling cold fronts. Quart. J. Roy. Meteor. Soc., 136, 21472161.

  • Murray, R. J., and I. Simmonds, 1995: Responses of climate and cyclones to reductions in Arctic winter sea ice. J. Geophys. Res., 100, 47914806.

    • Search Google Scholar
    • Export Citation
  • Neu, U., and Coauthors, 2013: IMILAST—A community effort to intercompare extratropical cyclone detection and tracking algorithms. Bull. Amer. Meteor. Soc., 94, 529–547.

    • Search Google Scholar
    • Export Citation
  • Payer, M., N. F. Laird, R. J. Maliawco, and E. G. Hoffman, 2011: Surface fronts, troughs, and baroclinic zones in the Great Lakes region. Wea. Forecasting, 26, 555563.

    • Search Google Scholar
    • Export Citation
  • Punkka, A. J., and M. Bister, 2005: Occurrence of summertime convective precipitation and mesoscale convective systems in Finland during 2000–01. Mon. Wea. Rev., 133, 362373.

    • Search Google Scholar
    • Export Citation
  • Punkka, A. J., J. Teittinen, and R. H. Johns, 2006: Synoptic and mesoscale analysis of a high-latitude derecho–severe thunderstorm outbreak in Finland on 5 July 2002. Wea. Forecasting, 21, 752763.

    • Search Google Scholar
    • Export Citation
  • Purvis, R. M., and Coauthors, 2003: Rapid uplift of nonmethane hydrocarbons in a cold front over central Europe. J. Geophys. Res., 108, 4224, doi:10.1029/2002JD002521.

    • Search Google Scholar
    • Export Citation
  • Raible, C. C., P. M. Della-Marta, C. Schwierz, H. Wernli, and R. Blender, 2008: Northern Hemisphere extratropical cyclones: A comparison of detection and tracking methods and different reanalyses. Mon. Wea. Rev., 136, 880897.

    • Search Google Scholar
    • Export Citation
  • Renard, R. J., and L. C. Clarke, 1965: Experiments in numerical objective frontal analysis 1. Mon. Wea. Rev., 93, 547556.

  • Sanders, F., 1955: An investigation of the structure and dynamics of an intense surface frontal zone. J. Meteor., 12, 542552.

  • Sanders, F., 1999: A proposed method of surface map analysis. Mon. Wea. Rev., 127, 945955.

  • Sanders, F., and C. A. Doswell III, 1995: A case for detailed surface analysis. Bull. Amer. Meteor. Soc., 76, 505521.

  • Sanders, F., and E. G. Hoffman, 2002: A climatology of surface baroclinic zones. Wea. Forecasting, 17, 774782.

  • Schultz, D. M., 2004: Cold fronts with and without prefrontal wind shifts in the central United States. Mon. Wea. Rev., 132, 20402053.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., A. H. Lynch, and M. P. Clark, 2001: The Arctic frontal zone as seen in the NCEP–NCAR reanalysis. J. Climate, 14, 15501567.

    • Search Google Scholar
    • Export Citation
  • Shafer, J. C., and W. J. Steenburgh, 2008: Climatology of strong intermountain cold fronts. Mon. Wea. Rev., 136, 784807.

  • Shapiro, M. A., T. Hampel, D. Rotzoll, and F. Mosher, 1985: The frontal hydraulic head: A micro-α scale (~1 km) triggering mechanism for mesoconvective weather systems. Mon. Wea. Rev., 113, 11661183.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., K. Keay, and J. A. T. Bye, 2012: Identification and climatology of Southern Hemisphere mobile fronts in a modern reanalysis. J. Climate, 25, 19451962.

    • Search Google Scholar
    • Export Citation
  • Sinclair, V. A., S. L. Gray, and S. E. Belcher, 2008: Boundary-layer ventilation by baroclinic life cycles. Quart. J. Roy. Meteor. Soc., 134, 14091424.

    • Search Google Scholar
    • Export Citation
  • Sinclair, V. A., S. L. Gray, and S. E. Belcher, 2010: Controls on boundary layer ventilation: Boundary layer processes and large scale dynamics. J. Geophys. Res.,115, D11107, doi:10.1029/2009JD012169.

  • Sinclair, V. A., S. Niemelä, and M. Leskinen, 2012: Structure of a narrow cold front in the boundary layer: Observations versus model simulation. Mon. Wea. Rev., 140, 24972519.

    • Search Google Scholar
    • Export Citation
  • Smith, R. K., B. F. Ryan, A. J. Troup, and K. J. Wilson, 1982: Cold fronts research: The Australian summertime cool change. Bull. Amer. Meteor. Soc., 63, 10281028.

    • Search Google Scholar
    • Export Citation
  • Smith, R. K., M. J. Redder, N. J. Tapper, and D. R. Christie, 1995: Central Australian cold fronts. Mon. Wea. Rev., 123, 1638.

  • Stohl, A., S. Eckhardt, C. Forster, P. James, and N. Spichtinger, 2002: On the pathways and timescales of intercontinental air pollution transport. J. Geophys. Res., 107, 4684, doi:10.1029/2001JD001396.

    • Search Google Scholar
    • Export Citation
  • Uccellini, L. W., S. F. Corfidi, N. W. Junker, P. J. Kocin, and D. A. Olson, 1992: Report on the surface analysis workshop at the National Meteorological Center 25–28 March 1991. Bull. Amer. Meteor. Soc., 73, 459471.

    • Search Google Scholar
    • Export Citation
  • Wernli, H., and C. Schwierz, 2006: Surface cyclones in the ERA-40 dataset (1958–2001). Part I: Novel identification method and global climatology. J. Atmos. Sci., 63, 24862507.

    • Search Google Scholar
    • Export Citation
  • Young, G. S., and R. H. Johnson, 1984: Meso- and microscale features of a Colorado cold front. J. Climate Appl. Meteor., 23, 13151325.

    • Search Google Scholar
    • Export Citation
  • Fig. 1.

    (a) The location of interest, which is shown by the black rectangle near 60°N 25°E, relative to the rest of northern Europe. The dashed line shows the Arctic Circle. (b) Zoomed-in map over the rectangle shown in (a). The H indicates the center of Helsinki city, S indicates the location of the SMEAR III station, and K indicates the location of the Kivenlahti mast. Note that the distance between K and S is 17 km.

  • Fig. 2.

    Number of cold (black), warm (gray), and occluded (white) fronts: (a) total number of fronts per year, (b) mean number of fronts per month, (c) total number of fronts analyzed as a function of SWC analysis times, and (d) total number of fronts observed in 5-m observations from the Kivenlahti mast as a function of diurnal cycle.

  • Fig. 3.

    Direction from which fronts approach Helsinki as determined from the SWCs. Percentages are per total number of each type of front: (left) cold, (center) warm, and (right) occluded fronts.

  • Fig. 4.

    (a) Histogram of the 5-m temperature change measured at the Kivenlahti mast due to warm (gray) and cold (black) fronts and (b) the 5-m temperature change vs the 296-m temperature change for both warm (positive temperature change) and cold (negative temperature change) fronts. The shading scale indicates the lapse rate (°C km−1) of the boundary layer immediately ahead of the front. The percentages indicate the percentage of warm and cold fronts in each sector.

  • Fig. 5.

    The mean (a) temperature (solid lines) and dewpoint (dashed lines) at SMEAR III and temperature at Kivenlahti (dash–dotted lines), (b) surface pressure, (c) wind speed (solid lines) and wind direction (dashed lines), and (d) rain rate as a function of time relative to the time that the surface front was observed. The gray lines represent the composite warm front, and black lines represent the composite cold front. SMEAR III observations are from 51 m, and Kivenlahti data are at 48 m. All composites are constructed from 10-min observations, except precipitation, which is based on 30-min averages.

  • Fig. 6.

    Histograms of the (a) change in wind direction and (b) change in wind speed for cold (black) and warm (gray) fronts. The change is calculated over a 12-h period. The median value of the wind direction change and the median of the absolute change in wind speed are displayed in each panel.

  • Fig. 7.

    Temperature time series for the composite (a) cold and (b) warm fronts based on Kivenlahti mast observations at the heights indicated.

  • Fig. 8.

    Potential temperature (contour interval: 0.5°C) for the composite (a) cold and (b) warm fronts as a function of height and the time relative to the time that the surface front was observed. All data are from the Kivenlahti mast. Note that in this reference frame the fronts appear to be moving from left to right.

  • Fig. 9.

    (top) Temperature change (°C) and (bottom) total rain (mm) due to (a),(c) cold and (b),(d) warm fronts as a function of the direction from which the fronts approached Helsinki. Percentages are per total number of each type of front.

  • Fig. 10.

    Box-and-whiskers plots of temperature change (°C) due to (a),(c) cold and (b),(d) warm fronts. Shown are (top) diurnal and (bottom) seasonal variation. The central line is the median, the edges of the box are the 25th and 75th percentiles, the whiskers extend to the most extreme data points that are not considers to be outliers, and the outliers are plotted individually. Data points are considered to be outliers if they are larger than Q3 + 1.5 × (Q3 − Q1) or smaller than Q1 − 1.5 × (Q3 − Q1), where Q1 and Q3 are the 25th and 75th percentiles, respectively. The overplotted solid black lines show the mean temperature change for each hour or month.

  • Fig. 11.

    As in Fig. 10, but for the total precipitation due to (a),(c) cold and (b),(d) warm fronts.

  • Fig. 12.

    The temperature change due to (a) cold and (b) warm fronts as a function of the 3-h average, prefrontal lapse rate derived between 5 and 296 m.

  • Fig. 13.

    Potential temperature (contour interval: 0.5°C) as a function of height and time relative to the time that the surface front was observed for (a) cold and (b) warm fronts that have a stable prefrontal boundary layer and for (c) cold and (d) warm fronts that have an unstable prefrontal boundary layer. Note that within this reference frame, the fronts appear to be moving from left to right.

  • Fig. 14.

    The total amount of rain measured during a 12-h period centered on the surface front as a function of the temperature change due to (a) cold and (b) warm fronts.

  • Fig. 15.

    Box-and-whiskers plot of (a) temperature change and (b) rate of temperature change for cold fronts with wind direction changes that are located before (prefront), after (postfront), and exactly at the same time (collocated) as the surface front as identified from temperature time series. The central line is the median, the edges of the box are the 25th and 75th percentiles, the whiskers extend to the most extreme data points that are not considered to be outliers, and the outliers are plotted individually. Data points are considered to be outliers if they are larger than Q3 + 1.5 × (Q3 − Q1) or smaller than Q1 − 1.5 × (Q3 − Q1), where Q1 and Q3 are the 25th and 75th percentiles, respectively.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 633 225 25
PDF Downloads 335 104 3