Abstract

A 9-yr (2000–08) analysis of precipitation characteristics for the central and western European Alps has been generated from ground-based operational weather radar data provided by the Swiss radar network. The radar-based precipitation analysis focuses on the relationship between synoptic-scale weather patterns and mesoscale precipitation distribution over complex alpine terrain. The analysis divides the Alps into six regions (each approximately 200 × 200 km2 in size)—one on the northern side, two each on the western and southern sides of the Alps, and one in the Massif Central—representing various orographic aspects and localized climates within the radar coverage area. For each region, estimated precipitation rate derived from radar data is analyzed on a seasonal basis for total daily precipitation and frequency of high-precipitation-rate events. The summer season has the highest total daily precipitation for all regions in the study, whereas median values of daily precipitation in winter are less than one-half of median daily precipitation for summer. For all regions, high-precipitation-rate events occur most frequently in the summer. Daily synoptic-scale weather patterns are associated with total daily precipitation and frequency of high precipitation rate to show that an advective synoptic-scale pattern with southerly midtropospheric flow results in the highest median and 90th-quantile values for total daily precipitation and that a convective synoptic-scale pattern results in elevated frequency of extreme-precipitation-rate events.

1. Introduction

Precipitation is a primary element of the water cycle, and precipitation characteristics such as frequency, intensity, type, and duration are likely to be affected by changes in global and regional climate. The Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC; Kundzewicz et al. 2007) emphasizes that changes in precipitation related to climate change will exacerbate the current stress on water resources from population growth and land-use change. Mountains play an especially important role in the hydrologic cycle, with implications for runoff, flooding, water storage, and glacier-mass balance (Beniston 2005). Changes in precipitation amount and intensity in mountainous regions are of major concern, affecting both the localized alpine area and the lowlands that utilize freshwater supplied from upper-elevation catchments and are exposed to flooding risk that results from extreme precipitation. Understanding past and future precipitation characteristics in mountainous regions is essential to develop water policy and management strategies for responding to the impact of climate change. The 2008 IPCC technical paper on climate change and water notes that there is insufficient information on the effect of climate change in regions where topography generates fine spatial scales in climate, and it also notes that further development is needed of catchment-scale climate models that are more relevant to water management (Bates et al. 2008, 133–137).

The intent of this study is to utilize operational radar data to analyze precipitation distribution in the European Alps. Here, the distribution of total daily precipitation and occurrence of high-precipitation-rate events over the years 2000–08 are summarized for regions located to the north, northwest, west, and south of the main crest of the Swiss Alps. The goals for development of this analysis are 1) to demonstrate the use of radar-based precipitation estimates in creating a decade-length precipitation record, 2) to compare regional precipitation distributions within the Alps, and 3), beyond direct analysis of precipitation characteristics, to understand the relationship between precipitation distribution and synoptic-scale weather patterns. Systematic classification of synoptic-scale weather patterns into discrete categories over the time span of the analysis enables correlations between synoptic patterns and resulting precipitation characteristics.

The analysis presented herein is unique because of its operational radar basis. Previous European Alps precipitation climatological studies have reported precipitation quantity and intensity as inferred from daily rain gauge values (e.g., Schönwiese et al. 1994; Widmann and Schär 1997; Frei and Schär 1998; Schmidli et al. 2002; Klein Tank and Können 2003; Begert et al. 2005; Schmidli and Frei 2005; Moberg et al. 2006; Zolina et al. 2008, 2010). Although rain gauge–based precipitation “climatologies” provide valuable information about seasonal and regional precipitation attributes, they are limited by the spatial resolution of the rain gauge network. Smith et al. (2003) showed that precipitation distribution in complex Alpine terrain varies at scales of <10 km, and yet rain gauge–based precipitation climatologies for the Alps only attain spatial resolution of approximately 25 km (Frei and Schär 1998). The radar-based precipitation data presented here have underlying spatial resolution of 2 × 2 km2.

Furthermore, the radar basis of this study provides higher temporal resolution than most nonautomated rain gauges. Rain gauge–based climatologies typically characterize data on a daily basis (Frei and Schär 1998; Klein Tank and Können 2003; Schmidli and Frei 2005; Moberg et al. 2006; Zolina et al. 2008, 2010), whereas radar provides precipitation estimates at 5–15-min intervals, thereby enabling detailed precipitation rate analysis. Wüest et al. (2010) recognized the time-scale advantage of radar precipitation estimates and applied hourly radar analyses to rain gauge data to improve the temporal resolution of a rain gauge–based climatology for the Alps, but they did not incorporate spatial advantages inherent to radar. Subhour radar-based precipitation rate was leveraged by Carbone et al. (2002) and Ahijevych et al. (2004) to investigate diurnal precipitation patterns in the continental United States. Germann et al. (2009) even demonstrate the usefulness of radar-derived precipitation-rate estimation as an input to a real-time hydrologic model for a mountainous region. The distribution of high precipitation rate is important to consider and is included in this study because of the implications for flooding (Allamano et al. 2009).

This study is also distinguished by the complex terrain of the European Alps. A 10-year radar-based precipitation climatology of total precipitation and extreme precipitation events was developed previously for the relatively flat terrain of the Netherlands (Overeem et al. 2009a,b). In earlier work, Schiesser et al. (1995) demonstrated the utility of radar for characterizing the structure of severe mesoscale precipitation systems over Switzerland for a 5-year period (1985–89), but radar instrument capability has significantly improved since this work was published. Panziera and Germann (2010) demonstrated the use of Swiss radar data from a single site to characterize airflow and precipitation rate for 58 orographic precipitation events occurring in 2004–08. Here, the time span is extended to the entirety of 2000–08 and the dataset is a composite of three Swiss radar sites (Fig. 1). Usage of ground-based radar in complex terrain presents many challenges. Obstacles faced by modern radar installations and associated verification of radar precipitation estimates in mountainous areas, specifically the central European Alps, are described in Germann et al. (2006, 2009).

Fig. 1.

Area of investigation covered by the three MeteoSwiss radars (locations indicated with squares): Albis, La Dole, and Monte Lema. The boundaries of the six investigation regions are shown with thick black lines. Thin lines indicate 800 m MSL topographic contours.

Fig. 1.

Area of investigation covered by the three MeteoSwiss radars (locations indicated with squares): Albis, La Dole, and Monte Lema. The boundaries of the six investigation regions are shown with thick black lines. Thin lines indicate 800 m MSL topographic contours.

2. Data and method

a. Data

Data from the operational Swiss Weather Service (MeteoSwiss) radar network form the basis of this analysis. The radar network includes three C-band Doppler weather radars at Lema, Albis, and La Dole (located near Locarno, Zurich, and Geneva, respectively; see Fig. 1). Each of the radars scans 20 elevations every 5 min, monitoring radar reflectivity up to a range of 230 km. MeteoSwiss composites the three Swiss radars to estimate surface precipitation in a data product called “RAIN.” RAIN provides a 30-min running average of surface precipitation rate on a 2 × 2 km2 Cartesian grid and is updated every 5 min. Thorough descriptions of the Swiss radar network and data products are found in Joss et al. (1998), Germann and Joss (2004), Germann et al. (2006, 2009), and Panziera and Germann (2010). MeteoSwiss radar-based precipitation estimates incorporate multiple quality-control measures, including automatic hardware calibration; removal of ground clutter and false echoes; corrections for visibility to account for and properly weight pulse volumes that are not fully visible by the radar antenna; corrections for vertical profile of reflectivity to account for precipitation phase change and growth between the height of the visible measurement and the ground; and bias correction for nonuniform beam filling, low-level growth unaccounted for by the vertical profile correction, and attenuation (Joss et al. 1998; Germann and Joss 2004; Germann et al. 2006).

Precipitation fields derived from the Swiss radars are adjusted to rain gauge measurements on an annual basis. Analyses of measurement uncertainty revealed that the adjustment was most valuable when applied over a large space–time window, that is, a single value for each year (Germann et al. 2006). Germann et al. (2006) have previously evaluated precipitation estimates from the Swiss radar network by comparison of total daily rain gauge and radar precipitation measurements and showed improvement over 1997–2004 that resulted in a global bias across the whole of Switzerland of −1.3% for the daily radar:gauge ratio for summer 2004. Germann et al. (2006) also quantifies localized bias between radar-estimated and rain gauge–measured precipitation. In terms of geography, bias varies from 0.2% on the south side of the Alps near the Monte Lema radar to 13% in the interior of the Alps where the terrain is more complex.

b. Seasons and regions

The time intervals defined for analysis of seasonal precipitation distribution and patterns are spring (March–May), summer (June–August), autumn (September–November), and winter (December–February; the year associated with the winter season is the year of January/February).

The Swiss Alps are divided into six regions to facilitate spatial analysis of the interaction between synoptic weather conditions and topography. The first requirement for defining the regions is that each is contained within the geographic boundaries of radar coverage. Second, the regions are defined such that they are roughly aligned with the main Alpine crest. With these requirements in mind, precipitation distributions are calculated for each season of 2000–08 for the six regions shown in Fig. 1: northern Alps (N), southern Alps (S), northwestern Alps (NW), southwestern Alps (SW), western Alps (W), and Massif Central (MC). The regions range in size from 15 000 to 40 000 km2 (3700 to 10 000 pixels). Note that the NW, N, and S regions each contain one of the three radar sites, and therefore data for these regions result from shorter distances and lower heights of the radar beam and have different data quality than the more distant MC region. Also noted is absence of a region covering the eastern Alps because of lack of radar visibility.

c. Total precipitation

Total daily precipitation is estimated at each pixel within each region by summing the estimated precipitation rate (mm h−1) at the surface for each 30-min period (Precip_Rate30min) over the course of each day [Eq. (1)]. The daily totals are then averaged over each season for all pixels in a given region to give average daily precipitation:

 
formula

Equation (1) reveals a difference between the time series of average daily precipitation calculated here and time series of daily precipitation reported by rain gauge–based climatologies. The radar-based daily precipitation in this analysis is a temporal and spatial average of precipitation received at each pixel (representing 2 × 2 km2) within a region, whereas rain gauge–based precipitation is generally reported as time series for specific points that can be interpolated to provide regional precipitation averages.

d. High precipitation rate

The 30-min time resolution of the radar-based precipitation rate data is leveraged to develop an analysis that includes precipitation rate as well as quantity. Therefore, in addition to total daily precipitation, the regional and seasonal distributions of high-precipitation-rate events are also investigated. Precipitation rate is analyzed in terms of geographic extent and duration of heavy precipitation (i.e., number of pixels having a precipitation rate of ≥20 mm h−1 over the course of a season) and is subsequently referred to as frequency of high precipitation rate F [Eq. (2)]:

 
formula

The 20 mm h−1 threshold is confirmed as a high, rarely occurring value (extreme value) for precipitation rate by observing that over the period of 2000–08 a precipitation rate of 20 mm h−1 or greater occurred in 0.39% of the sampled pixels that indicated occurrence of precipitation.

Although quality-control (QC) measures are built into the dataset as described in section 2a, an additional measure was applied to the frequency of high-precipitation-rate data to further remove nonprecipitating echoes. The QC procedure involved gathering frequency data during a time frame in which no precipitation was received and radar reflectivity is related to clear-air echoes. Pixels that showed a rain rate of >8 mm h−1 during this time frame were omitted from the frequency-of-high-precipitation-rate analysis.

e. Synoptic weather patterns

Distribution of total precipitation and high precipitation rate is generally coupled with the associated synoptic weather situation (e.g., Lin et al. 2001; Rudari et al. 2004). The identification of weather patterns most likely to result in the greatest daily precipitation and the most widespread occurrence of high precipitation rate allows prediction of future precipitation amounts related to changes in natural modes of the atmospheric circulations such as the North Atlantic Oscillation. Schüepp’s weather classification is a system that provides daily categorization of the synoptic pattern in the central Alps from 1945 to the present. A detailed description of Schüepp’s weather classification is provided by Schüepp (1979), Wanner et al. (1998), and Stefanicki et al. (1998). Schüepp’s classification system is a subjective method because the weather pattern definitions are known in advance. It is used for this analysis because of its established usage in the Swiss Alpine area. Daily Schüepp synoptic weather pattern classifications for 2000–08 were obtained from MeteoSwiss.

The main three classes of the Schüepp system (convective, advective, and mixed) are determined by the strength of the surface pressure gradient across a 444-km diameter centered in Switzerland at 46.5°N, 9°E. The convective type is characterized by lack of surface pressure gradient, and the advective classification is associated with the existence of a gradient in surface pressure of >5 hPa across the diameter. The mixed classification results when surface and 500-hPa pressure gradients are at odds (strong gradient at one level while weak at the other). The main classes are further divided into subclasses: the convective subclasses are convective high (CH), convective flat (CF), and convective low (CL) (on the basis of 500-hPa geopotential height relative to the annual mean); the advective subclasses are advective north (AN), advective south (AS), advective east (AE), and advective west (AW) (on the basis of the direction of the 500-hPa geostrophic wind); and the mixed subclass (MIX), which indicates no further subclassification. A final distinction into 40 weather situations is made by evaluating the angle between surface isobars and 500-hPa contours and the 500-hPa wind speed and geopotential height from soundings in Payerne, Switzerland; Milan, Italy; and Munich, Germany.

Maps of 1000- and 500-hPa geopotential height are generated for Schüepp’s classifications that are herein identified as associated with days having high total precipitation or frequency of high precipitation rate. The geopotential data were retrieved from the European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA) Interim dataset with a horizontal resolution of 1.5° × 1.5°.

3. Results and analysis

a. Total precipitation

Total precipitation for each region is compared within each season to indicate which region has the highest precipitation amount for 2000–08 (Fig. 2). The SW has the highest median value in winter, spring, and autumn and is among the group of regions having the greatest average daily precipitation for all seasons (based on pairwise comparison using a t test with 95% confidence). The N region has highest median in summer, however. The W and MC regions are consistently in the group with the lowest median values of total precipitation in all seasons. Also, the maximum values of total precipitation in each region qualitatively follow the same pattern as the distribution of median values for each region.

Fig. 2.

Average daily precipitation between 2000 and 2008 for each season is plotted as a function of region. Diamonds indicate maximum values, boxes indicate median values, and the lines indicate interquartile range at the hash marks (within each region). The regions receiving the most total precipitation during each season are shaded gray (based on pairwise comparison with 95% confidence as discussed in section 3a).

Fig. 2.

Average daily precipitation between 2000 and 2008 for each season is plotted as a function of region. Diamonds indicate maximum values, boxes indicate median values, and the lines indicate interquartile range at the hash marks (within each region). The regions receiving the most total precipitation during each season are shaded gray (based on pairwise comparison with 95% confidence as discussed in section 3a).

To indicate which season provides the highest values of total precipitation, each season is compared within each region (Fig. 3). Summer has the highest median and maximum value of total precipitation for all regions and is the wettest season in all regions, with median values of total precipitation for summer being more than 2 times the winter median values. The SW is the exception for which a significant difference does not exist between any pair of seasons (at 95% confidence level; Fig. 3). The result agrees with Frei and Schär (1998), who also reported summer to be the wettest season across the Alps, with the exception of the southwestern Alps, where minimal seasonal differences were found.

Fig. 3.

Average daily precipitation between 2000 and 2008 for each region is plotted as a function of season. Diamonds indicate maximum values, boxes indicate median values, and the lines indicate interquartile range at the hash marks (within each season). The seasons receiving the most total precipitation within each region are shaded gray (based on pairwise comparison with 95% confidence).

Fig. 3.

Average daily precipitation between 2000 and 2008 for each region is plotted as a function of season. Diamonds indicate maximum values, boxes indicate median values, and the lines indicate interquartile range at the hash marks (within each season). The seasons receiving the most total precipitation within each region are shaded gray (based on pairwise comparison with 95% confidence).

The 10-day running mean (from −5 days to +4 days) of total daily precipitation averaged over 2000–08 as shown in Fig. 4 provides some insight into the relative seasonal values of total precipitation. Each region follows a similar annual cycle. Daily precipitation reaches its minimum during winter, steadily increases during spring, reaches maximum values during the summer, begins to decrease in late summer, and remains at lower values with more variability during autumn as compared with summer. Late spring and early–midautumn both contain individual days that have total precipitation in several regions that is similar to that of the summer season; when averaged together, however, with drier days earlier in spring and later in autumn, both seasonal means of total precipitation are less than in summer, which culminates in the relative mean values of the seasonal distributions (Fig. 3). The high variability of total precipitation in spring and autumn may be reflective of the increased occurrence of strong southerly flow that results in heavy precipitation and flooding during April/May and October/November as reported by Grazzini (2007). Most of the heavy-precipitation events, especially in spring and autumn, are linked to upper-level troughs (Lin et al. 2001; Rudari et al. 2004). The lee cyclone, also called the Genoa cyclone, forms in the lee of the southern Alps close to the city of Genoa, Italy. This weather situation first causes heavy precipitation on the southern side of the Alps followed by heavy precipitation on the northern side (e.g., areas of Switzerland received 200–300 mm of precipitation over a 5-day period during the August 2005 flood; Hohenegger et al. 2008).

Fig. 4.

The 10-day running-average daily precipitation received over 2000–08 is plotted for each day of the year and for each region. The 10-day running average over all regions is indicated by the black line.

Fig. 4.

The 10-day running-average daily precipitation received over 2000–08 is plotted for each day of the year and for each region. The 10-day running average over all regions is indicated by the black line.

b. Frequency of high precipitation rate

To provide insight into local areas of maximum precipitation rate and the spatial distribution within each region, the number of occurrences of precipitation rate exceeding 20 mm h−1 is summed at each pixel over each season (resulting in resolution of 2 × 2 km2). Figure 5 is shown as an example of the output generated for each season of 2007. In 2007 summer was the season with the largest area that experienced a precipitation rate of >20 mm h−1. The area south of the Alps contained more numerous and widespread high-precipitation events than did the other regions and seasons. Spring was the season with the second largest area of high precipitation rate over the Alps. The Lago Maggiore area in the southern region of the Alps had the greatest frequency of high-precipitation events in spring, summer, and autumn of 2007.

Fig. 5.

Horizontal distribution of high-precipitation-rate frequency (number of occurrences of precipitation rate of >20 mm h−1) is plotted for all seasons of 2007. Black lines indicate 800 m MSL topography.

Fig. 5.

Horizontal distribution of high-precipitation-rate frequency (number of occurrences of precipitation rate of >20 mm h−1) is plotted for all seasons of 2007. Black lines indicate 800 m MSL topography.

Frequency of high precipitation rate—that is, total seasonal count of pixels that exceed a precipitation rate of 20 mm h−1 divided by the surface area of each region—is summarized for each region and season over 2000–08 in Fig. 6. Larger values indicate greater geographic extent or the combined effect of widespread and frequent high-precipitation events. The overall maximum frequency of high precipitation rate observed (summer in the NW region, Figs. 6 and 7) results both from increased horizontal distribution of high-precipitation-rate events, meaning that more surface area received at least one instance of high precipitation rate, and increased occurrence of high precipitation rates in localized areas, specifically north and northwest of the Alps. Summer months have the highest frequency in all regions over the period of 2000–08, and winter has the lowest frequency (Figs. 6 and 7). The S region has the highest overall mean frequency value (all seasons; not shown), and the W region has the lowest overall mean frequency.

Fig. 6.

Frequency of high precipitation rate between 2000 and 2008 for each season is plotted as a function of region. Diamonds indicate maximum values, boxes indicate median values, and the lines indicate interquartile range at the hash marks (within each region). The regions with the greatest frequency of high-precipitation-rate events during each season are shaded gray (based on pairwise comparison with 95% confidence).

Fig. 6.

Frequency of high precipitation rate between 2000 and 2008 for each season is plotted as a function of region. Diamonds indicate maximum values, boxes indicate median values, and the lines indicate interquartile range at the hash marks (within each region). The regions with the greatest frequency of high-precipitation-rate events during each season are shaded gray (based on pairwise comparison with 95% confidence).

Fig. 7.

Frequency of high precipitation rate between 2000 and 2008 for each region is plotted as a function of season. Diamonds indicate maximum values, boxes indicate median values, and the lines indicate interquartile range at the hash marks (within each region). The seasons with the greatest frequency of high-precipitation-rate events in each region are shaded gray (based on pairwise comparison with 95% confidence). Note the scale differences on the y axis for the various regions.

Fig. 7.

Frequency of high precipitation rate between 2000 and 2008 for each region is plotted as a function of season. Diamonds indicate maximum values, boxes indicate median values, and the lines indicate interquartile range at the hash marks (within each region). The seasons with the greatest frequency of high-precipitation-rate events in each region are shaded gray (based on pairwise comparison with 95% confidence). Note the scale differences on the y axis for the various regions.

In contrast to total precipitation shown in Fig. 2, seasonal comparison of frequency of high precipitation rate between each region with 95% confidence shows the most high-precipitation-rate events in winter occur in the N region (Fig. 6). In spring and summer, the W region has the least frequency of high-precipitation-rate events, and in autumn the S and SW regions have the greatest frequency of high-precipitation-rate events.

Comparison of the frequency of high precipitation rate between seasons (with threshold set at 20 mm h−1 for all seasons) within each region shows that frequency in summer is significantly different than that for each of the other seasons in all regions (Fig. 7). Spring, autumn, and winter are not found to be significantly different. Therefore, in general, summer has the most widespread occurrence of high-precipitation-rate events, and the result is consistent with previous studies that show that the majority of severe mesoscale convective systems and heavy-precipitation events in Switzerland occur during summer (Schiesser et al. 1995; Laing and Fritsch 1997).

c. Correlation of synoptic weather pattern with total precipitation and high precipitation rate

As already discussed in sections 2e and 3a, most of the precipitation in the Alps is related to the synoptic-scale weather pattern. If the 9-yr (2000–08) precipitation distribution presented here is related to synoptic weather patterns, it may enable prediction of future precipitation amounts and occurrence of high precipitation rate based on changes in occurrence of synoptic weather patterns as predicted by climate models. First, the synoptic weather patterns are identified that correspond to the highest median and 90th quantile (Q90) of total daily precipitation and frequency of high precipitation rate for each region over 2000–08 (Table 1). Days characterized by an advective pattern have both the highest median and Q90 of total daily precipitation. This implies that advective patterns are associated with the highest values of total daily precipitation. Furthermore, the AS pattern accounts for the highest median and Q90 values of total daily precipitation in all regions, except the N. Advective patterns in the Schüepp classification exhibit a surface-level pressure gradient and are further classified by the direction of the 500-hPa geostrophic wind. The AS pattern is generally associated with a midtropospheric southerly flow that carries moist air from the Mediterranean Sea toward the Alps and provides the necessary moisture to result in large daily totals of precipitation as shown in Fig. 8. The subcategories of the advective Schüepp classification are limited to four cardinal directions (north, south, east, west), and the AS pattern includes advection from the southwest, as well as the south, whereas advection from the southeast is classified as AE (Schüepp 1979; Wanner et al. 1998; Stefanicki et al. 1998).

Table 1.

Schüepp’s weather classifications corresponding to largest median and largest 90%-quantile (Q90) values for total daily precipitation and frequency of high precipitation rate are shown for each region for 2000–08. The synoptic weather pattern abbreviations are given in section 2e.

Schüepp’s weather classifications corresponding to largest median and largest 90%-quantile (Q90) values for total daily precipitation and frequency of high precipitation rate are shown for each region for 2000–08. The synoptic weather pattern abbreviations are given in section 2e.
Schüepp’s weather classifications corresponding to largest median and largest 90%-quantile (Q90) values for total daily precipitation and frequency of high precipitation rate are shown for each region for 2000–08. The synoptic weather pattern abbreviations are given in section 2e.
Fig. 8.

The 1000-hPa geopotential height (solid contours; dam) and 500-hPa geopotential height (dashed contours; dam) averaged for the top five total precipitation days during 2000–08 (26 Oct 2004, 31 Oct 2003, 13 Jul 2008, 8 Dec 2006, and 6 Dec 2006) classified as the AS Schüepp weather pattern.

Fig. 8.

The 1000-hPa geopotential height (solid contours; dam) and 500-hPa geopotential height (dashed contours; dam) averaged for the top five total precipitation days during 2000–08 (26 Oct 2004, 31 Oct 2003, 13 Jul 2008, 8 Dec 2006, and 6 Dec 2006) classified as the AS Schüepp weather pattern.

Figure 8 shows the 1000- and 500-hPa geopotential height contours averaged for the five days classified as AS that produced the greatest amount of total daily precipitation as indicated by operational radar across all regions during 2000–08. The surface pressure gradient across Switzerland leads to the advective classification while the south-southwest 500-hPa geostrophic wind direction further classifies the situation as AS. The predominant feature is the passage of an upper-level trough west of the Alps that usually extends from England to the Mediterranean and causes moist flow from over the Mediterranean to interact with the Alpine topography to result in heavy precipitation. In addition, the westward movement of the trough is slowed by a ridge to the east of the Alps. The synoptic situation displayed in Fig. 8 is similar to the elongated trough, or potential vorticity streamer, described as an ingredient for heavy Alpine precipitation in Doswell et al. (1998), Massacand et al. (1998), Lin et al. (2001), Martius et al. (2006), Grazzini (2007), and references within. Although the North Atlantic Ocean has been identified as the greatest contributor of moisture to annual precipitation for the overall Alpine area (Sodemann and Zubler 2010), the association of the AS pattern with the highest daily values of precipitation suggests that the Mediterranean is the moisture source that produces the largest daily totals of precipitation across the Alpine area, with the exception of the north region.

Days characterized by a convective pattern account for the both the greatest median and Q90 of high-precipitation-rate frequency in all regions (Table 1). Therefore, across all regions, convective patterns are associated with the greatest geographic extent of high precipitation rate. In fact, the CH pattern accounts for the greatest median and Q90 values of high-precipitation-rate frequency in all regions, except for the S region. Convective patterns are characterized by a dome or ridge of high pressure at the surface that creates a weak pressure gradient and subsequently results in light low-level geostrophic winds. Weak surface-level winds may allow localized convective systems producing high precipitation rates to develop, thereby enabling the relationship between a convective synoptic pattern and high precipitation rate.

Figure 9 shows the 1000- and 500-hPa geopotential height contours averaged for the five days classified as CH that produced the greatest frequency of high precipitation rate as indicated by operational radar across all regions for 2000–08. The convective classification is associated with minimal surface pressure gradient across Switzerland (Fig. 9). The CH subclassification is determined when the 500-hPa geopotential height is in the upper quartile of the mean annual distribution of 500-hPa heights for soundings from Payerne, Milan, and Munich. Because the majority of high-precipitation-rate events are found in the warm summer season, as evidenced here and also in Schiesser et al. (1995), it is not surprising that the associated 500-hPa geopotential heights are in the upper quartile. Also notable in Fig. 9 is the southwesterly geostrophic wind at 500 hPa. Again, similar to the advective situation, the southerly component of the midtropospheric flow carries moisture from the Mediterranean toward the Alps. Vertical wind shear, as indicated by the difference in geostrophic wind between the surface and midtroposphere, also creates an environment favorable for development of intense storms capable of producing high precipitation rates (Weisman and Klemp 1984).

Fig. 9.

The 1000-hPa geopotential height (solid contours; dam) and 500-hPa geopotential height (dashed contours; dam) averaged for the top five frequency-of-high-precipitation-rate days during 2000–08 (24 Jun 2005, 12 Jul 2006, 20 Jun 2007, 12 Jun 2003, and 27 Jun 2006) classified as the CH Schüepp weather pattern.

Fig. 9.

The 1000-hPa geopotential height (solid contours; dam) and 500-hPa geopotential height (dashed contours; dam) averaged for the top five frequency-of-high-precipitation-rate days during 2000–08 (24 Jun 2005, 12 Jul 2006, 20 Jun 2007, 12 Jun 2003, and 27 Jun 2006) classified as the CH Schüepp weather pattern.

The relationship between total precipitation and frequency of high precipitation rate is analyzed with conditional probability. To be specific, the conditional probability that a given day is above the 90th quantile for total precipitation given that the frequency of high precipitation rate is above the 90th quantile, noted as P(total90|rate90) is defined as

 
formula

The conditional probability, P(total90|rate90), is a measure of the causal effect of high precipitation rate on large amounts of total daily precipitation. Larger values indicate that high precipitation rate (in this case, ≥20 mm h−1) contributes more to total precipitation. Likewise, smaller conditional probabilities result when large quantities of total precipitation result from longer-duration, lower-precipitation-rate events.

Figure 10 shows conditional probability, P(total90|rate90), for each Schüepp synoptic pattern and region. The CH pattern has the highest conditional probabilities for all regions. Therefore, under a CH synoptic pattern, high precipitation rates are more likely to result in large quantities of total precipitation. In contrast, the AS synoptic pattern has lower conditional probability for each region, an indication that large-quantity-of-precipitation events occurring in an AS pattern are associated with persistent precipitation at lower rates.

Fig. 10.

Conditional probability that daily total rainfall (mm day−1) is above the 90th quantile given that the frequency of high precipitation rate (No. per kilometer squared) for the day is above the 90th quantile, shown for each Schüepp synoptic classification and region.

Fig. 10.

Conditional probability that daily total rainfall (mm day−1) is above the 90th quantile given that the frequency of high precipitation rate (No. per kilometer squared) for the day is above the 90th quantile, shown for each Schüepp synoptic classification and region.

Figure 11 further illustrates the use of conditional probability in evaluating the relationship between precipitation rate and quantity. Frequency of high precipitation rate is plotted against total daily precipitation for the south region for days classified as AS and CH, with the Q90 values for frequency of high precipitation rate and total daily precipitation indicated with dashed lines. Data points in the upper-right sector demarked by the 90th-quantile lines are values that are above the 90th quantile for both precipitation rate and quantity. For the AS pattern, the values greater than Q90 for total precipitation are dispersed across both sides of the Q90 line for precipitation rate. The CH pattern exhibits a stronger relationship between total precipitation and precipitation rate, thereby explaining the larger conditional probability for the CH pattern as shown in Fig. 10.

Fig. 11.

Frequency of high precipitation rate for each day (No. per day per kilometer squared) is plotted vs total daily precipitation (mm day−1) for the south region for (left) days classified as AS and (right) days classified as CH. The data include all days over 2000–08. The vertical dashed line is the 90th quantile of total daily precipitation for the south region, and the horizontal dashed line is the 90th quantile for frequency of high precipitation rate.

Fig. 11.

Frequency of high precipitation rate for each day (No. per day per kilometer squared) is plotted vs total daily precipitation (mm day−1) for the south region for (left) days classified as AS and (right) days classified as CH. The data include all days over 2000–08. The vertical dashed line is the 90th quantile of total daily precipitation for the south region, and the horizontal dashed line is the 90th quantile for frequency of high precipitation rate.

Table 2 shows the relative frequencies of the main Schüepp weather classifications for the period of 1945–94 [as reported in Wanner et al. (1998)] and also for the period of 2000–08. The data indicate an increased frequency of convective days at the expense of a decrease in advective days. The increase in convective days and decrease in advective days for 2000–08 as compared with 1945–94 is consistent with trends reported in Stefanicki et al. (1998). On the basis of the association of convective patterns with high precipitation rate and the relationship between precipitation rate and total precipitation for convective patterns, it may be expected that an increasing frequency of convective patterns will lead to intense precipitation events accounting for a greater portion of total precipitation. In fact, several studies provide evidence that extreme events are increasing in probability and increasing their contribution to total seasonal precipitation amounts (Klein Tank and Können 2003; Groisman et al. 2005; Zolina et al. 2010).

Table 2.

Frequencies of main Schüepp’s weather classifications are shown for the periods of 1945–94 (Wanner et al. 1998) and 2000–08.

Frequencies of main Schüepp’s weather classifications are shown for the periods of 1945–94 (Wanner et al. 1998) and 2000–08.
Frequencies of main Schüepp’s weather classifications are shown for the periods of 1945–94 (Wanner et al. 1998) and 2000–08.

4. Summary and conclusions

A 9-yr (2000–08) analysis of precipitation for the central and western European Alps has been assembled from Swiss operational radar data. Six regions, aligned with the general topography of the Alps and the radar coverage, were established to enable subanalysis at the scale of hundreds of kilometers. The seasonal distribution of total daily precipitation and the frequency of high precipitation events were characterized on the basis of radar-estimated precipitation rate at the surface.

Total daily precipitation was found to follow a similar annual time series across all regions with minimum values in the winter, values steadily increasing during the spring, consistent highest values in the summer, and then values decreasing in late summer and being more variable throughout the autumn. The SW region exhibits less of a seasonal trend in daily precipitation and also is among the regions that received the most total precipitation per kilometer squared in all seasons. For all regions, high precipitation events (precipitation rate ≥ 20 mm h−1) were most frequent in summer and least frequent in winter over 2000–08.

The highest median and Q90 values of total daily precipitation occurred on days with advective synoptic weather patterns—a synoptic situation that carries moist air from the Mediterranean toward the Alps. A convective synoptic weather classification having minimal surface pressure gradient was associated with the highest median and Q90 values of high-precipitation-rate frequency. Under a convective high synoptic pattern all regions are more likely to receive large total amounts of precipitation resulting from high precipitation rate. Conversely, under an advective south pattern large amounts of total daily precipitation are more likely to result from longer-duration, lower-intensity precipitation rates.

A potentially more valuable and available extension of this analysis is application of the radar-based precipitation distribution at the resolution of the base data (2 × 2 km2) and therefore potential use as a catchment- or glacier-specific precipitation characterization (Machguth et al. 2009). Furthermore, a probability distribution of precipitation at each grid point is envisioned that, when related to synoptic weather patterns, provides precipitation distribution and frequency for potential use in coupling catchment-scale hydrologic models to coarser-resolution climate models. Relating the operational radar-based precipitation analysis to climate models also provides an opportunity for assessment and validation of downscaled climate-model precipitation output, such as in Smiatek et al. (2009) in which multiple regional climate models for the European Alps are reported but are not thoroughly compared because of lack of high-resolution precipitation data. Additional future work is planned to leverage the three-dimensional spatial nature of radar data to study features of vertical structure such as melting layer and distribution of precipitation type (snow/rain).

Acknowledgments

This research was supported by NSF Grant AGS-0937035 and the University of Colorado at Boulder Department of Atmospheric and Oceanic Sciences. The primary authors thank MeteoSwiss for providing operational radar data products and ECMWF for providing free Internet access to ERA Interim data, and we also thank three anonymous reviewers for providing valuable comments.

REFERENCES

REFERENCES
Ahijevych
,
D. A.
,
C. A.
Davis
,
R. E.
Carbone
, and
J. D.
Tuttle
,
2004
:
Initiation of precipitation episodes relative to elevated terrain
.
J. Atmos. Sci.
,
61
,
2763
2769
.
Allamano
,
P.
,
P.
Claps
, and
F.
Laio
,
2009
:
Global warming increases flood risk in mountainous areas
.
Geophys. Res. Lett.
,
36
,
L24404
,
doi:10.1029/2009GL041395
.
Bates
,
B. C.
,
Z. W.
Kundzewicz
,
S.
Wu
, and
J. P.
Palutikof
, Eds.,
2008
:
Climate Change and Water
.
Tech. Paper VI, IPCC Secretariat, 200 pp. [Available online at http://www.ipcc.ch/pdf/technical-papers/climate-change-water-en.pdf.]
Begert
,
M.
,
T.
Schlegel
, and
W.
Kirchhofer
,
2005
:
Homogeneous temperature and precipitation series of Switzerland from 1864 to 2000
.
Int. J. Climatol.
,
25
,
65
80
.
Beniston
,
M.
,
2005
:
The risks associated with climatic change in mountain regions
.
Global Change and Mountain Regions, U. M. Huber, H. K. M. Bugmann, and M. A. Reasoner, Eds., Springer, 511–520
.
Carbone
,
R. E.
,
J. D.
Tuttle
,
D. A.
Ahijevych
, and
S. B.
Trier
,
2002
:
Inferences of predictability associated with warm season precipitation episodes
.
J. Atmos. Sci.
,
59
,
2033
2056
.
Doswell
,
C. A.
,
C.
Ramis
,
R.
Romero
, and
S.
Alonso
,
1998
:
A diagnostic study of three heavy precipitation episodes in the western Mediterranean region
.
Wea. Forecasting
,
13
,
102
124
.
Frei
,
C.
, and
C.
Schär
,
1998
:
A precipitation climatology of the Alps from high-resolution rain-gauge observations
.
Int. J. Climatol.
,
18
,
873
900
.
Germann
,
U.
, and
J.
Joss
,
2004
:
Operational measurement of precipitation in mountainous terrain
.
Weather Radar: Principles and Advanced Applications, P. Meischner, Ed., Springer, 52–77
.
Germann
,
U.
,
G.
Galli
,
M.
Boscacci
, and
M.
Bolliger
,
2006
:
Radar precipitation measurement in a mountainous region
.
Quart. J. Roy. Meteor. Soc.
,
132
,
1669
1692
.
Germann
,
U.
,
M.
Berenguer
,
D.
Sempere-Torres
, and
M.
Zappa
,
2009
:
REAL—Ensemble radar precipitation estimation for hydrology in a mountainous region
.
Quart. J. Roy. Meteor. Soc.
,
135
,
445
456
.
Grazzini
,
F.
,
2007
:
Predictability of large-scale flow conducive to extreme precipitation over the western Alps
.
Meteor. Atmos. Phys.
,
95
,
123
138
.
Groisman
,
P. Ya.
,
R. W.
Knight
,
D. R.
Easterling
,
T. R.
Karl
,
G. C.
Hegerl
, and
V. N.
Razuvaev
,
2005
:
Trends in intense precipitation in the climate record
.
J. Climate
,
18
,
1326
1350
.
Hohenegger
,
C.
,
A.
Walser
,
W.
Langhans
, and
C.
Schär
,
2008
:
Cloud resolving ensemble simulations of the August 2005 Alpine flood
.
Quart. J. Roy. Meteor. Soc.
,
134
,
889
904
.
Joss
,
J. S.
, and
Coauthors
,
1998
:
Operational Use of Radar for Precipitation Measurements in Switzerland
.
vdf Hochschulverlag AG an der ETH Zürich, 108 pp
.
Klein Tank
,
A. M. G.
, and
G. P.
Können
,
2003
:
Trends in indices of daily temperature and precipitation extremes in Europe, 1946–99
.
J. Climate
,
16
,
3665
3680
.
Kundzewicz
,
Z. W.
, and
Coauthors
,
2007
:
Freshwater resources and their management
.
Climate Change 2007: Impacts, Adaptation and Vulnerability, M. L. Parry et al., Eds., Cambridge University Press, 173–210. [Available online at http://www.ipcc.ch/pdf/assessment-report/ar4/wg2/ar4-wg2-chapter3.pdf.]
Laing
,
A. G.
, and
J. M.
Fritsch
,
1997
:
The global population of mesoscale convective complexes
.
Quart. J. Roy. Meteor. Soc.
,
123
,
389
405
.
Lin
,
Y. L.
,
S.
Chiao
,
T. A.
Wang
,
M. L.
Kaplan
, and
R. P.
Weglarz
,
2001
:
Some common ingredients for heavy orographic rainfall
.
Wea. Forecasting
,
16
,
633
660
.
Machguth
,
H.
,
F.
Paul
,
S.
Kotlarski
, and
M.
Hoelzle
,
2009
:
Calculating distributed glacier mass balance for the Swiss Alps from regional climate model output: A methodical description and interpretation of the results
.
J. Geophys. Res.
,
114
,
D19106
,
doi:10.1029/2009JD011775
.
Martius
,
O.
,
E.
Zenklusen
,
C.
Schwierz
, and
H.
Davies
,
2006
:
Episodes of Alpine heavy precipitation with an overlying elongated stratospheric intrusion: A climatology
.
Int. J. Climatol.
,
26
,
1149
1164
.
Massacand
,
A. C.
,
H.
Wernli
, and
H. C.
Davies
,
1998
:
Heavy precipitation on the Alpine southside: An upper-level precursor
.
Geophys. Res. Lett.
,
25
,
1435
1438
.
Moberg
,
A.
, and
Coauthors
,
2006
:
Indices for daily temperature and precipitation extremes in Europe analyzed for the period 1901–2000
.
J. Geophys. Res.
,
111
,
D22106
,
doi:10.1029/2006JD007103
.
Overeem
,
A.
,
T. A.
Buishand
, and
I.
Holleman
,
2009a
:
Extreme rainfall analysis and estimation of depth-duration-frequency curves using weather radar
.
Water Resour. Res.
,
45
,
W10424
,
doi:10.1029/2009WR007869
.
Overeem
,
A.
,
I.
Holleman
, and
A.
Buishand
,
2009b
:
Derivation of a 10-year radar-based climatology of rainfall
.
J. Appl. Meteor. Climatol.
,
48
,
1448
1463
.
Panziera
,
L.
, and
U.
Germann
,
2010
:
The relation between airflow and orographic precipitation on the southern side of the Alps as revealed by weather radar
.
Quart. J. Roy. Meteor. Soc.
,
136
,
222
238
.
Rudari
,
R.
,
D.
Entekhabi
, and
G.
Roth
,
2004
:
Terrain and multiple-scale interactions as factor in generating extreme precipitation events
.
J. Hydrometeor.
,
5
,
390
404
.
Schiesser
,
H.
,
R.
Houze
, and
H.
Huntrieser
,
1995
:
The mesoscale structure of severe precipitation systems in Switzerland
.
Mon. Wea. Rev.
,
123
,
2070
2097
.
Schmidli
,
J.
, and
C.
Frei
,
2005
:
Trends of heavy precipitation and wet and dry spells in Switzerland during the 20th century
.
Int. J. Climatol.
,
25
,
753
771
.
Schmidli
,
J.
,
C.
Schmutz
,
C.
Frei
,
H.
Wanner
, and
C.
Schär
,
2002
:
Mesoscale precipitation variability in the region of the European Alps during the 20th century
.
Int. J. Climatol.
,
22
,
1049
1074
.
Schönwiese
,
C.-D.
,
J.
Rapp
,
T.
Fuchs
, and
M.
Denhard
,
1994
:
Observed climate trends in Europe 1891–1990
.
Meteor. Z.
,
3
,
22
28
.
Schüepp
,
M.
,
1979
:
Klimatologie der Schweiz, Band III: Witterungsklimatologie (Climatology of Switzerland, Volume 3: Medium-Range Weather Climatology)
.
Schweizerische Meteorologische Anstalt, 93 pp
.
Smiatek
,
G.
,
H.
Kunstmann
,
R.
Knoche
, and
A.
Marx
,
2009
:
Precipitation and temperature statistics in high-resolution regional climate models: Evaluation for the European Alps
.
J. Geophys. Res.
,
114
,
D19107
,
doi:10.1029/2008JD011353
.
Smith
,
R. B.
,
Q. F.
Jiang
,
M. G.
Fearon
,
P.
Tabary
,
M.
Dorninger
,
J. D.
Doyle
, and
R.
Benoit
,
2003
:
Orographic precipitation and air mass transformation: An Alpine example
.
Quart. J. Roy. Meteor. Soc.
,
129
,
433
454
.
Sodemann
,
H.
, and
E.
Zubler
,
2010
:
Seasonal and inter-annual variability of the moisture sources for Alpine precipitation during 1995–2002
.
Int. J. Climatol.
,
30
,
947
961
.
Stefanicki
,
G.
,
P.
Talkner
, and
R. O.
Weber
,
1998
:
Frequency changes of weather types in the Alpine region since 1945
.
Theor. Appl. Climatol.
,
60
,
47
61
.
Wanner
,
H.
,
E.
Salvisberg
,
R.
Rickli
, and
M.
Schüepp
,
1998
:
50 years of Alpine weather statistics (AWS)
.
Meteor. Z.
,
7
,
99
111
.
Weisman
,
M. L.
, and
J. B.
Klemp
,
1984
:
The structure and classification of numerically simulated convective storms in directionally varying wind shears
.
Mon. Wea. Rev.
,
112
,
2479
2498
.
Widmann
,
M.
, and
C.
Schär
,
1997
:
A principal component and long-term trend analysis of daily precipitation in Switzerland
.
Int. J. Climatol.
,
17
,
1333
1356
.
Wüest
,
M.
,
C.
Frei
,
A.
Altenhoff
,
M.
Hagen
,
M.
Litschi
, and
C.
Schär
,
2010
:
A gridded hourly precipitation dataset for Switzerland using rain-gauge analysis and radar-based disaggregation
.
Int. J. Climatol.
,
30
,
1764
1775
.
Zolina
,
O.
,
C.
Simmer
,
A.
Kapala
,
S.
Bachner
,
S. K.
Gulev
, and
H.
Maechel
,
2008
:
Seasonally dependent changes of precipitation extremes over Germany since 1950 from a very dense observational network
.
J. Geophys. Res.
,
113
,
D06110
,
doi:10.1029/2007JD008393
.
Zolina
,
O.
,
C.
Simmer
,
S. K.
Gulev
, and
S.
Kollet
,
2010
:
Changing structure of European precipitation: Longer wet periods leading to more abundant rainfalls
.
Geophys. Res. Lett.
,
37
,
L06704
,
doi:10.1029/2010GL042468
.