• Burger, A. E., , and E. Ekhart, 1937: Über die tageszeitliche Zirkulation der Atmosphäre im Bereich der Alpen (On the daily circulation of the atmosphere over the Alps). Gerlands Beitr. Geophys., 49 , 341367.

    • Search Google Scholar
    • Export Citation
  • Chapman, S., , and R. S. Lindzen, 1970: Atmospheric Tides. D. Reidel, 201 pp.

  • Crook, N. A., , T. L. Clarke, , and M. W. Moncrieff, 1990: The Denver Cyclone. Part I: Generation in low Froude number flow. J. Atmos. Sci., 47 , 27252742.

    • Search Google Scholar
    • Export Citation
  • Crook, N. A., , T. L. Clarke, , and M. W. Moncrieff, 1991: The Denver Cyclone. Part II: Interaction with the convective boundary layer. J. Atmos. Sci., 48 , 21092126.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., 1989: Numerical study of convection observed during the Winter Monsoon Experiment using a mesoscale two-dimensional model. J. Atmos. Sci., 46 , 30773107.

    • Search Google Scholar
    • Export Citation
  • Egger, J., , S. Bajarachaya, , U. Egger, , R. Heinrich, , J. Reuder, , P. Shakaya, , H. Wendt, , and V. Wirth, 2000: Diurnal winds in the Himalayan Kali Gandaki Valley. Part I: Observations. Mon. Wea. Rev., 128 , 11061122.

    • Search Google Scholar
    • Export Citation
  • Egger, J., and Coauthors, 2002: Diurnal winds in the Himalayan Kali Gandaki Valley. Part III: Remotely piloted aircraft soundings. Mon. Wea. Rev., 130 , 20422058.

    • Search Google Scholar
    • Export Citation
  • Egger, J., and Coauthors, 2005: Diurnal circulation of the Bolivian Altiplano. Part I: Observations. Mon. Wea. Rev., 133 , 911924.

  • Ekhart, E., 1944: Strömungsaerologie der Talwinde (Aerology of valley winds). Meteor. Z., 61 , 217231.

  • Freytag, C., 1985: MERKUR—Results: Aspects of the temperature fields and the energy budget in a large alpine valley during mountain and valley wind. Contrib. Atmos. Phys., 58 , 458476.

    • Search Google Scholar
    • Export Citation
  • Freytag, C., , and B. Hennemuth, 1983: MERKUR: Mesoskaliges Experiment im Raum Kufstein-Rosenheim (Mesoscale experiment in the region of Kufstein-Rosenheim). Wiss. Mitt. Meteor. Inst. Univ. München, 48, 132 pp.

    • Search Google Scholar
    • Export Citation
  • Grell, G. A., , J. Dudhia, , and D. R. Stauffer, 1995: A description of the fifth-generation Penn State/NCAR Mesoscale Model (MM5). NCAR Tech. Note NCAR/TN-398+STR, 122 pp.

    • Search Google Scholar
    • Export Citation
  • Hafner, T. A., , M. E. Reinhardt, , E. L. Weisel, , and H. P. Fimpel, 1987: Boundary layer aspects and elevated heat source effects of the Alps. Meteor. Atmos. Phys., 36 , 6173.

    • Search Google Scholar
    • Export Citation
  • Haurwitz, B., , and A. D. Cowley, 1973: The diurnal and semi-diurnal barometric oscillations: Global distribution and annual variation. Pure. Appl. Geophys., 102 , 193222.

    • Search Google Scholar
    • Export Citation
  • Heimann, D., 1997: Mesoscale surface-wind characteristics and potential gravity wave formation during cross-alpine airflow. Meteor. Atmos. Phys., 62 , 4970.

    • Search Google Scholar
    • Export Citation
  • Hoinka, K. P., 1980: Synoptic-scale atmospheric features and foehn. Beitr. Phys. Atmos., 53 , 486508.

  • Kain, J. S., 2004: The Kain–Fritsch convective parameterization: An update. J. Appl. Meteor., 43 , 170181.

  • Klemp, J. B., , and D. R. Durran, 1983: An upper boundary condition permitting internal gravity wave radiation in numerical mesoscale models. Mon. Wea. Rev., 111 , 430444.

    • Search Google Scholar
    • Export Citation
  • Lugauer, M., , and P. Winkler, 2005: Thermal circulation in south Bavaria—Climatology and synoptic aspects. Meteor. Z., 14 , 1430.

  • Mlawer, E. J., , S. J. Taubman, , P. D. Brown, , M. J. Iacono, , and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmosphere: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102 , 1666316682.

    • Search Google Scholar
    • Export Citation
  • Müller, H., , and C. D. Whiteman, 1988: Breakup of a nocturnal temperature inversion in the Dischma Valley during DISKUS. J. Appl. Meteor., 27 , 188194.

    • Search Google Scholar
    • Export Citation
  • Müller, H., , R. Reiter, , and R. Sládkovič, 1984: Die vertikale Windstruktur beim MERKUR-Schwerpunkt ‘Tagesperiodische Windsysteme’ aufgrund von aerologischen Messungen im Inntal und im Rosenheimer Becken (On the vertical wind structure of the diurnal wind system within the Inn Valley and the adjacent plain: Results of aerological soundings during the field experiment MERKUR). Arch. Meteor. Geophys. Bioklimatol., B33 , 359372.

    • Search Google Scholar
    • Export Citation
  • Pamperin, H., , and G. Stilke, 1985: Nächtliche Grenzschicht und LLJ im Alpenvorland nahe dem Inntalausgang (Nocturnal boundary layer and LLJ in the pre-alpine region near the Inn Valley exit). Meteor. Rundsch., 38 , 145156.

    • Search Google Scholar
    • Export Citation
  • Reisner, J., , R. M. Rasmussen, , and R. T. Bruintjes, 1998: Explicit forecasting of supercooled liquid water in winter storms using the MM5 mesoscale model. Quart. J. Roy. Meteor. Soc., 124 , 10711107.

    • Search Google Scholar
    • Export Citation
  • Sládkovič, R., , and H-J. Kanter, 1977: Low-level jet in the Bavarian pre-Alpine region. Arch. Meteor. Geophys. Bioklimatol., A25 , 343355.

    • Search Google Scholar
    • Export Citation
  • Smith, R. B., 1979: The influence of mountains on the atmosphere. Adv. Geophys., 21 , 87230.

  • Steinacker, R., 1984: Area height distribution of a valley and its relation to the valley wind. Contrib. Atmos. Phys., 57 , 6471.

  • Szoke, E. J., 1991: Eye of the Denver Cyclone. Mon. Wea. Rev., 119 , 12831292.

  • Szoke, E. J., , M. L. Weisman, , J. M. Brown, , F. Caracena, , and T. W. Schlatter, 1984: A synoptic analysis of the Denver tornadoes of 3 June 1981. Mon. Wea. Rev., 112 , 790808.

    • Search Google Scholar
    • Export Citation
  • Thompson, G., , R. M. Rasmussen, , and K. Manning, 2004: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part I: Description and sensitivity analysis. Mon. Wea. Rev., 132 , 519542.

    • Search Google Scholar
    • Export Citation
  • Wagner, A., 1932: Der tägliche Luftdruck- und Temperaturgang in der freien Atmosphäre und in Gebirgstälern (Diurnal variations of pressure and temperature in the free atmosphere and in valleys). Gerlands Beitr. Geophys., 37 , 315344.

    • Search Google Scholar
    • Export Citation
  • Weissmann, M., , F. J. Braun, , L. Gantner, , G. J. Mayr, , S. Rahm, , and O. Reitebuch, 2005: The Alpine mountain–plain circulation: Airborne Doppler lidar measurements and numerical simulations. Mon. Wea. Rev., 133 , 30953109.

    • Search Google Scholar
    • Export Citation
  • Whiteman, C. D., 1990: Observations of thermally developed wind systems in mountainous terrain. Atmospheric Processes over Complex Terrain, Meteor. Monogr., No. 45, Amer. Meteor. Soc., 5–42.

    • Search Google Scholar
    • Export Citation
  • Whiteman, C. D., , and J. C. Doran, 1993: The relationship between overlying synoptic-scale flows and winds within a valley. J. Appl. Meteor., 32 , 16691682.

    • Search Google Scholar
    • Export Citation
  • Wippermann, F., , and G. Gross, 1981: On the construction of orographically influenced wind roses for given distributions of the large-scale wind. Beitr. Phys. Atmos., 54 (4) 492501.

    • Search Google Scholar
    • Export Citation
  • Zängl, G., 2002a: An improved method for computing horizontal diffusion in a sigma-coordinate model and its application to simulations over mountainous topography. Mon. Wea. Rev., 130 , 14231432.

    • Search Google Scholar
    • Export Citation
  • Zängl, G., 2002b: Stratified flow over a mountain with a gap. Linear theory and numerical simulations. Quart. J. Roy. Meteor. Soc., 128 , 927949.

    • Search Google Scholar
    • Export Citation
  • Zängl, G., 2003: A generalized sigma-coordinate system for the MM5. Mon. Wea. Rev., 131 , 28752884.

  • Zängl, G., 2004: A reexamination of the valley wind system in the Alpine Inn Valley with numerical simulations. Meteor. Atmos. Phys., 87 , 241256.

    • Search Google Scholar
    • Export Citation
  • Zängl, G., 2007: Small-scale variability of orographic precipitation in the Alps: Case studies and semi-idealized numerical simulations. Quart. J. Roy. Meteor. Soc., 133 , 17011716.

    • Search Google Scholar
    • Export Citation
  • Zhang, D., , and R. A. Anthes, 1982: A high-resolution model of the planetary boundary layer—Sensitivity tests and comparisons with SESAME-79 data. J. Appl. Meteor., 21 , 15941609.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Schematic illustration of different valley exits with possible flow features. (a) Daytime up-valley flow and (b) nocturnal jetlike down-valley outflow for a symmetric valley exit, respectively. (c) Possible up- and down-valley flow for an asymmetric valley exit. (d) Impact of a background flow on the thermally driven circulation in an asymmetric valley exit area.

  • View in gallery

    Topographic map of the Ostallgäu and the Lech Valley. Letters indicate AWS positions [V1 (Holzgau, 1093 m), V2 (Stanzach, 940 m), V3 (Reutte, 841 m), P1 (Schwangau, 792 m), P2 (Buching, 796 m), P3 (Steingaden, 786), and P4 (Lengenwang, 811 m)]. The pertinent mountain ranges are also indicated, as well as the Vilser Mountains (VM, highlighted by the dashed circle) and the Halblech Valley (HV). The line indicates the position of the cross section in Fig. 12.

  • View in gallery

    Scatterplot of 3-hourly mean wind for V2 vs P2, morning (0430–0730 CEST; square) and afternoon (1530–1830 CEST; circle). Filled symbols indicate that the daily mean global radiation exceeded 10 MJ m−2. A threshold for wind speed >0.5 m s−1 was applied. The horizontal and vertical lines indicate the upward (solid) and downward (dashed) direction of the valley axis. The diagonal is indicative for identical values at V2 and P2 and the dotted diagonals are the 45° deviation from it.

  • View in gallery

    MSLP with contour interval (CI) of 2 hPa (solid) and geopotential (CI = 20 gpm) at 500 hPa (dashed) for 1200 UTC 28 Jul. Topography shaded in increments of 500 m (white for sea level). The X marks the area of the field experiment.

  • View in gallery

    Time series of wind speed (solid lines), υ (m s−1), and direction (°, crosses) for selected AWS stations—(a) P3, (b) P2, (c) P1, and (d) V2—for 28 Jul.

  • View in gallery

    Potential temperature θ and specific humidity q from descents of Kali and wind speed, υ, and direction, dir, from pilot balloon ascents during 28 Jul near P2. Different times during the day are indicated by different line styles or symbols.

  • View in gallery

    (top) Time series of surface pressure at P1 and (bottom) surface pressure gradients between selected AWSs (P3, P2, V2) and P1 for 28 Jul.

  • View in gallery

    As in Fig. 4, but for 1200 UTC 17 Aug.

  • View in gallery

    As in Fig. 5, but for 17 Aug.

  • View in gallery

    As in Fig. 7, but for 17 Aug.

  • View in gallery

    Wind vectors at ∼180 m AGL, with shading indicating total wind speed, for (a) 0800 and (b) 1600 CEST 28 Jul. Topography of the model domain is contoured; CI = 200 m.

  • View in gallery

    Vertical cross section along the line marked in Fig. 2 showing contours θ (K) and the projected wind vectors and wind speed (shaded) onto the cross section. Both panels are valid for 1600 CEST (a) 28 Jul and (b) 17 Aug.

  • View in gallery

    Pressure perturbation (shaded) and wind vectors at 1200 m above sea level at 1600 CEST for (a) 28 Jul and (b) 17 Aug. Topography of the model domain is contoured; CI = 400 m.

  • View in gallery

    Time series of surface pressure gradients from MM5 model simulations for (a) 28 Jul and (b) 17 Aug. All stations (P3, P2, V2) were reduced to the height of P1.

  • View in gallery

    As in Fig. 11, but for (a) 0800, (b) 1100, and (c) 1600 CEST 17 Aug.

  • View in gallery

    Wind vectors at ∼180 m AGL, shading indicating total wind speed, for idealized setup with easterly flow (a) including real topography and (b) without the Ammer Mountains, both at 1100 CEST. Topography of the model domain is contoured; CI = 200 m.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 4 4 2
PDF Downloads 1 1 0

Thermally Driven Flows at an Asymmetric Valley Exit: Observations and Model Studies at the Lech Valley Exit

View More View Less
  • 1 Institute for Atmospheric and Climate Science, ETH Zurich, Zurich, Switzerland
  • 2 Institute for Geophysics and Meteorology, University of Cologne, Cologne, Germany
  • 3 Meteorological Institute Munich, University of Munich, Munich, Germany
© Get Permissions
Full access

Abstract

The summertime thermal circulation in the region of an asymmetric valley exit is investigated by means of observations and high-resolution model simulations. The northeastward-oriented Alpine Lech Valley opening into the Bavarian Alpine foreland has an eastern slope exceeding the western slope by about 15 km. Northerly winds along the eastern slope are frequently observed, reaching substantial strength during fair weather conditions. A field experiment has been conducted to explore this phenomenon and to pinpoint the connection of the northeasterly flow to the Lech Valley wind circulation. Numerical simulations have also been carried out to support the interpretation of the observations. It is found that the northerlies owe their existence to the dominantly easterly flow along the foothills of the Alps, which is partly induced by the Alpine heat low but may be strengthened by favorable synoptic conditions. Examples for both situations will be discussed. The diurnal flow in the Lech Valley has little obvious impact on these northeasterlies. On days with moderate synoptic easterly flow, a wake is present on the lee of the eastern slope of the exit region, accompanied by a shear zone along the edge of the wake. This shear zone is forced southward during the daytime because of thermally initiated pressure gradients between the Alpine foreland and the Alps, leading to sudden wind changes in the exit area at the time of its passage.

* Current affiliation: Deutscher Wetterdienst, Offenbach, Germany.

Corresponding author address: Thomas Spengler, Institute for Atmospheric and Climate Science, ETH Zurich, Universitätsstrasse 16, ETH Zentrum, CH-8092 Zürich, Switzerland. Email: thomas.spengler@env.ethz.ch

Abstract

The summertime thermal circulation in the region of an asymmetric valley exit is investigated by means of observations and high-resolution model simulations. The northeastward-oriented Alpine Lech Valley opening into the Bavarian Alpine foreland has an eastern slope exceeding the western slope by about 15 km. Northerly winds along the eastern slope are frequently observed, reaching substantial strength during fair weather conditions. A field experiment has been conducted to explore this phenomenon and to pinpoint the connection of the northeasterly flow to the Lech Valley wind circulation. Numerical simulations have also been carried out to support the interpretation of the observations. It is found that the northerlies owe their existence to the dominantly easterly flow along the foothills of the Alps, which is partly induced by the Alpine heat low but may be strengthened by favorable synoptic conditions. Examples for both situations will be discussed. The diurnal flow in the Lech Valley has little obvious impact on these northeasterlies. On days with moderate synoptic easterly flow, a wake is present on the lee of the eastern slope of the exit region, accompanied by a shear zone along the edge of the wake. This shear zone is forced southward during the daytime because of thermally initiated pressure gradients between the Alpine foreland and the Alps, leading to sudden wind changes in the exit area at the time of its passage.

* Current affiliation: Deutscher Wetterdienst, Offenbach, Germany.

Corresponding author address: Thomas Spengler, Institute for Atmospheric and Climate Science, ETH Zurich, Universitätsstrasse 16, ETH Zentrum, CH-8092 Zürich, Switzerland. Email: thomas.spengler@env.ethz.ch

1. Introduction

Thermally driven wind circulations can have a significant impact on the diurnal flow evolution in the vicinity of valley exit regions. At the Inn Valley exit, a day–night asymmetry of the flow was identified during the Mesoscale Experiment in the Region of Kufstein-Rosenheim (MERKUR; Freytag and Hennemuth 1983; Müller et al. 1984; Freytag 1985; Pamperin and Stilke 1985). Although the daytime up-valley flow was found to be rather weak in the Inn Valley exit region, strong nocturnal low-level jets (10–14 m s−1) extending some tens of kilometers into the foreland were observed during clear nights. Similar diurnal flow evolutions were also found at the Loisach (Sládkovič and Kanter 1977) and Salzach (Ekhart 1944) Valley exits. The nocturnal low-level jet can generally be linked to hydraulic theory applied at a constriction at the valley exit (Zängl 2004).

Despite the often very complex and different nature of these valleys, they have one feature in common: a symmetric opening at the valley exit (i.e., the bounding slopes of the valley terminate at the same location; see Figs. 1a,b). Henceforth we will refer to valley exits as being asymmetric when one bounding slope is longer than the other at the valley exit, as shown in Figs. 1c,d.

The daytime and nocturnal flow conditions for a symmetric setup are shown in Figs. 1a,b. In the exit region weak winds are expected during the daytime (Fig. 1a), whereas a strong jetlike outflow during the night reaches a significant distance downstream from the valley exit into the foreland (Fig. 1b).

Although many examples of symmetric exits have been discussed in the literature, there has not been much focus on the effects of asymmetry of valley exits on the diurnal flow evolution. Figures 1c and 1d illustrate two scenarios for an asymmetric valley exit region. Figure 1c illustrates the possible daytime (nighttime) flow setup with up-valley (down-valley) flow conditions. The question mark highlights the issue of how the asymmetry might affect the daytime (nighttime) inflow (outflow) in the foreland region—that is, whether the mass inflow (outflow) for the up-valley (down-valley) winds reaches farther into the foreland compared to a symmetric valley exit. Finally, Fig. 1d depicts the open question of how an ambient flow in the foreland might change the inflow (outflow) in the vicinity of an asymmetric valley exit. There have been some studies on the impact of large-scale flow over valleys on the valley flow but none on the flow near valley exits.

Wippermann and Gross (1981) showed that the pressure gradient related to the ambient flow can lead to a channeling of the flow in wide valleys such as the Rhine Valley. Whiteman and Doran (1993) presented similar channeled flow behavior for the Tennessee Valley in the United States. The Denver Cyclone (Szoke 1991), or Denver convergence vorticity zone (Szoke et al. 1984), is an example of flow over and around a promontory, a situation which might be of relevance with respect to Fig. 1d. Crook et al. (1990) showed that a low Froude number flow setup yields flow around the topography, which generates a vortex in the lee of the Palmer Divide. The vortex subsequently moves downstream, leaving behind a wake of stagnant air. The effects of diabatic heating were included in Crook et al. (1991), showing strong interactions of shear and thermal instabilities along the convergence vorticity zone.

In the European Alps there are several large valleys with asymmetric openings into the Alpine foreland (e.g., the Rhine Valley at Lake Constance and the Rhone Valley at Lake Geneva). Another asymmetric Alpine valley exit is that from the Lech River into the Ostallgäu, in southwestern Bavaria (Fig. 2). The Allgäu Field Experiment (AllgEx) was conducted in the latter area to explore the impact of the geographic asymmetry on the flow evolution near the exit region.

a. Geographic setting

The rather flat Ostallgäu area is bounded by the Ammer Mountains to the southeast and by the Allgäuer Alps to the southwest (Fig. 2). The Lech River Valley is bound by the Allgäuer Alps to the northwest and the Lechtaler Alps to the south. The highest adjacent mountains are around 3000 m. The valley slopes from approximately the southwest to the northeast with a varying bottom width between about 300 m near Holzgau (V1; 1093 m) and 4–5 km in the basin of Reutte (V3; 838 m). The valley has several tributaries mostly directed toward the south into the Lechtaler Alps. Between Reutte (V3) and Schwangau (P1) the Lech River Valley is bound by the Vilser Mountains (VM), which rise about 400 m above the valley floor, creating an obstacle to atmospheric flow between the valley and the foreland. The asymmetry of the exit region is manifested by the distinct protrusions of the ranges embedding the Lech Valley exit area, where the northeastern flank of the valley exit (Ammer Mountains) extends a significant distance into the Alpine foreland, producing a barrier in ambient easterly flow. The major measuring site was located at Buching (P2), which is located northwest of the nearby Halblech Valley (HV). There is also a small ridge of about 100 m height between Buching (P2) and Steingaden (P3).

b. Summer flow conditions in the Ostallgäu

During the summer months between April and September, a thermally driven daytime northeasterly flow near the ground is regularly observed in the Ostallgäu. The onset of the northeasterly flow is usually characterized by a continuous change from almost calm conditions in the morning to northeasterlies peaking around 5–8 m s−1 in the afternoon. This feature is well known by the local residents and is often used by paragliders and hang gliders for dynamic soaring in the area. These winds are most likely related to the larger-scale cyclonic flow around the Alps in summer due to the Alpine heat low (Burger and Ekhart 1937; Hafner et al. 1987). Weissmann et al. (2005) pinpointed that a significant transport of air mass from the Alpine foreland into the Alps is driven by this thermal circulation with maximum northeasterly flow in the late afternoon and evening.

During the nighttime, rather calm conditions dominate the Ostallgäu. Strong nocturnal jetlike outflow, as in the aforementioned Inn Valley, is usually not observed near the surface away from the Lech Valley exit.

A remarkable feature of the daytime northeasterly flow in the area around Buching is its aforementioned strength since it is still located in the Alpine foreland. The nearby Lech Valley circulation might be one driver of the flow, but theory would suggest that the strongest winds during the daytime occur in the valley itself.

In addition, and in contrast to the daytime evolution described above, abrupt onset of northeasterly flow is also observed in Buching on certain days during summer. This can be a hazard to the local pilots because wind speeds sometimes exceed the maximum flight velocity of paragliders, implying that they have to land while flying backward. A preliminary analysis carried out together with the local paragliding school (DAeC Gleitschirmschule) showed that events of strong northeasterly flow in the area are often accompanied by the approach of a low with its frontal system from the northwest. This synoptic setup usually has a tendency toward foehn conditions in southern Bavaria, which are often accompanied by northeasterly flow in the Bavarian Alpine foreland (Hoinka 1980; Heimann 1997). This northeasterly flow, however, is not observed in the area around Buching until about noon.

The primary goals of AllgEx were to identify the role of the Lech Valley in driving the afternoon northeasterly flow in the Ostallgäu area and to explore the dynamics of the sudden onset of northeasterly flow on certain days during the summer months. These two questions strongly relate to the question marks highlighted in Figs. 1c and 1d.

This paper is organized as follows: The instrumentation of the field campaign is described in section 2. Selected surface observations and soundings from the area of investigation are presented in section 3, followed by section 4 with numerical simulations. Discussion of the results and concluding remarks are presented in section 5.

2. Instrumentation

A variety of observing systems were in use during AllgEx, including a network of seven surface automatic weather stations (AWSs) with continuous measurements during the general observation period (GOP) from 1 July 2005 until 25 August 2005, as well as pilot balloons and remotely piloted vehicles equipped with sensor units, which were used during the 17 days of intensive observation periods (IOPs; see Table 1). IOPs were carried out on subjectively selected days when the thermal circulation was expected to develop during the daytime; in particular, we looked for no frontal activities in the area of interest, rather weak upper level flow, and low cloudiness (i.e., unimpeded insolation).

a. Automatic weather stations

AWSs were deployed by the Meteorological Institute Munich (MIM) (Fig. 2). The stations Steingaden (P3), Buching (P2), Schwangau (P1), Reutte (V3), Stanzach (V2), and Holzgau (V1) are approximately aligned along a line from northeast to southwest, stretching from the Bavarian Alpine foreland into the upper Lech Valley in Austria. An additional station was created in Lengenwang (P4), which is located about 20 km west of P3, and is primarily utilized as a reference station for the ambient flow to the north of the area of interest. Each of these AWSs is equipped with sensors for temperature, humidity, wind speed and direction, and pressure (resolution 0.1 hPa, accuracy ±0.3 hPa). Two-minute mean values were sampled at a height of 2 m, whereas pressure was measured near the ground.

At the beginning of the campaign we determined the difference of each station barometer to a reference barometer (dp0). However, this was not accurate enough to adjust for “correct” absolute pressure differences between different stations. To obtain pressure measurements of maximum accuracy, we used the following procedure: Assuming a linear vertical temperature profile between the stations, the pressure of each station was reduced to the height of P1 (pi) and averaged over the entire GOP (pi). It should be noted that other temperature dependencies (e.g., mean temperature) yield negligible differences in pressure with the given height differences. Using the mean value (pi), we determined the mean difference to P1 (dpi = ) for each station. We assume that the pressure gradients relative to P1 average out over the GOP and correct for the difference of the P1 barometer to the reference barometer (dp0), yielding the “exact” pressure for every station: pi(t) = pi(t) − dp0. One should be aware that the assumption that pressure gradients average out over the GOP is violated if a mean pressure gradient between different stations exists. However, we did not find systematic deviations in the mean values and thus argue that the proposed procedure is an appropriate approach to obtain a consistent dataset.

For a dynamical interpretation of the wind evolution, we calculated the 20-min running mean distance-weighted pressure differences between different stations relative to P1 [(pipP1)/dL, where dL, is the distance between the respective stations]. We will refer to these weighted differences as pressure gradients.

Further access to data from mountain AWSs in the vicinity was provided by the local avalanche services in Bavaria and Tyrol (10-min-mean data of temperature, humidity, wind speed and direction, and global radiation). The station located at Tegelberg (TB in Fig. 2) on a mountaintop south of P1 was closest to the area of interest.

b. Pilot balloon soundings with theodolite

Two theodolites tracking a helium-filled balloon were located 2 km southwest of P2 with a basis length and orientation of 1561 m and 117°, respectively. The orientation of the basis was chosen to be perpendicular to the main wind direction in the target area. The measuring system is identical to that described in Egger et al. (2000). During the IOPs, 149 ascents were carried out, aiming for an ascent every hour during the daytime. The vertical resolution is ∼20 m and, depending on cloudiness and visibility, profiles were usually taken up to 2–4 km above ground level (AGL).

c. Remotely piloted vehicles (Kali)

Battery-powered miniature airplanes (1.29 m in length with a wingspan of 2.10 m) collected vertical profiles of pressure, temperature, and humidity. The observing system is named Kali and is described in further detail in Egger et al. (2002, 2005). During the IOP, 97 flights were launched, always aiming for parallel profiling with the pilot balloon. An average height of about 1000 m (maximum ≈2000 m) AGL was reached with a vertical resolution of ∼7 m. The use of an aircraft in a highly populated area such as Germany is subject to several rules. However, the AllgEx research area was far away from restriction zones of civil and military air fields and thus no explicit permissions were necessary.

3. Results

During the 56 days of the GOP we identified three different thermal flow regimes in the area of interest. Days with a flat synoptic pressure distribution and daily total global radiation exceeding 10 MJ m−2 (total daily insolation end of July at 48°N is ∼35 MJ m−2 at the top of the atmosphere) are defined as days with a thermal circulation if the diurnal circulation is subjectively identified in the daily wind cycle. In the first regime, a thermally driven circulation is evident in the Lech Valley and the foreland (25 days). In the second, only the Lech Valley experiences a thermal circulation (7 days), and in the third, no thermal circulation developed during the day (24 days). During the observation period, the circulation in the foreland never occurred without the Lech Valley circulation being evident as well. Whereas the first regime is found on days with weak synoptic-scale forcing, significant large-scale pressure gradients favor the third regime, where the wind at the experimental site was dominated by the large-scale synoptic flow.

To pinpoint some differences between the characteristics of the flow behavior in the Lech Valley and the foreland, we briefly discuss the 3-hourly mean wind direction in the morning and afternoon at V2 and P2 (Fig. 3). In the morning [0430–0730 central European summer time (CEST, which will be used throughout and which is UTC + 2 h); squares], a channeling along the Lech Valley axis (dashed line) is clearly evident at V2, whereas for P2 the direction is much less constrained, indicating that the flow at P2 is not influenced by the Lech Valley outflow at this time. During the afternoon (1530–1830 CEST; circles) the flow is again channeled along the Lech Valley axis (solid line) at V2. At P2 two different regimes are evident during the afternoon, favoring either southerly or northerly flow. The latter direction corresponds to the wind direction of the thermal circulation at P2.

The data points can be further distinguished by separating them into dates on which the daily mean global radiation is above (filled symbols) or below (blank symbols) 10 MJ m−2. For values above this threshold, there are more days with up-valley flow at V2 and concomitant northerly flow at P2 in the afternoon, indicative of a daytime thermal circulation in the foreland. There are a few outliers in the top left corner (solid circles), which can mainly be related to south foehn events and days with strong synoptic westerly flow with reduced cloudiness. Southerly (during foehn) and westerly flow are able to penetrate down to the valley floor because of the south-southwest orientation of the Lech Valley.

In the following we focus on the first regime where the thermal circulation is also evident in the Alpine foreland. The selected behavior is depicted by two cases, 28 July and 17 August. Whereas the synoptic forcing is almost absent for the first case, weak large-scale easterly flow is present in the second case. The rapid onset of northeasterly flow at P2 mentioned in the introduction was observed once during the GOP (17 August) but fell outside an IOP.

a. 28 July

The flow evolution on 28 July appeared to be typical (according to the obtained dataset) for conditions with weak background flow. Rather weak mean sea level pressure (MSLP) gradients are present in the area of interest with a low pressure system southwest of Great Britain, and at 500 hPa the geopotential field implies a mild southwesterly flow across the Alps (Fig. 4).

A down-valley wind is evident at V2 in the early morning with wind speeds of 2 m s−1 (Fig. 5). Southerly wind (≈4 m s−1) at P1 indicates some influence from the Lech Valley outflow, whereas almost no wind was recorded at P2 and southeasterly winds around 1–2 m s−1 were evident at P3. Around 0830 CEST sudden changes at P1, P2, and P3 mark the breakup of the surface-based temperature inversion. Northerly wind speed at P2 increases with time after 1000 CEST, marking the onset of the thermal circulation in the Alpine foreland, which lasts until 1900 CEST (at P2). It should be noted that the flow at P4 is easterly, commencing at 1030 CEST and lasting until 2030 CEST (not shown). Maximum surface wind speeds at P2 (∼4 m s−1) were sustained between 1400 and 1800 CEST, with the evolution at P1 being similar but with maximum wind speeds 1–2 m s−1 lower. The evolution at P3 is less clear because of strong variations in wind speed and direction.

V2 shows the clear signature of a diurnal valley wind circulation with the aforementioned down-valley winds decreasing from 1000 to 1200 CEST before the wind direction changes to 0°, indicating up-valley flow. The onset of the up-valley wind is rather abrupt around 1300 CEST with wind speeds reaching 4 m s−1 between 1400 and 1800 CEST.

Inspecting the potential temperature profile, a shift toward higher values is evident over the course of the day indicating warm air advection, as also shown by a clockwise turning (with height) of the wind direction (Fig. 6). Around 1000 CEST a temperature inversion around 400 m AGL is evident, accompanied by weak wind speed and varying wind directions in the lower 1000 m. At 1148 CEST, the mixed layer depth (defined by constant potential temperature with height) reaches 600 m along with increased wind speed up to 4 m s−1 over the same vertical extent at a fixed wind direction of 45°. The low-level wind speed peaks around 1340 CEST at 8 m s−1 at a height of 200 m. It is interesting to note that the depth of the layer with constant wind direction (750 m) remains constant after 1208 CEST. This vertical extent of northeasterly flow appeared to be typical for days with a thermal circulation near P2 and corresponds to the height of the nearby mountains to the south. The wind above 1000 m AGL is from southerly directions, implying a counterflow setup in comparison to the lower levels, and increases from 0943 to 1208 CEST but decreases thereafter. It is interesting to note that the pilot balloons commonly recrossed the baseline of the theodolites some 20 min after release at about 2500 m AGL if the thermal circulation was present at P2.

The surface pressure time series at P1 (Fig. 7a) has a diurnal cycle with a peak-to-peak amplitude of about 1.8 hPa, with a maximum in the morning hours and a minimum in the evening. This diurnal cycle is typical for thermally driven circulations like sea breezes and valley winds. However, one should keep in mind the atmospheric pressure tide, which has a diurnal contribution as well (Chapman and Lindzen 1970). According to the empirical formula of Haurwitz and Cowley (1973), one would expect the diurnal maximum at P3 around 0930 CEST and the diurnal minimum 12 h later around 2130 CEST with an amplitude of about 0.38 hPa. Hence, the diurnal tide might explain up to a quarter of the signal evident in Fig. 7a. The pressure gradient P3–P1 (Fig. 7b) remains almost constant around zero, while the gradient P2–P1 has a minimum of −0.05 hPa km−1 in the afternoon. This implies a local pressure minimum at P2, which was observed on several days with a thermal circulation present in the Alpine foreland. The reversal in sign of the pressure gradient P3–P2 (which can be inferred from the pressure gradients P3–P1 and P2–P1) at 0900 CEST leads the onset of northeasterly flow at P2 by about one hour. The increase in wind speed until 1400 CEST can also be explained by the increasing pressure difference between P2 and P3. Bearing in mind that the distance P3–P2 is about twice the distance P2–P1, the southward acceleration P3–P2 of the flow acts over a larger horizontal extent than its deceleration P2–P1. Thus, the weak northeasterly flow at P1 is due to the inertia of the accelerated air parcels. The wind variation before 0930 CEST at P1 can be explained by the peaks in the pressure gradients P3–P1 and P2–P1 and is most likely linked to a local flow phenomenon rather than to a larger-scale thermal circulation.

The diurnal cycle for the pressure gradient V2–P1 shows forcing of down-valley flow in the morning until the pressure gradient changes its sign at 0930 CEST, marking the onset of up-valley flow forcing accompanied by a continuous deceleration of down-valley flow thereafter until the wind changes its direction to up-valley flow at 1200 CEST. The sign of the pressure gradient changes again around 2030 CEST, coincident with the onset of the down-valley flow. However, there might still be inflow at 100 m AGL. Similar time offsets in the flow response were also found for the Inn Valley (Zängl 2004).

The wind at TB (not shown) was from the south with 5 m s−1 in the morning until 1200 CEST and then shifted toward the northeast, peaking around 3 m s−1 in the late afternoon. At 1900 CEST the wind weakened and turned southerly again.

Analysis of the full dataset reveals that during fair weather conditions the thermal circulation features northerlies along the foothills of the Alps (see Fig. 3), accompanied by a consistently identified local pressure minimum at P2. The most plausible cause for the afternoon pressure minimum at P2 is a pressure perturbation pattern initiated by a gravity wave response to the ridge between P2 and P3 as well as to the ridge between P1 and V3 (Vilser Mountains). For two-dimensional (xz), hydrostatic, Boussinesq flow over orography, gravity wave theory yields ′ = u (e.g., Smith 1979), where ^ denotes the Fourier transform of the variable and p′, ,, N, and h denote the perturbation pressure, background density, basic state flow, Brunt–Vaisala frequency, and the height of the orography, respectively. The factor i = −1 indicates that the perturbation pressure is 90° out of phase with the orography, yielding positive (negative) pressure perturbations on the windward (lee) side of the ridge. In prevailing northeasterly flow, P2 is on the lee of the P3–P2 ridge and P1 is on the windward side of the P1–V3 ridge (Vilser Mountains). Thus, we expect a negative gravity wave pressure perturbation at P2 and a positive one at P1. We can estimate the relative contribution to the observed pressure gradients by assuming N = 0.01 s−1, ρ = 1 kg m−3, and u = 5 m s−1. The amplitude of the pressure perturbation corresponding to the ridge between P3 and P2 (100 m) together with the pressure perturbation related to the Vilser Mountains (400 m) yields a pressure gradient of −0.038 hPa km−1 for P2–P1, which explains about 80%–90% of the observed values (see Fig. 7b). The residual is most likely due to the pressure deficit at P2 connected to the thermal circulation of the adjacent Halblech Valley.

It should be noted that the time series of the pressure gradient P3–P1, which levels around zero, is somewhat equivocal. In general one would expect a gradual pressure decrease toward the Alpine crest in the afternoon related to the larger-scale Alpine heat low. The rather flat curve implies a balance between the pressure decrease in the area of P1 and the increase in pressure due to local effects such as the gravity wave response. However, the distance between the AWS sites is too large to fully resolve this phenomenon.

b. 17 August

The flow evolution on 17 August features the sudden onset of northeasterly flow in the area of P2 mentioned in the introduction. The MSLP field implies easterly surface flow with rather weak gradients of geopotential at 500 hPa (Fig. 8).

The wind speed (≈6 m s−1) and direction (≈120°) at P3 are almost constant after 0900 CEST for the entire day (Fig. 9), which is also the case for the AWS at P4 (not shown). This continuance is in contrast to the flow evolution at the other stations, where the wind speed at P2 remains below 2 m s−1 in the morning until 1130 CEST when a sudden onset of northeasterly flow occurs, with surface wind speeds peaking around 7 m s−1 in the early afternoon. The time series at P1 shows very similar features with calm conditions before 1330 CEST followed by a sudden onset of northeasterly flow peaking around 5 m s−1. The striking difference between the time evolution at P2 and P1 is the time shift in the onset of northeasterly flow by about 2 h. The wind at P2 weakens around 1500 CEST but northeasterly flow is evident until the late evening, whereas at P1 the wind speed decreases after 1900 CEST and shifts to southerly directions, attaining wind speeds around 2 m s−1.

At V2 the diurnal valley wind circulation is evident with down-valley flow until 1100 CEST and increasing up-valley flow thereafter, with maximum values around 6 m s−1. The upslope flow starts to weaken after 1900 CEST. The wind at TB (not shown) was from southerly to easterly directions for the entire day with wind speeds between 1 and 5 m s−1.

The surface pressure time series at P1 (Fig. 10) features a maximum in the morning and a minimum in the late afternoon similar to 28 July. The pressure gradient P3–P1 indicates southward forcing of flow from 1100 CEST onward with almost zero pressure gradients P2–P1, indicating that the flow is first accelerated between P3 and P2. It should be noted that the pressure gradient increase leads the sudden onset of wind at P2 by about half an hour. The marked pressure gradient increase for P2–P1 at 1400 CEST is followed by a sudden onset of wind at P1 within 15 min. The V2–P1 time series behaves very similarly to 28 July.

The analyses presented above indicate that P3 and P4 tend to record the synoptic easterly flow, whereas P1 and P2 experience rather calm conditions in the morning. Thus, a zone of strong surface wind shear must exist between P2 and P3. The presence of the Ammer Mountains on the windward side suggests that P1 and P2 might be in the wake of this ridge, which possibly generates the shear zone on its lee. Once the pressure gradients are initiated by heating, it appears that the shear zone is forced southward. It should also be noted that the large-scale thermal circulation is stronger than on 28 July.

To support the hypothesis of the moving shear zone and to obtain a more complete three-dimensional picture of the flow evolution, we conducted several model studies to pinpoint the inherent dynamical features.

4. MM5 experiments

a. Setup

The numerical simulations have been conducted with the fifth-generation Pennsylvania State University–National Center for Atmospheric Research (NCAR) Mesoscale Model (MM5), version 3.6 (Grell et al. 1995). Five two-way nested model domains are used with a horizontal mesh size of 64.8 km, 21.6 km, 7.2 km, 2.4 km, and 800 m, respectively; 38 levels are used in the vertical. The lowermost is located at about 10 m AGL. The vertical distance between the model levels increases from about 40 m close to the ground to 800 m near the upper boundary at 100 hPa. The initial and boundary conditions used for the real-case simulations are obtained from operational European Centre for Medium-Range Weather Forecasts (ECMWF) analysis data. In addition, two semi-idealized simulations have been conducted by combining realistic and slightly modified topography with idealized large-scale flow conditions. Further details are given in the appendix.

b. Results

1) Case study: 28 July

The flow at 180 m AGL at 0800 CEST features a down-valley flow along the Lech Valley with a strong jetlike outflow at the Lech Valley exit reaching maximum wind speeds around 10 m s−1 in the area near P1 but also extending some 30 km to the north of the valley exit (Fig. 11a). Inspection of the vertical structure suggests that the wind maximum at the exit is caused by hydraulic effects with a coincident dipping of isentropic surfaces in the lee of the Vilser Mountains (not shown). This strong outflow into the foreland is absent in the surface observations. However, rather intense winds during nocturnal and early morning valley outflow are well known by the local wind surfers of Lake Weissensee just on the lee side of the Vilser Mountains. The jetlike flow disappears rapidly after sunrise because of vertical mixing of momentum initiated by surface heating. It should be noted that the early morning temperature has a positive bias in the model simulations, most likely due to an underestimation of the surface-based nocturnal temperature inversion. However, during the day differences in temperature are usually smaller than 1 K. Probably because of the underestimation of the low-level inversion, the surface winds are overestimated in the model compared to the observations. The wind during the daytime has small deviations compared to the observations.

In the afternoon (Fig. 11b), up-valley flow is evident in the Lech Valley, accompanied by north to northeasterly flow in the Alpine foreland with wind speeds around 4 m s−1, which is in accordance with the observations. A vertical cross section through the line marked in Fig. 2 is shown in Fig. 12a, indicating that the flow crosses the Vilser Mountains southwest of P1 and subsequently contributes to the up-valley flow.

Figure 13a shows pressure perturbation and wind at 1200 m above sea level at 1600 CEST, indicating a pressure gradient forcing flow from the foreland toward the Alps and up-valley. The wind in the vicinity of the valley follows the pressure gradient as is usual for highly ageostrophic thermally driven flow. In the Alpine foreland a geostrophic (easterly) component is evident.

In the pressure gradient time series for selected stations (Fig. 14a), a clear diurnal signal is evident for V2–P1 depicting the forcing of the valley wind circulation; P3–P1 and P2–P1 have a much less pronounced diurnal cycle and indicate weak pressure gradients forcing flow toward P1 during most of the day. Comparing the time series of pressure gradients with observations (Fig. 7), a good qualitative agreement is evident but the model fails to predict the observed pressure minimum at P2 and tends to underestimate the amplitude of the diurnal pressure gradient V2–P1. However, the difference between model and observations is smaller than 0.02 hPa km−1, except at P2. The absence of the P2 pressure minimum in the numerical simulations is most likely due to the coarser resolution of the topography, which does not feature the ridge to the north of P2 and the finer structure of the Halblech Valley and which underestimates the height of the Vilser Mountains between P1 and V3.

2) Case study: 17 August

At 0800 CEST (Fig. 15) valley outflow is only evident in the region of the lower Lech Valley, corresponding to very calm conditions in the P1 and P2 area. However, a strong easterly flow is present to the north of the Ammer Mountains and at the crests of the mountains in the area. This easterly flow is related to the synoptic conditions (see Fig. 8). By 1100 CEST (Fig. 15b) up-valley flow develops in the Lech Valley, whereas wind speeds near P1 and P2 are still insignificant. However, there is a zone of strong wind shear to the north of P2 near P3. This shear zone subsequently progresses southward, penetrating into the P2 area. By 1200 CEST the shear zone reaches the Lech Valley exit, leaving northeasterly flow behind it (not shown). At 1600 CEST the model produces rather uniform northeasterly flow in the Ostallgäu (Fig. 15c). In the vertical cross section at 1600 CEST (Fig. 12b), some reminiscence of the wind shear zone is evident near V3 (x ≈ 20 km). The time evolution of the surface flow compares well with the observations, especially the wake in the P2 area and the easterly flow farther away from the Alps. However, the modeled passage of the wind shear zone at P2 and P1 leads the observations by one and three hours, respectively.

Stronger gradients of pressure perturbations, compared to the 17 July case, are evident in Fig. 13b, forcing flow toward the south and up the Lech Valley. This fact is also depicted in Fig. 14b, where the time series of the pressure gradients is shown. Again, the forcing of the valley wind circulation is clearly seen in the V2–P1 time series. Looking at the pressure gradients P3–P1 and P2–P1, a sudden increase between 1100 and 1200 CEST is evident, with a time lag of about 30 min between the two curves indicating that a positive gradient is first established between P2 and P3. The amplitudes of the diurnal cycles are about 2–3 times larger than the ones for the 28 July case. However, one should bear in mind the contribution from the synoptic-scale pressure gradient related to the easterly flow. Comparing the modeled pressure gradients P3–P1 and P2–P1 to observations (Fig. 10), the model leads the changes in pressure gradients by about one and three hours, respectively, in accordance with the wind (see above). The amplitude of the diurnal cycle of the pressure gradient V2–P1 is again underestimated by the model (maximum error ≈0.02 hPa km−1), yet the wind speed in the Lech Valley compares well to the observations. The model also produces higher wind speeds in the Lech Valley compared to 28 July, in agreement with the observations.

In general it should be noted that the model lacks the observed variations in wind, temperature, and pressure on time scales smaller than 1 h. However, this is not unexpected because the MM5 model relies on parameterizations of turbulent processes and is not able to explicitly resolve eddies on these time scales.

3) Idealized study with modified topography

The case study for 17 August suggests that the wake effect of the Ammer Mountains plays a crucial role in the generation of the aforementioned shear zone in prevailing synoptic easterly flow. To pinpoint the sensitivity of the flow evolution to the existence of the Ammer mountains, two semi-idealized simulations with easterly flow (see appendix for details) were carried out to investigate the dynamical differences.

There are striking similarities in the wind field between the case study for 17 August (Fig. 15b) and the idealized simulation with real topography (Fig. 16a). Note the resemblance of the shear zone just north of P2 in the lee of the Ammer Mountains. The shear zone is completely absent in Fig. 16b where the mountain range was removed, featuring prevailing easterly flow in the Alpine foreland. This test highlights the importance of the Ammer Mountains in generating the wake, which is accompanied by the shear zone. Another interesting feature to point out is the increased northeasterly flow component in the area of P2 evident in Fig. 16a relative to Fig. 16b, pinpointing a flow channeling by the asymmetric valley exit.

5. Discussion and concluding remarks

In this study, we tried to look for the relationship between the Lech Valley circulation and the northeasterly flow near P2. We found that the northeasterly flow near P2 only occurs in tandem with the Lech Valley circulation. However, we also detected several days on which the Lech Valley circulation was evident, while no northeasterly flow developed near P2. Analyzing the onset times of the thermal circulation, there is no evidence that the Lech Valley wind circulation is leading the northeasterly flow at P2.

Lugauer and Winkler (2005) suggested a local heat low situated to the west of the Ostallgäu as the forcing mechanism for the northeasterly flow. This hypothesis cannot be ruled out using our observations, but the model simulations did not produce a local heat low in the proposed area in either of the case studies. The far-reaching (150–200 km outward from the main Alpine crest) convergent flow into the Alps on summer days forced by the Alpine heat low (Burger and Ekhart 1937; Weissmann et al. 2005) can be seen as a more general mechanism forcing the northeasterly flow near P2. Based on the arguments above, we argue that the Lech Valley circulation is not the sole driving mechanism. The northeasterly flow near P2 is rather a result of the easterly flow related to the Alpine heat low, which is partly deflected by the Ammer Mountains to the southwest into the Lech Valley and cannot be seen as a simple extension of the Lech valley winds.

We identified a local pressure minimum on days with a thermal circulation at P2, which is most likely due to hydrostatic pressure perturbations initiated by a gravity wave response above the ridges between P1 and V3 as well as between P3 and P2.

The higher frequency of the Lech Valley circulation compared to the occurrence of the thermal circulation in the Alpine foreland is most likely linked to the different spatial scales of the wind systems and a channeling of the flow in the valley. The difference in magnitude of thermally initiated pressure gradients can usually be attributed to the so-called volume effect [also called the topographic amplification factor (TAF); see Wagner 1932; Steinacker 1984; Whiteman 1990]. The TAF describes the ratio of the volume of air within a valley to a volume of air comprising the same horizontal area but over a flat surface. Assuming that the total solar energy input for a given surface area is the same, the TAF describes the ratio of energy being available for heating the volume of air. Thus the TAF is closely linked to differences in the diurnal temperature evolution between the valley and the plane and hence to the hydrostatic pressure gradients and the intensity of the diurnal valley winds. For an idealized V-shaped valley the TAF is exactly 2; if the valley slopes become more convex (concave), the TAF will become larger (smaller) than 2. It should be noted that if tributaries are included the TAF becomes even larger (Steinacker 1984). The Lech Valley TAF, excluding tributaries, is 2.02 and is thus comparable to the TAF of the Austrian Inn Valley including tributaries (TAF = 2.1; Steinacker 1984). For comparison, the smaller and higher Alpine valleys like the Stubai, Wipp, and Dischma Valleys have TAFs of 3, 3, and 2.7, respectively (Steinacker 1984; Müller and Whiteman 1988). Even though the TAF appears quite useful in comparing the strengths of different valley wind systems, it is not obvious how one can explain the difference in frequency between the two wind systems observed at the Lech Valley exit by only using TAF arguments.

Regarding the mechanism yielding the sudden onset of northeasterly flow near P2, the observations suggest the existence of a shear zone between P2 and P3 moving southward during the daytime. This hypothesis is supported by the numerical simulations, where synoptic easterly flow yields a wake in the lee of the Ammer Mountains during low Froude number conditions in the night and early morning. At the northern edge of the wake a shear zone is present (Fig. 15a), which is subsequently forced southward by the pressure gradients related to differential heating between the foreland and the Alps. According to Wippermann and Gross (1981) and Whiteman and Doran (1993), the pressure gradient related to the synoptic easterly flow can also lead to a channeling of flow in the presence of mountainous terrain. This could be an additional mechanism moving the shear zone southward. However, the pressure gradient related to a geostrophically balanced 5 m s−1 easterly wind only explains 10%–20% of the observed pressure gradients. Removing the Ammer Mountains in a semi-idealized model experiment illustrates the effects of the asymmetry of the valley exit, since no wind shear zone is generated with a symmetric valley exit and easterly flow prevails in the entire Alpine foreland (see Figs. 16a,b).

The Denver Cyclone (convergence vorticity zone; Szoke 1991) evolves in a similar flow setup with a wake behind a mountain range. Crook et al. (1990) showed that a low Froude number flow around a mountain range produces a vortex in the lee, which subsequently moves downstream leaving a wake behind. However, in our simulations neither a cyclone in the lee nor a subsequent downstream development is evident. There are two possible reasons for these differences: (i) different scales of the two phenomena, which are separated by an order of magnitude, or (ii) the wake in the lee of the Ammer Mountains is affected by the Lech Valley wind circulation. Rerunning the idealized model simulation but stretching the real topography by a factor of 5 yields basically the same results presented above, indicating that the scale is most likely not the key parameter determining the difference. We believe that the absence of the lee vortex in the Ostallgäu is primarily related to the wind field at the valley exit region. The nocturnal valley outflow counteracts the lee vortex formation and the daytime low-level wind field acts to remove low-level air into the Lech Valley hampering the wake formation.

The results for an asymmetric valley exit presented in this study are inherent to the Lech Valley. However, we are confident that the discussed features are also evident at other asymmetric valley exits, especially the difference between the higher frequency of valley winds compared to the frequency of the larger-scale thermal circulation in the foreland, as well as possible wake effects due to the topographic setting and the synoptic flow. It would be worthwhile to compare our findings to flow evolutions at other valley exits as well as utilizing idealized model studies to pinpoint the influence of the degree of asymmetry on the flow evolution.

Acknowledgments

The authors are grateful to the communities of Rieden and Buching for their support. Special thanks are given to the DAeC Gleitschirmschule for helpful input in the planning phase and the Forggensee Yachtschule, where we had our base during the entire campaign. We also thank all the farmers (A. and M. Falger, E. and I. Falger, T. Storf, T. Velle, J. and R. Bellmund, T. Häußrer, J. Christa, J. Fritsch and J. and H. Greißl) for providing some space for measuring equipment and last but not least the staff for AllgEx: H. Lösslein, H. Wendt, H. Aschenbrenner, P. Kolb, C. Schmidt, S. Lämmlein, M. Garhammer, I. Nudelmann, S. Raith, C. Schmidt, M. Zink, and M. Zwanzger. We also appreciate interesting discussions with C. Davis on the material pointing out some similarities to the Denver Cyclone. We thank Tracy Ewen for a careful proofreading of the manuscript. The comments by the anonymous referees were constructive and helpful.

REFERENCES

  • Burger, A. E., , and E. Ekhart, 1937: Über die tageszeitliche Zirkulation der Atmosphäre im Bereich der Alpen (On the daily circulation of the atmosphere over the Alps). Gerlands Beitr. Geophys., 49 , 341367.

    • Search Google Scholar
    • Export Citation
  • Chapman, S., , and R. S. Lindzen, 1970: Atmospheric Tides. D. Reidel, 201 pp.

  • Crook, N. A., , T. L. Clarke, , and M. W. Moncrieff, 1990: The Denver Cyclone. Part I: Generation in low Froude number flow. J. Atmos. Sci., 47 , 27252742.

    • Search Google Scholar
    • Export Citation
  • Crook, N. A., , T. L. Clarke, , and M. W. Moncrieff, 1991: The Denver Cyclone. Part II: Interaction with the convective boundary layer. J. Atmos. Sci., 48 , 21092126.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., 1989: Numerical study of convection observed during the Winter Monsoon Experiment using a mesoscale two-dimensional model. J. Atmos. Sci., 46 , 30773107.

    • Search Google Scholar
    • Export Citation
  • Egger, J., , S. Bajarachaya, , U. Egger, , R. Heinrich, , J. Reuder, , P. Shakaya, , H. Wendt, , and V. Wirth, 2000: Diurnal winds in the Himalayan Kali Gandaki Valley. Part I: Observations. Mon. Wea. Rev., 128 , 11061122.

    • Search Google Scholar
    • Export Citation
  • Egger, J., and Coauthors, 2002: Diurnal winds in the Himalayan Kali Gandaki Valley. Part III: Remotely piloted aircraft soundings. Mon. Wea. Rev., 130 , 20422058.

    • Search Google Scholar
    • Export Citation
  • Egger, J., and Coauthors, 2005: Diurnal circulation of the Bolivian Altiplano. Part I: Observations. Mon. Wea. Rev., 133 , 911924.

  • Ekhart, E., 1944: Strömungsaerologie der Talwinde (Aerology of valley winds). Meteor. Z., 61 , 217231.

  • Freytag, C., 1985: MERKUR—Results: Aspects of the temperature fields and the energy budget in a large alpine valley during mountain and valley wind. Contrib. Atmos. Phys., 58 , 458476.

    • Search Google Scholar
    • Export Citation
  • Freytag, C., , and B. Hennemuth, 1983: MERKUR: Mesoskaliges Experiment im Raum Kufstein-Rosenheim (Mesoscale experiment in the region of Kufstein-Rosenheim). Wiss. Mitt. Meteor. Inst. Univ. München, 48, 132 pp.

    • Search Google Scholar
    • Export Citation
  • Grell, G. A., , J. Dudhia, , and D. R. Stauffer, 1995: A description of the fifth-generation Penn State/NCAR Mesoscale Model (MM5). NCAR Tech. Note NCAR/TN-398+STR, 122 pp.

    • Search Google Scholar
    • Export Citation
  • Hafner, T. A., , M. E. Reinhardt, , E. L. Weisel, , and H. P. Fimpel, 1987: Boundary layer aspects and elevated heat source effects of the Alps. Meteor. Atmos. Phys., 36 , 6173.

    • Search Google Scholar
    • Export Citation
  • Haurwitz, B., , and A. D. Cowley, 1973: The diurnal and semi-diurnal barometric oscillations: Global distribution and annual variation. Pure. Appl. Geophys., 102 , 193222.

    • Search Google Scholar
    • Export Citation
  • Heimann, D., 1997: Mesoscale surface-wind characteristics and potential gravity wave formation during cross-alpine airflow. Meteor. Atmos. Phys., 62 , 4970.

    • Search Google Scholar
    • Export Citation
  • Hoinka, K. P., 1980: Synoptic-scale atmospheric features and foehn. Beitr. Phys. Atmos., 53 , 486508.

  • Kain, J. S., 2004: The Kain–Fritsch convective parameterization: An update. J. Appl. Meteor., 43 , 170181.

  • Klemp, J. B., , and D. R. Durran, 1983: An upper boundary condition permitting internal gravity wave radiation in numerical mesoscale models. Mon. Wea. Rev., 111 , 430444.

    • Search Google Scholar
    • Export Citation
  • Lugauer, M., , and P. Winkler, 2005: Thermal circulation in south Bavaria—Climatology and synoptic aspects. Meteor. Z., 14 , 1430.

  • Mlawer, E. J., , S. J. Taubman, , P. D. Brown, , M. J. Iacono, , and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmosphere: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102 , 1666316682.

    • Search Google Scholar
    • Export Citation
  • Müller, H., , and C. D. Whiteman, 1988: Breakup of a nocturnal temperature inversion in the Dischma Valley during DISKUS. J. Appl. Meteor., 27 , 188194.

    • Search Google Scholar
    • Export Citation
  • Müller, H., , R. Reiter, , and R. Sládkovič, 1984: Die vertikale Windstruktur beim MERKUR-Schwerpunkt ‘Tagesperiodische Windsysteme’ aufgrund von aerologischen Messungen im Inntal und im Rosenheimer Becken (On the vertical wind structure of the diurnal wind system within the Inn Valley and the adjacent plain: Results of aerological soundings during the field experiment MERKUR). Arch. Meteor. Geophys. Bioklimatol., B33 , 359372.

    • Search Google Scholar
    • Export Citation
  • Pamperin, H., , and G. Stilke, 1985: Nächtliche Grenzschicht und LLJ im Alpenvorland nahe dem Inntalausgang (Nocturnal boundary layer and LLJ in the pre-alpine region near the Inn Valley exit). Meteor. Rundsch., 38 , 145156.

    • Search Google Scholar
    • Export Citation
  • Reisner, J., , R. M. Rasmussen, , and R. T. Bruintjes, 1998: Explicit forecasting of supercooled liquid water in winter storms using the MM5 mesoscale model. Quart. J. Roy. Meteor. Soc., 124 , 10711107.

    • Search Google Scholar
    • Export Citation
  • Sládkovič, R., , and H-J. Kanter, 1977: Low-level jet in the Bavarian pre-Alpine region. Arch. Meteor. Geophys. Bioklimatol., A25 , 343355.

    • Search Google Scholar
    • Export Citation
  • Smith, R. B., 1979: The influence of mountains on the atmosphere. Adv. Geophys., 21 , 87230.

  • Steinacker, R., 1984: Area height distribution of a valley and its relation to the valley wind. Contrib. Atmos. Phys., 57 , 6471.

  • Szoke, E. J., 1991: Eye of the Denver Cyclone. Mon. Wea. Rev., 119 , 12831292.

  • Szoke, E. J., , M. L. Weisman, , J. M. Brown, , F. Caracena, , and T. W. Schlatter, 1984: A synoptic analysis of the Denver tornadoes of 3 June 1981. Mon. Wea. Rev., 112 , 790808.

    • Search Google Scholar
    • Export Citation
  • Thompson, G., , R. M. Rasmussen, , and K. Manning, 2004: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part I: Description and sensitivity analysis. Mon. Wea. Rev., 132 , 519542.

    • Search Google Scholar
    • Export Citation
  • Wagner, A., 1932: Der tägliche Luftdruck- und Temperaturgang in der freien Atmosphäre und in Gebirgstälern (Diurnal variations of pressure and temperature in the free atmosphere and in valleys). Gerlands Beitr. Geophys., 37 , 315344.

    • Search Google Scholar
    • Export Citation
  • Weissmann, M., , F. J. Braun, , L. Gantner, , G. J. Mayr, , S. Rahm, , and O. Reitebuch, 2005: The Alpine mountain–plain circulation: Airborne Doppler lidar measurements and numerical simulations. Mon. Wea. Rev., 133 , 30953109.

    • Search Google Scholar
    • Export Citation
  • Whiteman, C. D., 1990: Observations of thermally developed wind systems in mountainous terrain. Atmospheric Processes over Complex Terrain, Meteor. Monogr., No. 45, Amer. Meteor. Soc., 5–42.

    • Search Google Scholar
    • Export Citation
  • Whiteman, C. D., , and J. C. Doran, 1993: The relationship between overlying synoptic-scale flows and winds within a valley. J. Appl. Meteor., 32 , 16691682.

    • Search Google Scholar
    • Export Citation
  • Wippermann, F., , and G. Gross, 1981: On the construction of orographically influenced wind roses for given distributions of the large-scale wind. Beitr. Phys. Atmos., 54 (4) 492501.

    • Search Google Scholar
    • Export Citation
  • Zängl, G., 2002a: An improved method for computing horizontal diffusion in a sigma-coordinate model and its application to simulations over mountainous topography. Mon. Wea. Rev., 130 , 14231432.

    • Search Google Scholar
    • Export Citation
  • Zängl, G., 2002b: Stratified flow over a mountain with a gap. Linear theory and numerical simulations. Quart. J. Roy. Meteor. Soc., 128 , 927949.

    • Search Google Scholar
    • Export Citation
  • Zängl, G., 2003: A generalized sigma-coordinate system for the MM5. Mon. Wea. Rev., 131 , 28752884.

  • Zängl, G., 2004: A reexamination of the valley wind system in the Alpine Inn Valley with numerical simulations. Meteor. Atmos. Phys., 87 , 241256.

    • Search Google Scholar
    • Export Citation
  • Zängl, G., 2007: Small-scale variability of orographic precipitation in the Alps: Case studies and semi-idealized numerical simulations. Quart. J. Roy. Meteor. Soc., 133 , 17011716.

    • Search Google Scholar
    • Export Citation
  • Zhang, D., , and R. A. Anthes, 1982: A high-resolution model of the planetary boundary layer—Sensitivity tests and comparisons with SESAME-79 data. J. Appl. Meteor., 21 , 15941609.

    • Search Google Scholar
    • Export Citation

APPENDIX

Model Setup

The simulations presented in this study use a complete set of physics parameterizations. Cloud microphysics is treated with the Reisner–Thompson scheme (Reisner et al. 1998; Thompson et al. 2004). Subgrid-scale convection is treated with the updated Kain–Fritsch scheme (Kain 2004) in the three outer domains, whereas no convection parameterization is used in domains 4 and 5. Radiation is calculated with the so-called cloud radiation scheme for the solar part (Dudhia 1989) and with the Rapid Radiative Transfer Model (RRTM) scheme (Mlawer et al. 1997) for the longwave part. The planetary boundary layer is parameterized with the Blackadar scheme (Zhang and Anthes 1982). It is coupled with a simple predictive scheme for soil moisture, which is initialized with climatological data depending on the land-use class. Numerical diffusion of temperature and of the mixing ratios of water vapor and cloud water is computed with the improved scheme described in Zängl (2002a). At the upper model boundary, an improved version of Klemp and Durran’s (1983) radiative boundary condition is used to prevent spurious reflections of vertically propagating gravity waves (Zängl 2002b). All simulations are conducted with the generalized vertical coordinate system described by Zängl (2003), which allows for a rapid decay with height of small-scale topographic structures in the coordinate surfaces.

The initial data from ECMWF is used in the pressure-level version on a 0.25° × 0.25° latitude–longitude grid. The simulations are started at 0000 UTC 28 July 2005 and 0000 UTC 17 August 2005, respectively, and are integrated for 22 h in each case.

The procedure to generate the idealized flow conditions for the semi-idealized experiments is described in Zängl (2007). The flow field considered here has a wind direction of 90° and positive vertical shear with the flow speed increasing from 7.5 m s−1 at sea level to 20 m s−1 at 250 hPa, with constant values higher above. The ambient wind field is in geostrophic balance, assuming a constant Coriolis parameter of 10−4 s−1. The temperature profile in the domain center starts from a sea level temperature of 15°C, followed by an isothermal layer up to a pressure of 900 hPa. Higher above, the temperature decreases at an average rate of 7 K km−1 up to the tropopause, which is located at 250 hPa. The stratosphere is again assumed to be isothermal. To avoid the formation of clouds in the idealized simulation, the relative humidity is set to 50% within the low-level isothermal layer and to 30% in the remainder of the troposphere. Radiation in the semi-idealized simulations is computed for an artificial date of 10 August.

Fig. 1.
Fig. 1.

Schematic illustration of different valley exits with possible flow features. (a) Daytime up-valley flow and (b) nocturnal jetlike down-valley outflow for a symmetric valley exit, respectively. (c) Possible up- and down-valley flow for an asymmetric valley exit. (d) Impact of a background flow on the thermally driven circulation in an asymmetric valley exit area.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 2.
Fig. 2.

Topographic map of the Ostallgäu and the Lech Valley. Letters indicate AWS positions [V1 (Holzgau, 1093 m), V2 (Stanzach, 940 m), V3 (Reutte, 841 m), P1 (Schwangau, 792 m), P2 (Buching, 796 m), P3 (Steingaden, 786), and P4 (Lengenwang, 811 m)]. The pertinent mountain ranges are also indicated, as well as the Vilser Mountains (VM, highlighted by the dashed circle) and the Halblech Valley (HV). The line indicates the position of the cross section in Fig. 12.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 3.
Fig. 3.

Scatterplot of 3-hourly mean wind for V2 vs P2, morning (0430–0730 CEST; square) and afternoon (1530–1830 CEST; circle). Filled symbols indicate that the daily mean global radiation exceeded 10 MJ m−2. A threshold for wind speed >0.5 m s−1 was applied. The horizontal and vertical lines indicate the upward (solid) and downward (dashed) direction of the valley axis. The diagonal is indicative for identical values at V2 and P2 and the dotted diagonals are the 45° deviation from it.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 4.
Fig. 4.

MSLP with contour interval (CI) of 2 hPa (solid) and geopotential (CI = 20 gpm) at 500 hPa (dashed) for 1200 UTC 28 Jul. Topography shaded in increments of 500 m (white for sea level). The X marks the area of the field experiment.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 5.
Fig. 5.

Time series of wind speed (solid lines), υ (m s−1), and direction (°, crosses) for selected AWS stations—(a) P3, (b) P2, (c) P1, and (d) V2—for 28 Jul.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 6.
Fig. 6.

Potential temperature θ and specific humidity q from descents of Kali and wind speed, υ, and direction, dir, from pilot balloon ascents during 28 Jul near P2. Different times during the day are indicated by different line styles or symbols.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 7.
Fig. 7.

(top) Time series of surface pressure at P1 and (bottom) surface pressure gradients between selected AWSs (P3, P2, V2) and P1 for 28 Jul.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 8.
Fig. 8.

As in Fig. 4, but for 1200 UTC 17 Aug.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 9.
Fig. 9.

As in Fig. 5, but for 17 Aug.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 10.
Fig. 10.

As in Fig. 7, but for 17 Aug.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 11.
Fig. 11.

Wind vectors at ∼180 m AGL, with shading indicating total wind speed, for (a) 0800 and (b) 1600 CEST 28 Jul. Topography of the model domain is contoured; CI = 200 m.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 12.
Fig. 12.

Vertical cross section along the line marked in Fig. 2 showing contours θ (K) and the projected wind vectors and wind speed (shaded) onto the cross section. Both panels are valid for 1600 CEST (a) 28 Jul and (b) 17 Aug.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 13.
Fig. 13.

Pressure perturbation (shaded) and wind vectors at 1200 m above sea level at 1600 CEST for (a) 28 Jul and (b) 17 Aug. Topography of the model domain is contoured; CI = 400 m.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 14.
Fig. 14.

Time series of surface pressure gradients from MM5 model simulations for (a) 28 Jul and (b) 17 Aug. All stations (P3, P2, V2) were reduced to the height of P1.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 15.
Fig. 15.

As in Fig. 11, but for (a) 0800, (b) 1100, and (c) 1600 CEST 17 Aug.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Fig. 16.
Fig. 16.

Wind vectors at ∼180 m AGL, shading indicating total wind speed, for idealized setup with easterly flow (a) including real topography and (b) without the Ammer Mountains, both at 1100 CEST. Topography of the model domain is contoured; CI = 200 m.

Citation: Monthly Weather Review 137, 10; 10.1175/2009MWR2779.1

Table 1.

Dates of intensive observation periods (IOPs) during AllgEx 2005 with no. of pilot balloons (PB) released and Kali ascents (KA).

Table 1.
Save