Abstract

Operational radar data reveal that precipitation systems occurring on the southern side of the Alps near Locarno, Switzerland, follow seasonal patterns of vertical reflectivity structure. Storms occurring in summer are more convective than winter season storms as indicated by more frequent observation of reflectivity at higher altitudes during summer. Individual precipitation events occurring year-round are classified by comparison to average seasonal vertical reflectivity structure. Seasonal classification of individual storms reveals a transition between winter- and summer-type storms during spring and fall that follows changes in average daily surface temperature. In addition to distinct vertical structure, summer- and winter-type storms have differences in duration, intensity, and interval between storms. Although summer- and winter-type storms result in a similar amount of total precipitation, summer-type storms have shorter duration, and therefore greater intensity. The dependence of storm types on temperature has implications for intensification of the hydrologic cycle due to climate change. Warmer winter, spring, or fall surface temperatures may affect average precipitation intensity by increasing the number of days per year that experience more intense convective precipitation while decreasing the probability of less intense stratiform precipitation.

1. Introduction

Climate aspects related to precipitation are commonly discussed in terms of spatial distribution of rainfall at the surface (i.e., amount, rate, and duration) while long-term patterns of vertical precipitation structure remain elusive. In this study we use multiyear operational radar reflectivity data from southern Switzerland to investigate the seasonal variability and annual cycle of vertical precipitation structure. Results from previous experimental campaigns are used to relate seasonal and annual reflectivity patterns to atmospheric stability and cloud microphysics observed during individual cases. Furthermore, we examine the effects of surface temperature on the vertical structure of reflectivity and implications for the impact of climate change on the hydrologic cycle.

To our knowledge, multiple-year observations of vertical precipitation structure in the European Alps have not been investigated. Most precipitation climatologies are based on surface observations from rain gauges (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; Molnar and Burlando 2008). A few studies have used multiyear radar data as a basis for a regional precipitation climatology (Overeem et al. 2009; Rudolph et al. 2011). However, none of these climatologies investigate long-term patterns of vertical precipitation structure. Houze et al. (2001) analyzed orographically influenced vertical reflectivity structure in southern Switzerland over two fall seasons in 1998 and 1999. Our study is unique because we include 7 years of data in our investigation of vertical reflectivity structure. In addition, we explore the relationship between total amount, duration, and intensity of precipitation that results from changes in the vertical structure. It is important to establish a link (or lack thereof) between vertical precipitation structure and surface precipitation measurements in order to understand the implications that various storm types, as identified by the vertical structure of reflectivity, have for the hydrologic cycle.

Results from the Mesoscale Alpine Programme (MAP) characterize the vertical structure of individual storms that occurred in southern Switzerland during the fall of 1999 (Houze et al. 2001; Medina and Houze 2003; Yuter and Houze 2003; Rotunno and Houze 2007). The MAP cases provide unique information about the vertical motion and microphysics responsible for orographically enhanced precipitation, such as whether each case has a more convective or stratiform precipitation structure. But are these individual cases from the 1999 fall season relevant to the rest of the year and representative for the typical precipitation structure in fall for southern Switzerland? Here, we apply results from the MAP case studies to explain seasonal precipitation characteristics. We infer the mesoscale conditions and microphysical processes that dominate precipitation in each season from similarities in the vertical structure of reflectivity between seasonal averages and individual MAP cases.

In this study we also investigate the relationship between temperature and seasonal changes in the precipitation structure as an indicator of potential climate change impacts on the hydrologic cycle. Warming due to anthropogenic climate change is expected to affect both the space and time distribution of precipitation through changes in precipitation intensity, duration, and interval between events (Trenberth et al. 2003). Global and regional climate models predict that changes in surface precipitation characteristics will coincide with an overall intensification of the hydrologic cycle (Solomon et al. 2007; Giorgi et al. 2011). In addition, our previous work indicates that changes in synoptic patterns over the twenty-first century in Europe will increase the proportion of convective precipitation for Swiss river basins (Rudolph et al. 2012). This study further examines the possibility that precipitation events in southern Switzerland may become more convective as temperature increases and how an increase in convective precipitation may contribute to an increase in precipitation intensity.

2. Data

a. Monte Lema radar

The MeteoSwiss radar network provides high-quality three-dimensional multiyear reflectivity and estimated rainfall data for the Swiss Alps. The radar network comprises three C-band Doppler weather radars that scan 20 elevations every 5 min and monitor radar reflectivity up to a height of 12 km above mean sea level (MSL) and a range of 230 km. MeteoSwiss radar data 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). Further detailed descriptions of MeteoSwiss radar and data products are found in Joss et al. (1998), Germann and Joss (2004), Germann et al. (2006, 2009), and Panziera and Germann (2010).

OVERVIEW is a MeteoSwiss data product that is particularly useful for the analysis of vertical reflectivity structure. The OVERVIEW dataset provides reflectivity at 1-km intervals between a vertical range of 1–12 km MSL with a vertical resolution of ±0.5 km (i.e., data at height 2 km contain measurements made between 1.5 and 2.5 km MSL; Joss et al. 1998). The horizontal resolution is 2 × 2 km2 and the time resolution is 5 min. The reflectivity data are stored as 16 classes with the lowest class for reflectivity <13 dBZ and subsequent classes in 3-dBZ intervals (13–16 dBZ, 16–19 dBZ, 19–22 dBZ, etc.) up to the highest class for reflectivity >55 dBZ (Joss et al. 1998). OVERVIEW data from the MeteoSwiss Monte Lema radar (Fig. 1) for March 2004 through February 2011 are used for this study. It is noted that MeteoSwiss is in the process of upgrading their radar network with two new radars and the addition of dual-polarization capability at new and existing radar locations (Gabella et al. 2009). As part of the upgrade project the Monte Lema radar was replaced in May 2011, which limits the data analysis to February 2011.

Fig. 1.

Location of Locarno (red star), Monte Lema radar (ML; red box), and horizontal range of radar visibility used for this study (yellow boxes). The study area is divided into north and south sectors. Visibility is limited in the north sector by topography and restricted in the south sector to maintain a similar area as the north. Radar data have spatial resolution of 2 × 2 km2 (shown for reference in lower right). Locations of rain gauge stations (blue boxes) are also indicated (abbreviations as in Table 1). Topography contours are in 1000 m with white denoting elevation <1000 m MSL.

Fig. 1.

Location of Locarno (red star), Monte Lema radar (ML; red box), and horizontal range of radar visibility used for this study (yellow boxes). The study area is divided into north and south sectors. Visibility is limited in the north sector by topography and restricted in the south sector to maintain a similar area as the north. Radar data have spatial resolution of 2 × 2 km2 (shown for reference in lower right). Locations of rain gauge stations (blue boxes) are also indicated (abbreviations as in Table 1). Topography contours are in 1000 m with white denoting elevation <1000 m MSL.

The Monte Lema radar is located near Locarno, Switzerland, at the southern edge of the European Alps (Fig. 1). Locarno provides an interesting study location since it experiences an annual climate cycle that is influenced by close proximity to the Alps and the Mediterranean Sea (Frei and Schär 1998; Rudolph et al. 2011). An additional benefit of this location is the inclusion of data from the Monte Lema radar in intensive observation periods (IOPs) and subsequent analysis resulting from the 1999 MAP experiment (Bougeault et al. 2001; Houze et al. 2001; Medina and Houze 2003; Yuter and Houze 2003; Rotunno and Houze 2007).

Because of the difference in terrain on the north and south sides of the radar, the dataset used for this study was divided into north and south sectors (Fig. 1). The physical blockage imposed by the proximity of the Alps to the north results in a reduced range of visibility at lower elevations on the northern side of the radar. The Po River valley lies to the south of the radar, so southward visibility is largely unimpeded by the terrain (Germann et al. 2006). Since the radar’s physical elevation is 1625 m MSL, only data from ≥2 km MSL are utilized. At an altitude of 2 km, the radar beam is primarily unblocked toward the north up to a range of 60 km from the radar (Fig. 1). Therefore, we limit the horizontal cross sections of reflectivity at all altitudes in both the north and south sectors to a range of 60 km to maintain consistency in the volume of data analyzed at each altitude and in each sector. The radar’s maximum scan elevation (40°) inhibits measurement of reflectivity directly above the radar at higher altitudes and results in a “donut hole” of missing data over the radar (Germann et al. 2006). The scan geometry causes the radius of missing data above the radar to increase with height, so it is maximized at 12 km MSL. To maintain consistency in the number of pixels represented at each altitude, pixels with missing data at 12 km (representing a 6-km radius around location of radar) were omitted from lower altitudes prior to analysis.

b. Rain gauge network

The selection criteria for rain gauges used in this study are 1) the rain gauges are located at elevations below 2 km, so they are beneath the minimum altitude of analyzed radar reflectivity, and 2) the rain gauge locations form a north–south transect across the radar range. Six rain gauge stations located within the range of selected radar data meet these criteria (Table 1 and Fig. 1). Utilizing data from multiple locations within the radar range increases the probability that localized storms captured by the radar also pass over the rain gauges. Since precipitation typically approaches the study area from the west, southwest, or south (Rudolph et al. 2011), the north–south alignment of the rain gauges also improves the likelihood of intercepting the eastward moving storms. The rain gauge data are available at 10-min intervals with precision of 0.1 mm (https://gate.meteoswiss.ch/idaweb). The rain gauge data are quality controlled by MeteoSwiss following guidelines from the World Meteorological Organization (WMO 1983). The first step includes a real-time quality control that covers the status of each instrument and plausible maximum and minimum values for each observation based on long-term climatological values. In a second step, correction and calibration procedures are applied. A temporal and spatial consistency check is then applied to the data from the entire network using Vienna Enhanced Resolution Analysis Quality Control (VERAQC; http://www.univie.ac.at/amk/vera/; Steinacker et al. 1997, 2000) and NORM90 (Begert et al. 2003). Detailed descriptions of the quality control steps and procedures are found in Perl et al. (2009).

Table 1.

Rain gauge stations used in this study.

Rain gauge stations used in this study.
Rain gauge stations used in this study.

3. Method

a. Determining seasonal vertical structure of reflectivity

In this study, seasonal contoured frequency by altitude diagrams (CFADs) are generated to provide a view of the average vertical reflectivity structure over the course of winter [December–February (DJF); the year associated with the winter season is the year of January/February], spring [March–May (MAM)], summer [June–August (JJA)], and fall [September–November (SON)]. CFADs provide visualization of the number of occurrences of specific radar reflectivity values at predefined altitudes over a defined time period (Yuter and Houze 1995). CFADs have previously been used to determine whether a specific storm has a more stratiform or convective vertical structure (Yuter and Houze 1995). For the seasonal CFADs, the count of observed reflectivity for each reflectivity class r at each altitude z (cr,z) is calculated as

 
formula

where N is the number of 5-min observation periods over the course of the season and obst,r,z is the number of pixels with reflectivity class r at altitude z in the observed horizontal cross section at time step t.

Normalizing cr,z by the number of pixels and time steps allows the comparison of CFAD contours between observation periods having various durations, whether it is seasons differing in number of days, or comparison of seasonal CFADs to a CFAD for an individual storm that may only last a few hours. The normalization also accounts for horizontal dimensions so that CFADs based on different spatial extents may be compared; therefore, we may directly compare CFADs for the north and south regions (Fig. 1) even though total surface area representing the north region is somewhat smaller than the south because of radar visibility. The normalization is carried out by dividing cr,z by the number of 5-min observation periods N and also the total number of pixels p observed at each time step:

 
formula

CFADs plotted with normalized counts as described above allow qualitative comparison of vertical structure between seasons and/or individual storms. However, statistical comparison of CFADs requires a more quantitative method, so the centroid is calculated for each CFAD. The CFAD centroid is essentially the center of mass of the CFAD plot and provides a quantitative measure suitable for statistical analysis. The centroid of the CFAD has x and y coordinates (x centroid and y centroid) with x corresponding to reflectivity and y corresponding to height or altitude.

b. Individual storm identification and classification of vertical structure

In addition to exploring vertical reflectivity structure on a seasonal basis, CFADs and CFAD centroids are also generated for individual storms [following Eqs. (1) and (2); section 3a]. Individual precipitation events are identified using rain gauge data from the six stations listed in Table 1. Several previous analyses have identified precipitation events using a threshold of 1-mm rainfall R received over a 24-h period (Groisman and Knight 2008; Zolina et al. 2010; Giorgi et al. 2011). A similar approach is taken for this study. The start time of each event is the first 10-min period when R ≥ 0.1 mm is recorded at any single rain gauge station and at least 1-mm precipitation is received in 24 h or less. The end time for an individual precipitation event is the last 10-min period when R ≥ 0.1 mm is recorded and no precipitation is measured over the next three consecutive hours at any of the six stations. The 3-h interval without precipitation used to indicate the separation between storms is consistent with the spectral analysis of continental European storm frequency reported in Fraedrich and Larnder (1993) and the threshold used in Molnar and Burlando (2008; P. Molnar 2011, personal communication). Using the precipitation gauge–based storm criteria we identified 559 precipitation events between March 2004 and February 2011. Rain gauge data from all six stations are used to derive storm duration and interval between storms. Total precipitation amount and maximum 10-min precipitation rate are from the Locarno station. The average precipitation rate for each storm is calculated using the total amount of precipitation received at Locarno and the storm duration. The analysis certainly does not capture all events (e.g., isolated connective cells that did not pass over the rain gauges and are therefore excluded). In the same way, shallow events (<1500 m MSL) are also not captured by the analysis. However, we believe that the 559 precipitation events included in our analysis are sufficient to provide a representative sample.

After identifying individual storms, a CFAD and associated CFAD centroid coordinates are generated for each precipitation event. As with the seasonal CFADs and CFAD centroids described above (section 3a), the storm-specific CFADs are also normalized following Eq. (2), so that the storm- and seasonal-based CFADs and centroids may be directly compared. In addition, we use the centroids from the individual storm CFADs to classify the vertical structure of reflectivity for each precipitation event as winter type (DJF), summer type (JJA), or spring/fall type (MAM/SON) by finding its nearest neighbor among the seasonal CFAD centroids. Seasonal CFAD centroids from the northern sector were used for classification of individual storms since most of the rain gauge stations are located north of the radar.

4. Results

a. Seasonal variation of vertical structure

CFADs for each season from March 2004 through February 2011 show distinct seasonal differences that are consistent from year to year in both the north and south sectors (Fig. 2). For this study, we define vertical extent of reflectivity as the maximum altitude where the 0.004 count contour (equivalent to a spatial frequency of 0.3 km−2 day−1) appears in the seasonal CFAD (Fig. 2). Vertical extent of reflectivity is maximized in summer (JJA) when reflectivity extends to 9–10 km MSL and minimized in winter (DJF) when reflectivity extends only to 5–6 km MSL both in the north and south sectors. Therefore, reflectivity of 15 dBZ is observed with the same frequency at nearly twice the altitude in summer as in winter and indicates that precipitating systems in summer extend higher into the atmosphere. Seasonal CFADs for spring (MAM) and fall (SON) have intermediate values of vertical extent of 6–7 km, which is 3–4 km lower than JJA and slightly higher (~1 km) than DJF.

Fig. 2.

Seasonal CFADs for the (left) north and (right) south sectors for March 2004–February 2011 (DJF—winter; MAM—spring; JJA—summer; SON—fall). Color contours represent average seasonal CFAD over all years. Black lines are the 0.004-count contour for each individual year.

Fig. 2.

Seasonal CFADs for the (left) north and (right) south sectors for March 2004–February 2011 (DJF—winter; MAM—spring; JJA—summer; SON—fall). Color contours represent average seasonal CFAD over all years. Black lines are the 0.004-count contour for each individual year.

The seasonal CFADs also show differences in magnitude of reflectivity (Fig. 2). Larger magnitude of reflectivity is more frequent near the ground in JJA than in DJF. In the north sector at 3 km MSL, reflectivity of 35 dBZ occurs in JJA with similar frequency as 27 dBZ in DJF. Maximum reflectivity in MAM and SON (33–34 dBZ at 2 km in north region) approaches the magnitude of maximum reflectivity seen in JJA (35 dBZ at 3 km in north region; Fig. 2). A distinct difference between MAM and SON versus JJA for both the north and south is that larger reflectivity values appear at 3 km MSL in JJA and at a lower altitude of 2 km in MAM and SON.

Although the seasonal CFADs from the north and south sectors are similar in vertical extent and magnitude of reflectivity, the frequency of observed reflectivity is larger on the north side of the radar (Fig. 2). For example, comparison of the north and south JJA seasonal CFADs shows that 20-dBZ reflectivity is approximately 50% more frequent at 3 km MSL in the north sector than in the south (0.012 count contour for north versus 0.008 count contour for south; JJA in Fig. 2). Higher frequency of observed reflectivity at each altitude on the north side of the radar is consistent throughout the seasonal CFADs.

Seasonal CFAD centroid locations for the north sector (as described in section 3a and based on Fig. 2) for each year of 2004 through 2010 (2005 through 2011 for DJF) are shown in Fig. 3 (bold diamond symbols). From this point on, our results and discussion of seasonal CFAD centroids are solely focused on the north sector because the seasonal CFAD centroids for the north and south sectors are similar (not shown), and the majority of rain gauges that we use for storm identification are located north of the radar (Fig. 1). Enhanced reflectivity values observed at higher elevations in JJA are reflected in the greatest magnitude y centroid (range of 4.1 to 4.7 km MSL) as determined from the CFAD (y axis corresponds to altitude; Fig. 3). The JJA seasons’ CFAD y centroids are significantly greater (with 95% confidence) than winter, spring, or fall. DJF has the lowest CFAD y centroids (95% confidence) ranging from 2.7 to 2.9 km. CFAD y centroids for MAM and SON are not found to significantly differ (ranging from 2.9 to 3.4 km). JJA also experiences larger reflectivity values than the other seasons (with 95% confidence, CFAD x axis corresponds to reflectivity in Fig. 3). This results in JJA having the largest CFAD x centroids (ranging from 21 to 23 dBZ). The x centroids (x axis corresponds to reflectivity) for DJF and MAM/SON vary between 18–21 and 19–22 dBZ, respectively. The intermediate values of the CFAD x and y centroids for MAM and SON as compared to JJA and DJF imply that radar-observed reflectivity structure undergoes a transition from DJF to JJA, and vice versa.

Fig. 3.

Centroids as determined from the north sector seasonal CFADs shown in Fig. 2 (bold diamond symbols) and individual storm CFAD centroids (square symbols) with x centroids representing reflectivity and y centroids representing altitude. Colors represent different seasons with winter (DJF) in blue, spring (MAM) in green, summer (JJA) in red, and fall (SON) in orange. Individual storms are classified by nearest neighbor among the seasonal CFAD centroids. Fall- and spring-type storms (MAM/SON; indicated in light green) are combined into a single class for the individual storm analysis because no significant difference is observed between the seasonal spring and fall x (reflectivity) and y (altitude) centroids.

Fig. 3.

Centroids as determined from the north sector seasonal CFADs shown in Fig. 2 (bold diamond symbols) and individual storm CFAD centroids (square symbols) with x centroids representing reflectivity and y centroids representing altitude. Colors represent different seasons with winter (DJF) in blue, spring (MAM) in green, summer (JJA) in red, and fall (SON) in orange. Individual storms are classified by nearest neighbor among the seasonal CFAD centroids. Fall- and spring-type storms (MAM/SON; indicated in light green) are combined into a single class for the individual storm analysis because no significant difference is observed between the seasonal spring and fall x (reflectivity) and y (altitude) centroids.

b. Classification of individual storms reveals annual pattern of vertical structure

The significant differences in the seasonal CFAD centroid locations (Fig. 3, bold diamond symbols) enable us to classify individual storms as winter- (DJF), summer- (JJA), or spring/fall (MAM/SON)-type systems based on the comparison of individual storm CFAD centroids to seasonal CFAD centroids. Individual storms are classified by identifying the nearest neighbor among the seasonal CFAD centroids (Fig. 3), so the individual storm classifications (Fig. 3, square symbols; section 3b) are independent of the season in which the storm occurred. MAM- and SON-type storms are grouped together as classification MAM/SON because their respective seasonal CFAD centroids are indistinguishable (not significantly different at 95% confidence). The seasonal and individual storm CFAD centroids in Fig. 3 also indicate that mean reflectivity increases (x centroid) with vertical extent (y centroid).

Table 2 shows the seasonal distribution of all individual precipitation events occurring between March 2004 and February 2011 and their corresponding storm classifications based on the storms’ vertical structure of reflectivity (square symbols in Fig. 3). An annual pattern is apparent with individual storm classifications following a progression from DJF- to MAM/SON- to JJA- to MAM/SON- to DJF-type storms as time moves forward through each year. The vertical structure of DJF-type storms (i.e., storms with vertical extent <5–6 km and reflectivity mainly <20–30 dBZ) is most common during winter, but also occurs in spring and fall and only occasionally in summer (Table 2 and Fig. 4). Vertical structure of reflectivity typical for MAM/SON-type storms is mainly observed during spring and fall. Although MAM/SON-type storms appear in all seasons, they are less common in winter. JJA-type storms (i.e., storms with vertical extent <9–10 km and high reflectivities of <30–35 dBZ) exist mainly during summer, but also appear in late spring and early fall.

Table 2.

The number (and percentage of seasonal total) of each storm-type classification that occurred in each season over the years 2004–11.

The number (and percentage of seasonal total) of each storm-type classification that occurred in each season over the years 2004–11.
The number (and percentage of seasonal total) of each storm-type classification that occurred in each season over the years 2004–11.
Fig. 4.

Occurrence of individual precipitation events (square symbols) for (a) spring and (b) fall of the years 2004–10 as a function of storm type based on Fig. 2. Solid lines indicate 3-day average temperature at Locarno for each year. Colors indicate individual years.

Fig. 4.

Occurrence of individual precipitation events (square symbols) for (a) spring and (b) fall of the years 2004–10 as a function of storm type based on Fig. 2. Solid lines indicate 3-day average temperature at Locarno for each year. Colors indicate individual years.

c. Relationship between vertical structure of reflectivity and surface temperature

Throughout spring a steady increase in average daily temperature is experienced with decreasing average daily temperature as fall progresses (solid lines in Fig. 4). The vertical structure of JJA-type storms becomes more common in late spring when average daily temperature is higher (Fig. 4a); likewise, the vertical structure associated with JJA-type storms becomes less frequent as average daily temperature decreases during fall (Fig. 4b). Therefore, as daily average temperature increases in spring, the frequency reflectivity observed at higher altitudes also increases. Similarly, individual storms that occur in early spring and late fall commonly have a vertical structure of DJF-type storms. MAM/SON-type vertical structure is more frequently observed during the middle of spring and fall as the reflectivity structure transitions from DJF- to JJA-type storms in spring and JJA- to DJF-type storms in fall. Based on the 2004–10 data, there is no indication that the occurrence of JJA- or DJF-type storms has consistently shifted to an earlier or later onset or decline in spring or fall.

To further explore potential effects of temperature on the vertical reflectivity structure, CFADs for individual storms and surface temperature from Locarno are combined to determine the conditional probability of experiencing a particular storm-type classification given the 3-day average daily temperature (Fig. 5). Average 3-day temperature is used as the basis for conditional probability in order to align average surface temperature with the observed frequency of continental European frontal systems that occur, on average, approximately every 3 days (Fraedrich and Larnder 1993). In Fig. 5, the relationship between storm type and surface temperature, as evidenced by conditional probability, is based on all storms occurring between March 2004 and February 2011. The smooth curves (black lines) in Fig. 5 are the result of fitting a vector general linearized model (referred to as VGLM) to the observed data following Yee and Hastie (2003) and Yee (2008). The VGLM fit results in coefficients an and bn for the log-linear relationships:

 
formula
 
formula

where PDJF + PMAM/SON + PJJA = 1. Here PDJF, PMAM/SON, and PJJA are the respective probabilities for classes DJF, MAM/SON, and JJA when the 3-day average temperature is T. Applying the VGLM fit to the 2004–11 data, the log-linear relationship results in the following coefficients: a1 = −2.31, b1 = 0.20, a2 = −11.68, and b2 = 0.70.

Fig. 5.

Probability of each storm-type class as a function of the 3-day average daily temperature for all observed precipitation systems occurring between March 2004 and February 2011: blue diamonds—DJF; green triangles—MAM/SON; red squares—JJA. Black lines are the VGLM fit of conditional probabilities as described in section 4c.

Fig. 5.

Probability of each storm-type class as a function of the 3-day average daily temperature for all observed precipitation systems occurring between March 2004 and February 2011: blue diamonds—DJF; green triangles—MAM/SON; red squares—JJA. Black lines are the VGLM fit of conditional probabilities as described in section 4c.

DJF-type vertical structure is the most probable storm classification when average daily temperature is less than 11°C (Fig. 5). MAM/SON becomes the most common storm type when the surface temperature ranges between 11° and 18°C. Above 18°C JJA-type vertical structure is the most likely classification, and only JJA-type vertical structure is observed when temperature exceeds 24°C. DJF-type vertical structure is not observed when daily average surface temperature rises above 23°C, and JJA-type storms are not observed below 14°C.

d. Surface precipitation characteristics associated with vertical structure of reflectivity

Beyond changes in vertical structure of radar reflectivity, the transition in storm type has additional implications for surface precipitation characteristics. Table 3 lists the mean values of total precipitation, storm duration, interval between storms, average precipitation rate over the entire storm, and maximum precipitation rate within 10 min for each storm-type classification regardless of the season in which the individual storms occurred. The entire range of data is shown in Fig. 6. JJA-type storms have the shortest duration (mean duration 16.7 h) as compared to MAM/SON and DJF types, each having a mean duration of 24.2 h (Table 3 and Fig. 6a). The precipitation characteristics of each storm type shown in Table 3 are compared by using a Wilcoxon nonparametric test at 95% confidence. No significant difference is found in the mean length of the dry intervals following JJA- and MAM/SON-type storms (~60 h; Table 3 and Fig. 6b). The dry interval of 95 h following DJF-type storms is the longest. Despite the difference in duration, the Wilcoxon nonparametric comparison finds no significant difference in total precipitation for JJA- and DJF-type precipitation events (Table 3 and Fig. 6c). The only type that produces significantly different total precipitation is MAM/SON (mean total of 27.2 mm), which is greater than both DJF and JJA (14.9 and 21.8 mm, respectively; Table 3 and Fig. 6c). Since no statistical difference is found in total amount of precipitation received from JJA- or DJF-type storms, but JJA-type storms have shorter duration, it follows that the average precipitation rate recorded during JJA-type storms (1.4 mm h−1) is more than 3 times greater than for DJF (0.4 mm h−1; Table 3). In addition, JJA-type storms produce the largest maximum 10-min precipitation rate [3.5 mm (10 min)−1], nearly double the maximum 10-min precipitation rate of MAM/SON-type storms [1.4 mm (10 min)−1] and 7 times greater than DJF-type vertical structure [0.5 mm (10 min)−1; Table 3 and Fig. 6d].

Table 3.

Mean precipitation characteristics for each storm-type classification. Letters [in brackets] indicate statistically significant groupings by Wilcoxon nonparametric comparison at 95% confidence. For instance, the total precipitation for DJF and JJA is not statistically different (both DJF and JJA denoted by [A]), but both DJF and JJA differ from MAM/SON (MAM/SON denoted by [B]). Average rate is statistically different for each classification (as denoted by [A] for DJF, [B] for MAM/SON, and [C] for JJA). The full range of data for each classification is shown in Fig. 6.

Mean precipitation characteristics for each storm-type classification. Letters [in brackets] indicate statistically significant groupings by Wilcoxon nonparametric comparison at 95% confidence. For instance, the total precipitation for DJF and JJA is not statistically different (both DJF and JJA denoted by [A]), but both DJF and JJA differ from MAM/SON (MAM/SON denoted by [B]). Average rate is statistically different for each classification (as denoted by [A] for DJF, [B] for MAM/SON, and [C] for JJA). The full range of data for each classification is shown in Fig. 6.
Mean precipitation characteristics for each storm-type classification. Letters [in brackets] indicate statistically significant groupings by Wilcoxon nonparametric comparison at 95% confidence. For instance, the total precipitation for DJF and JJA is not statistically different (both DJF and JJA denoted by [A]), but both DJF and JJA differ from MAM/SON (MAM/SON denoted by [B]). Average rate is statistically different for each classification (as denoted by [A] for DJF, [B] for MAM/SON, and [C] for JJA). The full range of data for each classification is shown in Fig. 6.
Fig. 6.

Precipitation characteristics of each storm type based on rain gauge data: (a) storm duration, (b) interval between storms, (c) total precipitation amount, and (d) maximum 10-min precipitation rate. Boxes indicate median and interquartile range (25th to 75th quantile) with extended hash marks at the 10% and 90% quantiles.

Fig. 6.

Precipitation characteristics of each storm type based on rain gauge data: (a) storm duration, (b) interval between storms, (c) total precipitation amount, and (d) maximum 10-min precipitation rate. Boxes indicate median and interquartile range (25th to 75th quantile) with extended hash marks at the 10% and 90% quantiles.

5. Discussion

a. Comparison of seasonal vertical structure to MAP case studies

The greater vertical extent, higher reflectivity values near ground level, and slightly elevated maximum reflectivity in the summer season provide evidence that summer precipitation is more buoyantly driven than precipitation occurring in winter (Fig. 2). The observation of enhanced reflectivity at higher altitudes that coincides with increased probability of JJA-type vertical structure as surface temperature increases is indicative of more unstable stratification of the boundary layer during the summer season. The increased convective nature of summertime precipitation implies stronger vertical air motion (Houze 1993, 1997). Weaker vertical air motion in winter, as evidenced by lower altitude reflectivity, suggests that wintertime precipitation has a more stratiform character. However, convective and stratiform precipitation are not mutually exclusive (Houze 1997). Instead, a single precipitation system may consist of both convective and stratiform areas depending on the range of vertical velocities within the storm. The presence of embedded convective cells can be associated with intense orographic precipitation, such as the historic Alpine flood in central Switzerland during August 2005 (Hohenegger et al. 2008; Langhans et al. 2011).

Several papers describe radar-observed vertical precipitation structure that occurred during the MAP experiment in 1999 (Houze et al. 2001, hereafter H01; Medina and Houze 2003, hereafter MH03; Yuter and Houze 2003, hereafter YH03; Rotunno and Houze 2007, hereafter RH07). In addition to the Monte Lema radar data used in our study, the MAP analyses utilize data from a temporarily deployed National Center for Atmospheric Research (NCAR) S-band dual-polarization Doppler radar (S-Pol) and a vertically pointing radar to derive vertical velocity and distributions of hydrometeor types (H01; MH03; YH03). The vertical structure depicted in the fall season CFAD for years 2004–10 (Fig. 2) has many similarities to the vertical cross sections of radar reflectivity from the fall seasons of 1998 and 1999 observed during MAP (H01). The current study and H01 both observed maximum reflectivity values at 2–3 km MSL and increased frequency of radar returns between 2–4 km MSL with greater magnitude of reflectivity and frequency toward the north of the Monte Lema radar as compared to the south side. H01 observed that the melting level occurred consistently between 2–3 km MSL over the fall seasons of 1998 and 1999. Therefore, as in H01, the maximum reflectivity values observed in our study at 2–3 km MSL also indicate the melting level. The larger frequency of reflectivity in the northern sector is likely due to orographic enhancement of precipitation when moisture flux encounters the main crests of the Alps (Fig. 1).

Analysis of individual storms that occurred during fall 1998 and 1999 revealed that seasonal average reflectivity comprised two types of storms: one type in which stratiform precipitation was enhanced by embedded convection and a second type that only had stratiform precipitation (H01). Individual cases from MAP provide examples of precipitation events with embedded convection (IOP 2b) and events characterized by stratiform structure (IOP 8; MH03; YH03; RH07). The seasonal JJA CFAD for the northern sector presented here (Fig. 2) has similarities to the vertical reflectivity cross section from MAP IOP 2b (MH03). Both show maximum reflectivity (35 dBZ for our JJA CFAD and 35–40 dBZ for IOP 2b) at 3 km MSL. In addition, 15-dBZ reflectivity is observed in the JJA CFAD at 8–10 km MSL and at 7 km MSL in IOP 2b. The resemblance in vertical structure provides an indication that mesoscale conditions and microphysics described for IOP 2b are typical for summertime precipitation systems. Orographic precipitation is strongly affected by atmospheric stability and moisture flux (H01; MH03; RH07; Panziera and Germann 2010). The precipitation structure during IOP 2b resulted from unstable and unblocked flow with a high Froude number. The precipitation is orographically enhanced by the development of embedded convective cells over the first mountain peaks (MH03; YH03; RH07). The similarity in vertical structure between our JJA-type storm classification and IOP 2b implies that the majority of summer precipitation occurs during unstable/unblocked mesoscale conditions. In IOP 2b fallout of rimed ice particles and graupel from convective cells and coalescence of raindrops beneath the melting level produced locally intense precipitation (MH03; YH03; RH07). The increased intensity associated with JJA-type vertical structure suggests that precipitation is enhanced by microphysical processes similar to those observed during IOP 2b. Also, MH03 note that MAP IOP 2b occurred on a day having average temperature of ~19°C. Figure 5 shows the vertical structure of reflectivity associated with the JJA storm type is the most probable vertical structure type at a temperature of 19°C and further supports the relevance of comparison between IOP 2b and JJA-type vertical structure.

In addition, the DJF seasonal CFAD from the north sector resembles the vertical reflectivity profile from MAP IOP 8 (MH03). Both vertical profiles of reflectivity show maximum reflectivity of 30 dBZ at 2 km MSL with 15-dBZ reflectivity limited vertically to the 5-km level in the DJF CFAD and 4 km for IOP 8. Furthermore, MAP IOP 8 occurred during a much cooler day with average temperature ~8°C (MH03). At 8°C the most probable vertical structure type is DJF (Fig. 5). The similarity in vertical profiles between our DJF-type vertical structure and IOP 8 could suggest that blocked flow conditions and associated microphysics described for IOP 8 dominate winter season precipitation systems. In contrast to IOP 2b, IOP 8 was found to be more stratiform as its structure is characterized as a layer of dry snow over wet snow with light rain below 2 km (MH03). Some overturning is present because of shear at the top of the blocked airflow, but the resulting vertical motion is not enough to trigger graupel formation or significant riming. Instead, vapor diffusion onto ice is hypothesized as the primary particle growth mechanism (MH03; YH03; RH07). The reduced precipitation intensity associated with DJF-type vertical structure hints that DJF-type storms may be primarily characterized by stratiform microphysical processes as described for IOP 8.

b. Effect of relationship between surface temperature and vertical reflectivity structure on observed seasonal climate

The relationships between temperature, vertical reflectivity structure, and surface characteristics provide insights concerning observed seasonal precipitation. Vertical structure of reflectivity appears to be related to surface temperature (Fig. 5), and different precipitation characteristics result from different vertical structures of reflectivity (Table 3 and Fig. 6). Rudolph et al. (2011) show that precipitation follows an annual pattern in southern Switzerland with daily precipitation at a minimum in winter, increasing in spring, maximized in summer, and decreasing, although highly variable, in fall. We have shown that JJA-type storms increase in probability with increasing surface temperature, and JJA-type storms have the highest precipitation rates, the shortest duration, and the shortest interval between precipitation events. Therefore, the current study indicates that the summertime maximum in precipitation is the result of more frequent and higher intensity precipitation. This aligns with Molnar and Burlando (2008), who show for southern Switzerland that summer has the shortest and most intermittent storms. Likewise, the annual minimum in daily precipitation experienced during winter stems from less frequent and lower intensity storms. Individual storms having MAM/SON-type vertical structure produce greater total amounts of precipitation than storms with DJF- or JJA-type vertical structure. The large amount of total precipitation produced by MAM/SON-type vertical structure, which becomes most common in mid-spring and fall, likely contributes to October having the maximum monthly precipitation for the area around the Monte Lema radar (Frei and Schär 1998). Rudolph et al. (2011) observed that 10-day average precipitation amounts in the fall season are occasionally similar to 10-day average precipitation amounts experienced during the summer season. This coincides with our current study’s finding that precipitation systems occurring in the warmer days of early fall often have JJA-type vertical structure (Fig. 4b). Then, in mid-fall as daily temperatures cool and MAM/SON-type storms become more likely, individual precipitation systems become less intense, but the average precipitation received from each storm increases (Table 3 and Fig. 4).

c. Future climate implications of relationship between surface temperature and vertical reflectivity structure

The conditional probability of storm type given the surface temperature (Fig. 5) combined with observed precipitation characteristics associated with each storm type (Table 3) may have implications concerning global warming’s effect on precipitation. Synoptic weather patterns and resulting moisture flux directed toward the Alps are also known to affect precipitation characteristics. Specifically, heavy precipitation in the southern Alps has been associated with southerly flow carrying moisture from the Mediterranean toward south-facing Alpine terrain (Doswell et al. 1998; Massacand et al. 1998; Lin et al. 2001; Rudari et al. 2004). Synoptic weather patterns conducive to moist southerly flow also follow seasonal patterns occurring most commonly in spring and fall (Martius et al. 2006; Grazzini 2007) during which the Mediterranean is the primary moisture source for the southern Alps (Sodemann and Zubler 2010). For this study we limit our analysis to the relationship between seasonal precipitation characteristics and surface temperature. This is clearly not the only important factor in consideration of future precipitation as atmospheric moisture, stability, or large-scale circulation may also be affected in a changing climate, so the interplay between surface temperature and these factors provides opportunity for further investigation. Nonetheless, Fig. 5 implies that changes in average daily temperature affect the probability of precipitation having JJA- versus MAM/SOM- versus DJF-type vertical structure. This may become especially relevant during the spring and fall seasons if climate change impacts average daily temperature by earlier onset of warmer temperatures in spring or later occurrence of cooler temperatures in fall. If the conditional probability of storm type given a certain surface temperature remains constant as spring and fall become warmer, longer portions of each season could experience more intense, summer-type storms. The transition to more frequent summer-type storms is similar to results from Beniston (2009) that indicate an increasing portion of European precipitation observed in the twentieth century and modeled for the twenty-first century occurs during “warm modes” (temperature above 75th quantile). In addition, the warm mode precipitation is more likely to have convective precipitation characteristics of short duration and high intensity (Beniston 2009), similar to the precipitation characteristics that accompany JJA-type vertical structure (Table 3).

Previous studies also provide evidence that winter precipitation may be more sensitive to temperature than summer season precipitation. Over the years 1864–2000 the winter season had both the largest increase in seasonal temperatures [0.9°–1.1°C (100 yr)−1 north of the Alps and ~0.6°C (100 yr)−1 south of the Alps] and the largest positive trend in precipitation at several locations across Switzerland, although notably the precipitation trend is 16%–37% (100 yr)−1 in northern Switzerland with no precipitation trend identified in southern Switzerland (Begert et al. 2005). Regional climate models (RCMs) for the twenty-first century also predict an increase in precipitation extremes for Europe during the winter season (Frei et al. 2006). In the context of our study, climate change related to an increase in surface temperature may lead to an increase of average total precipitation in winter because warmer daily temperatures increase the probability of MAM/SON-type vertical structure, which produce higher total amounts of precipitation than DJF-type storms (Table 3).

The increase in precipitation rate that accompanies increased surface temperature is in agreement with many observational and model-based studies. However, the decrease in storm interval is contrary to previous works that find precipitation rate increases under climate change are accompanied by increases in the time interval between storms (Hennessy et al.1997; Dai et al. 1998; Trenberth 1999; Giorgi et al. 2011). Likely, this contradiction arises because of climate effects other than surface temperature on the precipitation characteristics associated with each class of vertical structure. As such, climate change may affect the availability of atmospheric moisture from evaporation and advection because of changes in soil moisture and large-scale circulation. Subsequently, changes in moisture availability may change the distribution of precipitation characteristics for each type of storm (Trenberth et al. 2003). For example, Berg et al. (2009) find observed and modeled precipitation intensity is more dependent on temperature in winter than in summer with the wintertime increase in intensity attributed to increased atmospheric moisture content at higher temperatures because of the Clausius–Clapeyron relationship. However, temperature effects on precipitation intensity are limited in summer by moisture availability, potentially because of drier soil (Berg et al. 2009).

6. Conclusions

Analysis of operational radar data over March 2004–February 2011 from the Monte Lema radar located close to Locarno, Switzerland, indicates that the vertical structure of reflectivity follows a seasonal pattern. Frequency distributions of reflectivity versus altitude (CFADs) reveal that summer precipitation events have the highest vertical extent and greatest reflectivity near the surface as compared to any other season and indicate the increased convective nature of summertime precipitation. Winter CFADs portray a more stratiform character as reflectivity remains closer to ground level and at lower magnitude than the other seasons. CFADs for spring and fall show a vertical extent and magnitude of reflectivity in between the values observed for summer and winter. The consistency in seasonal CFADs allows classification of individual storms by similarity to vertical structure of reflectivity typically observed in summer, winter, or spring/fall.

Rain gauge data for individual precipitation events reveal unique precipitation characteristics for JJA-, MAM/SON-, and DJF-type storms. JJA-type precipitation events are more severe than DJF type (shorter duration and higher maximum precipitation rate) but result in similar total precipitation per event. Therefore, a change in the probability distribution of storm-type classification, such as for a specific day or over an entire season, has implications for precipitation characteristics at the surface. An increase in the probability of JJA-type precipitation events results in a shift toward shorter, more severe storms.

The differences in precipitation characteristics based on different vertical structure of reflectivity have relevance for predicting future precipitation for a changing climate because of the relationship between storm-type classification and surface temperature. This demonstrates the role that an increase in the occurrence of convective precipitation may have on hydrologic intensity as climate changes. In addition, the possibility that convective precipitation will play an increasing role in annual precipitation highlights the importance of accurately portraying convection in climate models, either through improved parameterization or spatial resolution appropriate for convective precipitation. Lastly, we point out that our simplified model of vertical reflectivity structure as a function of surface temperature assumes that seasonal relationships between temperature and Froude number remain constant. This raises questions about the localized impacts of climate change on atmospheric stability, moisture flux, and wind speed and direction as these parameters are especially relevant to blocked or unblocked flow in areas of complex terrain where orography influences precipitation distribution.

Acknowledgments

The authors thank Dr. Urs Germann and Marco Boscacci of MeteoSwiss for providing the operational radar data products. Rain gauge and surface temperature data were downloaded through the web-based interface IDAWEB portal operated by MeteoSwiss (https://gate.meteoswiss.ch/idaweb). The authors also thank two anonymous reviewers whose comments led to improvements in the manuscript and Dr. Balaji Rajagopalan of the University of Colorado at Boulder, Department of Civil, Environmental, and Architectural Engineering for his suggestions regarding the vector general linearized model. This research was supported by NSF Grant AGS-0937035 and the University of Colorado at Boulder, Department of Atmospheric and Oceanic Sciences.

REFERENCES

REFERENCES
Begert
,
M.
,
G.
Seiz
,
T.
Schlegel
,
M.
Musa
,
G.
Baudraz
, and
M.
Moesch
,
2003
: Homogenisierung von Klimamessreihen der Schweiz und Bestimmung der Normwerte 1961–1990. Schlussbericht des Projekts NORM90, MeteoSchweiz Veröffentlichungen Nr. 67, 170 pp.
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.
,
2009
:
Trends in joint quantiles of temperature and precipitation in Europe since 1901 and projected for 2100
.
Geophys. Res. Lett.
,
36
,
L07707
,
doi:10.1029/2008GL037119
.
Berg
,
P.
,
J. O.
Haerter
,
P.
Thejll
,
C.
Piani
,
S.
Hagemann
, and
J. H.
Christensen
,
2009
:
Seasonal characteristics of the relationship between daily precipitation intensity and surface temperature
.
J. Geophys. Res.
,
114
,
D18102
,
doi:10.1029/2009JD012008
.
Bougeault
,
P.
, and
Coauthors
,
2001
:
The MAP special observing period
.
Bull. Amer. Meteor. Soc.
,
82
,
433
462
.
Dai
,
A.
,
K. E.
Trenberth
, and
T. R.
Karl
,
1998
:
Global variations in droughts and wet spells: 1900–1995
.
Geophys. Res. Lett.
,
25
,
3367
3370
.
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
.
Fraedrich
,
K.
, and
C.
Larnder
,
1993
:
Scaling regimes of composite rainfall time series.
Tellus
,
45A
,
289
298
.
Frei
,
C.
, and
C.
Schär
,
1998
:
A precipitation climatology of the Alps from high-resolution rain-gauge observations
.
Int. J. Climatol.
,
18
,
873
900
.
Frei
,
C.
,
R.
Schöll
,
S.
Fukutome
,
J.
Schmidli
, and
P. L.
Vidale
,
2006
:
Future change of precipitation extremes in Europe: Intercomparison of scenarios from regional climate models
.
J. Geophys. Res.
,
111
,
D06105
,
doi:10.1029/2005JD005965
.
Gabella
,
M.
,
U.
Germann
,
M.
Boscacci
,
M.
Sartori
, and
O.
Progin
,
2009
: Planning acceptance tests of the next generation Swiss weather radar network. Extended Abstracts, 34th Conf. on Radar Meteorology, Williamsburg, VA, Amer. Meteor. Soc., P11.3. [Available online at http://ams.confex.com/ams/34Radar/techprogram/paper_154452.htm.]
Germann
,
U.
, and
J.
Joss
,
2004
: Operational measurement of precipitation in mountainous terrain. Weather Radar: Principles and Advanced Applications, P. Meischne, Ed., Springer-Verlag, 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
.
Giorgi
,
F.
,
E. S.
Im
,
E.
Coppola
,
N. S.
Diffenbaugh
,
X. J.
Gao
,
L.
Mariotti
, and
Y.
Shi
,
2011
:
Higher hydroclimatic intensity with global warming
.
J. Climate
,
24
,
5309
5324
.
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.
, and
R. W.
Knight
,
2008
:
Prolonged dry episodes over the conterminous United States: New tendencies emerging during the last 40 years
.
J. Climate
,
21
,
1850
1862
.
Hennessy
,
K. J.
,
J. M.
Gregory
, and
J. F. B.
Mitchell
,
1997
:
Changes in daily precipitation under enhanced greenhouse conditions
.
Climate Dyn.
,
13
,
667
680
.
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
.
Houze
,
R. A.
, Jr.
,
1993
: Cloud Dynamics. Academic Press, 573 pp.
Houze
,
R. A.
, Jr.
,
1997
:
Stratiform precipitation in regions of convection: A meteorological paradox?
Bull. Amer. Meteor. Soc.
,
78
,
2179
2196
.
Houze
,
R. A.
, Jr.
,
C. N.
James
, and
S.
Medina
,
2001
:
Radar observations of precipitation and airflow on the Mediterranean side of the Alps: Autumn 1998 and 1999
.
Quart. J. Roy. Meteor. Soc.
,
127
,
2537
2558
.
Joss
,
J. S.
, and
Coauthors
,
1998
: Operational Use of Radar for Precipitation Measurements in Switzerland. Hochschulverlag AG, 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
.
Langhans
,
W.
,
A.
Gohm
, and
G.
Zängl
,
2011
:
The orographic impact on patterns of embedded convection during the August 2005 Alpine flood
.
Quart. J. Roy. Meteor. Soc.
,
137
,
2092
2105
.
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
.
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
.
Medina
,
S.
, and
R. A.
Houze
Jr.
,
2003
:
Air motions and precipitation growth in Alpine storms
.
Quart. J. Roy. Meteor. Soc.
,
129
,
345
371
.
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
.
Molnar
,
P.
, and
P.
Burlando
,
2008
:
Variability in the scale properties of high-resolution precipitation data in the Alpine climate of Switzerland
.
Water Resour. Res.
,
44
, W10404, doi:10.1029/2007WR006142.
Overeem
,
A.
,
I.
Holleman
, and
A.
Buishand
,
2009
:
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
.
Perl
,
M.
,
C.
Haeberli
,
E.
Grueter
,
M.
Musa
,
U.
Kueng
, and
M.
Abbt
, cited
2009
: Werkzeug für die Qualitaetskontrolle und die Bearbeitung von meteorologischen Daten. [Available online at http://www.meteoswiss.ch/web/en/services/data_portal/data_quality.html.]
Rotunno
,
R.
, and
R. A.
Houze
,
2007
:
Lessons on orographic precipitation from Mesoscale Alpine Programme
.
Quart. J. Roy. Meteor. Soc.
,
133
,
811
830
.
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
.
Rudolph
,
J. V.
,
K.
Friedrich
, and
U.
Germann
,
2011
:
Relationship between radar-estimated precipitation and synoptic weather patterns in the European Alps
.
J. Appl. Meteor. Climatol.
,
50
,
944
957
.
Rudolph
,
J. V.
,
K.
Friedrich
, and
U.
Germann
,
2012
:
Model-based estimation of dynamic effect on twenty-first-century precipitation for Swiss river basins
.
J. Climate
, 25, 2897–2913.
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
.
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
.
Solomon
,
S.
,
D.
Qin
,
M.
Manning
,
M.
Marquis
,
K.
Averyt
,
M. M. B.
Tignor
,
H. L.
Miller
Jr.
, and
Z.
Chen
, Eds.,
2007
: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.
Steinacker
,
R.
,
W.
Pöttschacher
, and
M.
Dorninger
,
1997
:
Enhanced resolution analysis of the atmosphere over the Alps using the fingerprint technique
.
Ann. Meteor.
,
35
,
235
237
.
Steinacker
,
R.
,
C.
Häberli
, and
W.
Pöttschacher
,
2000
:
A transparent method for the analysis and quality evaluation of irregularly distributed and noisy observational data
.
Mon. Wea. Rev.
,
128
,
2303
2316
.
Trenberth
,
K. E.
,
1999
:
Conceptual framework for changes of extremes of the hydrological cycle with climate change
.
Climatic Change
,
42
,
327
339
.
Trenberth
,
K. E.
,
A.
Dai
,
R. M.
Rasmussen
, and
D. B.
Parsons
,
2003
:
The changing character of precipitation
.
Bull. Amer. Meteor. Soc.
,
84
,
1205
1217
.
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
.
WMO
,
1983
: Guide to climatological practices. WMO-No. 100, 2nd ed., 198 pp. [The current edition is available online at http://www.wmo.int/pages/prog/wcp/ccl/documents/WMO_100_en.pdf.]
Yee
,
T. W.
,
2008
: The VGAM package. R News, Vol. 8, No. 2, R Foundation for Statistical Computing, Hamilton, ON, Canada, 28–39.
Yee
,
T. W.
, and
T. J.
Hastie
,
2003
:
Reduced-rank vector generalized linear models
.
Stat. Modell.
,
3
,
15
41
.
Yuter
,
S. E.
, and
R. A.
Houze
Jr.
,
1995
:
Three-dimensional kinematic and microphysical evolution of Florida cumulonimbus. Part II: Frequency distributions of vertical velocity, reflectivity, and differential reflectivity
.
Mon. Wea. Rev.
,
123
,
1941
1963
.
Yuter
,
S. E.
, and
R. A.
Houze
Jr.
,
2003
:
Microphysical modes of precipitation growth determined by S-band vertically pointing radar in orographic precipitation during MAP
.
Quart. J. Roy. Meteor. Soc.
,
129
,
455
476
.
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.