Introduction
Orographic effects on atmospheric flow can produce or modify precipitating clouds through orographic lifting, triggering of convection, indirect effects of flow splitting or blocking, and induced waves. Of course, some of these effects can occur simultaneously, but for simplicity, each of them is examined independently. The most direct way a mountain influences precipitation is when the air flowing over is moist enough for condensation to occur through a deep layer (Fig. 2-1 b).
Figure 2-1. Mechanisms of orographic precipitation: (a) Seeder-feeder mechanism; (b)
upslope condensation; (c) upslope triggering of convection; (d) upstream triggering of
convection; (e) thermal triggering of convection; (f) leeside triggering of convection;
(g) leeside enhancement of convection. Slanted lines below cloud base indicate
precipitation. [From Fig. 12.24 of Houze, 1993].
Although precipitating clouds can form in pure orographic flow, such a process leads more generally to a local enhancement of precipitation within pre-existing clouds, such as those associated with a front passing over a mountain range. On the opposite, air is dried during the downward motions, and precipitation is significantly reduced, on the lee side of the mountain. Another effect is the "seeder-feeder" mechanism according to which precipitating hydrometeors that originate from a cloud layer aloft (the "seeder" cloud) grow at the expense of the water content of a cloud below (the "feeder" cloud) which, by itself, might not precipitate (Fig. 2-1 a). Stratus and small cumulus clouds orographically formed over hills or mountains can be particularly effective feeder clouds. Orographically induced clouds can take the form of cumulus or cumulonimbus when the air subjected to lifting is sufficiently moist and unstable (Fig. 2-1 c). Such orographic cumulonimbi can be very important precipitation producers, as it has been observed during past MAP-related events (e.g. the Brig, South Ticino and Piedmont cases). It is recognized however, that the complex interactions between cloud dynamics and microphysics, and orographic forcing are far from being completely understood. Thermal forcing occurs when daytime heating produces an elevated heat source and a corresponding thermally direct circulation, with convergence and convection at the top of the mountain (Fig. 2-1 e). Vertically propagating waves, which are associated with mountains of any size, can tilt upstream and trigger convection at some distance ahead of the barrier (Fig. 2-1 d). The combined effects of mid-level upward motions associated with such a wave and low-level thermally induced upslope flow can also enhance convection on the lee side (Fig. 2-1 g). Upstream lifting can also be triggered by blocking due to a large mountain barrier. A steady state can be achieved if the flow is partially blocked. At some distance upstream from the barrier, a stationary hydraulic jump develops where upward motions can initiate convective clouds (Fig. 2-1 d). On the lee side, severe downslope winds and rotor cloud formation are observed. Lee-side forcing results from low-Froude-number flow passing around an isolated obstacle: convergence induced by channelling of the flow in the lee of the mountain can trigger convection (Fig. 2-1 f).
In the Alps, heavy precipitation events are caused by synoptic-scale frontal systems whose mesoscale precipitation fields are enhanced by topography or by slowly moving deep convective systems developing over favourable mountain features. In both cases, orographic precipitation is the result of a complex interaction of dynamics and microphysics. Strong local upward air motions resulting from convective instability produce particles which grow by coalescence and riming. Weaker vertical air motions produce ice-phase precipitation particles growing by vapour deposition and aggregation of crystals. The weaker vertical motions can be the remains of earlier, stronger convective updrafts or they may be associated with forced ascent of stable air. Mesoscale convective cloud systems and frontal cloud systems contain both regions of young vigorous convective motions and weaker air motions. The microphysical processes feedback to the dynamics via thermodynamic effects. The correct prediction of the precipitation pattern entails getting the correct distributions of air motions down to the convective scale and predicting the right set of microphysical processes connected with those air motions.
Orographic precipitation in the Alps is prototypical of precipitation over mountain ranges (Cotton and Anthes, 1989; Banta, 1990; Houze, 1993). During the autumn, large amounts of precipitation occur in the MAP area on the southern slopes of the Alps. Climatologically, the rainfall is most concentrated over slopes exposed to air trajectories coming from the Western Mediterranean or the Adriatic sea, especially in the vicinity of certain valleys or indentations in the range (Fig. 2-2). Typical mesoscale weather systems and flow features in this region favour heavy and long lasting precipitation in the MAP area south of the Alps, especially Mediterranean depressions and fronts (e.g. Sénési et al., 1996). Multiple rainbands form within these larger-scale organized storms. Prior to their movement into the Alpine region, these storms take on a very well-defined organization understandable in terms of the dynamics of baroclinic instability and frontogenesis. Ahead of these systems, the warm moist flow is potentially unstable in the gravitational sense. Thus, the flow over the Alps offers the opportunity to study both the modification of stable balanced baroclinic systems and the release of instability induced by the orography. The knowledge gained by investigating these storms will be generally applicable to other orographic regimes.
Figure 2-2. Climatological frequency of heavy precipitation (over 20 mm d-1) over the Alps
during the month of October. Color coding shows percentage of days. Note the maxima over
Ticino and Friuli-Slovenian areas (after Frei and Schär, 1998)
Scientific Questions
1.1 How do special configurations of topography with respect to the large-scale flow (especially, the moist air flow in the low levels), such as curved mountain ranges, valleys, etc., concentrate precipitation into severe downpours and produce flash floods?
1.2 How does the flow over complex terrain modify the growth mechanisms of precipitation particles (vapour diffusion, riming, coalescence, and aggregation)?
1.3 How are pre-existing mesoscale systems (e.g. fronts, rainbands, convective systems) modified by the orography through uplift, mountain waves, blocking and channelling, and how do these modifications localize heavy and persistent precipitation?
Observational Requirements
Although the Alps are the world's most densely instrumented mountain range, routine measurements are not sufficient to support the development of high-resolution numerical forecast models, in order to forecast the location and amount of precipitation with high precision. Such forecasts require very accurate simulation of the precipitation growth processes, down to the microphysical scale. Standard operational and climatological meteorological observations are insufficient to provide the appropriate verification data.
Figure 2-3. Proposed triple ground radar array for dual-Doppler and microphysics
observations in the Ticino area. The color background plot indicates elevation in
kilometers above MSL. The yellow outlines indicate regions of low-resolution dual-Doppler
coverage between the MeteoSwiss radar at Monte Lema and the RONSARD at Casaleggio. The white
outlines indicate regions of high-resolution dual-Doppler coverage obtained between the
S-Pol at Vergiate and the other two radars (these locations are still tentative)
Detailed measurements of airflow and of the distribution of hydrometeors in time and space are needed, and in the current state of the technology, these can only be provided by properly instrumented radars. Doppler radars are needed to provide the airflow and radar reflectivity (which is related to the mass of hydrometeor water in a sample of air). Polarimetric radar measurements are needed to provide information on the spatial and temporal distribution of hydrometeor type (snow, rain, graupel, hail, etc.). The research radars used in combination with the rain gauges and disdrometers also increase the ability to map the rainfall in the spatial detail necessary to verify model results.
The approach for obtaining specialized radar measurements within this overall observing network is to use the NCAR S-Pol ground-based S-band polarimetric Doppler radar, the French RONSARD ground-based C-band Doppler radar, and the MeteoSwiss's C-band operational Doppler radar at Monte Lema in Ticino. These radars will be used in a coordinated program of ground-based radar measurements of the air motions and microphysics of precipitating cloud systems. The radars will cover the Po Valley and the windward slopes of the Alps in the vicinity of Lago Maggiore (Fig. 2-3). This region is located within a curved indentation of the Alpine ridge, where rainfall is concentrated on average when storms from the Mediterranean move over the Alps.
Within this larger valley, several deep local valleys may further enhance the precipitation. Mobile radars would be useful to increase the capability of measurements in the finer-scale river valleys embedded in the broader-scale mountain slopes. Facilities under consideration are the scanning C-band NSSL/NCAR Doppler radar on a roving vehicle ("Doppler on Wheels"), the transportable, K- and X-band vertically pointing radars of the University of Karlsruhe, and of ETH Zurich.
In addition to this specialized network, several operational meteorological radars around the Alps, especially in the Po valley, offer advanced capabilities, and some of them could be switched to a research scanning mode during the IOPs. Several Italian institutes already announced their interest to cooperate to the scientific objectives of MAP in this respect. Of particular interest are the radars at Bric della Croce (operational, close to Torino) and Spino d'Adda (research, close to Milano). The network of operational Doppler radars in the north-eastern part of Italy is well suited to conduct studies related to the precipitation maximum in the Italian-Slovenian area (Fig. 2-2).
The ground-based measurements will be complemented by airborne radar measurements. The Doppler radars onboard the NOAA P-3 and NCAR Electra aircraft would be particularly useful for mapping the detailed mesoscale airflow over the windward slopes of the Alps. Figure 2-4 contains a generic flight plan for collecting data with the aircraft. The plan consists of three modules: (1) thermodynamic data in the moist flow coming in from the Mediterranean, (2) high-resolution Doppler radar data depicting the precipitation and airflow over the slopes of the Alps, and (3) microphysical legs to supplement the radar data. Whenever possible the radar and microphysical modules will be used to supplement the ground-based measurements.
More special and conventional measurements are also needed to monitor the mesoscale 3D airflow within the area specially instrumented for MAP, and the flow modified by the Alpine topography. Essential data can be provided by aircraft observations, surface-based additional soundings and observational networks, wind profilers, rain gauges, and satellite data. The documentation of the upstream flow will be provided by the three regular sounding stations of the Po valley (Milano, Udine, and S. Pietro Capofiume). Additional conventional sounding stations that are needed to complement the existing stations in the region south of the Alps are planned in Nice, in Genova and near Verona.
Two wind profilers (UHF and VHF, see Fig. 2-5) and a radio-acoustic system of CNRM and of Italian institutes will be located as close as possible to the ground array of Doppler radars (Lago Maggiore area). They will provide high-resolution time-height description of the three wind components within the sampling domain of the Doppler radars, with particular emphasis on the PBL levels for the UHF profiler, and on the tropopause level for the VHF profiler.
An integrated station for PBL measurements will be set up by ISAO-CNR and IFA-CNR in the vicinity of the Doppler radar and wind profiler location south-west of the southern tip of Lago Maggiore. The following instruments are foreseen: sodar, tethered balloon, series of sonic anemometers, solar energy balance station, porometers and other ecophysiological instruments, and multi-collector lidar with Raman capabilities. Similar instrumentation will be also be set up in the inflow region of the Lago Maggiore in the Bellinzona plain by GIETH. Mass and energy exchanges will be studied together with the vertical distribution of water vapour. The purpose is to document the structure of the PBL at the initial stages of development of convection and precipitation, with special emphasis on deep convective events. For these aspects a strong link with P8 (structure of the planetary boundary layer studies) will be developed.
Altogether, the above measurements will allow for detailed studies of the lower-level moisture input and the mechanisms of troposphere-stratosphere exchanges during convective events.
The French Fokker27 research aircraft, equipped with the LEANDRE 2 backscattering lidar, will be available for this project during events of intense precipitation. This device will allow mapping the boundary layer wind, height, and water vapour content over the Po valley and the Ligurian or Adriatic sea, upstream of the main precipitation areas (Fig. 2-4). The soil moisture availability in the Po valley will also be assessed via low level measurements of the turbulent flux of humidity during dedicated flights.
Figure 2-4. The generic flight track design to obtain thermodynamic, high-resolution
Doppler radar, and microphysical data (Electra or P-3 aircraft, solid line). The generic
flight track design to obtain boundary layer height, moisture and wind (Fokker, broken
line)
Finally, a rapid scan strategy from METEOSAT 6 is planned by EUMETSAT allowing for a high-frequency monitoring of the MAP area in the visible, infrared and water vapour channels. Fully navigated and calibrated imagery will be available for quantitative studies on rainfall estimation, cloud classification and cloud radiative properties.
Introduction
A major aim of MAP is to prove that a combination of optimal analysis and sophisticated modelling can lead to reliable forecasts of the strength and location of severe precipitation episodes. Both ingredients require that sufficient consideration be given to tropopause-level processes which are at least as crucial as other atmospheric sub-systems. Western Europe is a preferred region for the occurrence of narrow meridionally elongated troughs, and these can extend, for example, equatorward from Britain towards the Iberian Peninsula (Holopainen and Rontu, 1981). These features have been termed "PV-streamers'' because they are characterized by distinctive structures in the potential vorticity (PV) field (Appenzeller and Davies, 1992). In addition to their synoptic scale pattern such streamers can possess a wealth of meso-scale structures, including a train of vortices aligned along the streamer (Appenzeller et al., 1996; Wirth et al., 1997).
This project is designed to examine the influence of a PV-streamer upon, and its interaction with, both weather and clear-air flow systems in the Alpine region. To this end it is conceptually convenient to distinguish two phases in a streamer's life cycle: (i) a pre-Alpine phase when the streamer is maturing west of and advecting toward the Alps; and (ii) an Alpine phase when the streamer encounters the Alps, and there is the potential for both orographic modification of the streamer and for the triggering or modulation of heavy precipitation over the south-western Alpine arc. This project's overall goal is to elucidate the linkages between PV-streamers and Alpine weather and flow. This requires the accurate determination of the structure, strength and evolution of a streamer in the pre-Alpine phase, and its orographic modification in the second phase.
Scientific Questions
2.1 The role of a PV-streamer as precursor of Alpine severe precipitation. The point is to investigate the streamer's contribution to heavy precipitation events (see e.g. Massacand et al., 1998). It includes the assessment of the sensitivity of the precipitation prediction to the accurate specification of the streamer in the initial field of the forecast.
2.2 The diabatic modification of a PV streamer. With the onset of heavy precipitation there is, in addition to the aforementioned linkages, a possible interaction between an approaching PV-streamer and the orographically bound PV-tower established by the sustained diabatic heating. The tower can induce a further distension of the PV-streamer and thereby favour the formation of a PV-cut-off.
2.3 The small-scale structures in stratospheric intrusions. Structures seen on water-vapour satellite images during the roll-up of a stratospheric streamer often, but not always, correspond well with tropopause-level flow structures. It has been hypothesized that differences might relate to the mesoscale patterns of vertical velocity. Sound understanding of these differences is a prerequisite for the correct interpretation of satellite water vapour images in terms of tropopause structures.
Observational Requirements
The horizontal resolution of current routine observational systems is too low to accurately describe the sub-structure of streamers. We propose to enhance the routine measurements by a combination of measurements by research aircraft and wind profilers.
Convenient aircraft measurements could be taken by the DLR Falcon. In order to cover questions 2.1 above, Fig. 2-6 shows possible flight tracks crossing the southern end of the PV-streamer. Dropsondes are needed between the flight track points 6 and 7. The best period for performing the flight pattern is 12-36 hours before a PV-streamer reaches the Alpine region. The measurements should include a line of drop sondes traversing the front, if possible several traverses, to determine the structure of the front and in situ upper level measurements to determine the position and structure of associated upper level features.
In relation to question 2.2, Fig. 2-7 shows a flight track crossing at various times the upper-level PV-streamer located at about 300 hPa above the Alpine region. Intense diabatic heating occurs at the southern side of the Alps. A second flight with the same pattern would be desirable at a higher or lower level.
Figure 2-5. Foreseen network of wind profilers for MAP.
In relation to question 2.3, the aircraft must cross the PV anomaly at two, or preferably three different levels (Fig. 2-8): (i) some two kilometers below the tropopause in order to measure water vapour, (ii) right at the tropopause level in order to get in situ data from the tropopause, and (iii) some two or three kilometers above the tropopause in order to release dropsondes. The release of dropsondes is desirable in order to obtain sound data for the meteorological background fields, while the first two crossings with in situ measurements will provide the high-resolution measurements of water vapour and potential vorticity. Simultaneous measurement of ozone is desirable as a check of the estimate of potential vorticity, since the latter is somewhat indirect. Profiles of water vapour through an airborne lidar should be taken.
On the other hand, the French experimental network of wind profilers (CNRS and CNRM) offers a unique opportunity to document the fine scale structure of the streamers as they approach the Alps, through high-frequency (15 minutes) and vertical resolution (375 m) profiles of the three wind components (e.g. Caccia and Cammas, 1998). Most of these devices will operate on their usual sites (La Ferte-Vidame, Toulon, Clermont-Ferrand, Saint-Michel de Provence). Two of them will be deployed at special locations (the CNRM profiler will be located close to the Lago Maggiore, and the LA profiler will be located near Annecy, France). See Fig. 2-5 for a general view of this network.
Figure 2-6.Flight pattern for scientific questions 2.1: horizontal (a) and vertical (b)
sketch.
Figure 2-7.Flight pattern for scientific question 2.2 (upper right panel).
Figure 2-8.Flight pattern for scientific question 2.3 (lower right panel).
Introduction
The accurate measurement of the rainfall input and of variables of the hydrological system such as soil moisture, water volume in reservoirs, snow depth, and river discharge is basic for real-time flood forecasting using coupled hydrological and meteorological models (Georgakakos and Foufoula-Georgiou, 1991). The measurement campaign intends to collect data on precipitation to assess the accuracy of NWP at space and time resolutions which are of interest for flood forecasting. These data will also be used as input into hydrological flood models (Obled et al., 1994). The value of precipitation measurements and snowpack at high altitudes, where the density of instruments is low will be assessed. The initial soil moisture conditions significantly influence the runoff volumes produced during a heavy precipitation event, but measurements are very difficult especially in rough topographic conditions.
Scientific Questions
3.1 Demonstrate in real-time or near real-time the forecasting capabilities of a hydrological flood model, forced by the special measurements of the SOP, or coupled with the advanced mesoscale atmospheric prediction models implemented at that time.
3.2 What is the role of some man-influenced initial conditions of the surface water system, such as the water storage in reservoirs, on the runoff generation during floods in mountainous regions?
3.3 Study the role of soil moisture conditions prior to flood events in determining the production of runoff and investigate the capabilities and limitations of some soil moisture monitoring techniques over rugged terrain.
Observational Requirements
Candidate areas for the observational campaign and the flood forecasts are the Ticino-Toce watershed (CH-I), in the core of the Lago Maggiore target area and the Ammer watershed (D), close to the Brenner Pass target area.
The operational experience and theoretical considerations based on the theory of spatial random processes, indicate that a rain gauge network connected in real-time having a density of 0.01-0.02 per km2 is needed to forecast with sufficient accuracy river discharges in watersheds of some thousands of square kilometres in size with a lead time of some hours. This space and time resolution is also compatible with the actual capabilities of meteorological models in predicting precipitation fields, so that such a raingauge density is needed for the verification of the NWP rainfall forecasts. Existing raingauge and radar networks will be improved by the installation of additional instruments in those sites where the density is insufficient. The use of vertically pointing X- and K-band radars installed in lateral valleys of the Ticino-Toce area will allow precipitation monitoring in sites where the use of the C- or S-band radars is not feasible or difficult due to ground clutter and orographic shielding. Portable raingauges are planned to be installed in those areas, especially at high altitudes, where the existing network density is smaller than 0.01 per km2. These measurement systems and the existing ones will be connected in real time, or in near real-time, to the POC and MOC to provide detailed precipitation fields to be used for flood forecasts during the IOP. The Hydrographic Services will provide streamgauge measurements to be used for flood forecasts and for the verification of NWP of precipitation volumes and rainfall measurements (see Fig. 2-9).
The initial conditions of a hydrological basin include also the water storage in reservoirs, which is a man-influenced state variable. Thus the importance of having access to, at least, aggregated information on reservoir volumes and operation rules will be assessed. The real-time access to reservoir regulation rules is not a standard practice in the complex alpine environment, but is useful for the efficient control of floods. Water volumes stored in reservoirs prior to the expected flood events as well as standard operational rules have to be known. Steps have already been taken to involve hydropower companies in the experiments in order to have access to water volumes and operational rules of reservoir regulation.
Figure 2-9. Ticino-Toce watersheds: Hydrometric stations (non real-time in white squares;
real-time in grey circles). (Source of coordinates: Swiss National Hydrological and
Geological Survey; Servizio Idrografico e Mareografico Nazionale; Settore Meteoidrografico
Regione Piemonte)
Field measurements of soil moisture are planned, possibly in connection with airborne microwave remote sensing to tackle the third scientific question. Spatial patterns of soil moisture fields have to be collected prior to expected flood events. During the SOP some open and flat areas with bare soil or short-cut vegetation cover will be selected for soil moisture measurements: gravimetric measurements of soil samples will be possible using a laboratory oven and the dielectric constant of soil (a function of the soil moisture content) will be measured using Time Domain Reflectometry (TDR) probes (Altese et al., 1996). The Bellinzona plane and the Toce valley, upstream of the Lago Maggiore, and flat areas in the Ammer watershed are candidate sites for those measurements. The sites will be selected also in coordination with the PBL Working Group which is planning to measure the turbulence structure and exchange processes in the same region. The field and laboratory measurements of soil moisture will be used as experimental evidence for the validation of the hydrological model simulations. These data would also be valuable for the calibration of soil moisture monitoring using an airborne antenna with 1.4 GHz (approximately) frequency (Ulaby et al., 1986; Pampaloni et al., 1990). Following a meteorological warning, 24-48 hours before the IOP, an airborne microwave sensor scan on the measurement fields and selected areas with similar characteristics could provide estimates of the soil moisture content prior to the expected flood events, to be used for the initialization of flood models.
Introduction
In contrast to studies of airflow over 2-D and 3-D obstacles, the problem of airflow over mountain passes is a largely neglected area of atmospheric research. However, it represents an important issue, as the dynamics of gap flow control the transport of airmasses, the generation of deep vertically propagating gravity waves and the possible generation of PV banners. Specific to the Alps, they control the flushing of polluted air from Alpine valleys and both shallow and deep Foehn events downstream.
Among the few existing investigations of "gap flow" is the report by Scorer (1952) on an easterly low-level jet of 100 knots through the Strait of Gibraltar when cold air fills the western Mediterranean. Other observational and modelling studies refer primarily to "outflows" from fjords and canyons (e.g. Jackson and Steyn, 1994; Levinson and Banta, 1995).
The flow beneath a strong low-level inversion may, in a first approximation, be studied using single layer reduced gravity hydraulics. High winds develop when the flow undergoes a hydraulic transition from sub-critical to super-critical flow. Such a transition can be induced by a pure horizontal contraction as in the Strait of Gibraltar or the combination of a horizontal contraction and a rise in the floor of the gap as in a mountain pass.
When the flowing layer is continuously stratified, as is the more characteristic case for atmospheric flows through mountain passes, the problem is considerably more complex. It's a fluid mechanics problem that is particularly interesting because, unlike a single layer flow, this flow has an infinite number of characteristic wave speeds, each associated with a different vertical structure.
Figure 2-10. Stratified flow through a channel with lateral contraction (see upper part of
the figure), after Armi and Williams, 1993. Note the formation of a low-level jet topped
by a stagnant mass downstream (left) of the narrowest section. The bright, near-horizontal
lines are constant density surfaces. The 4 near-vertical lines are velocity profiles
indicating self-similar flow. (They are formed by dropping dye into the channels).
Figure 2-11. Vertical cross-section of the stratified flow over a ridge showing
stagnant airmass over low-level jet (after Smith, 1985).
A series of experiments in a hydraulic channel and theoretical solutions discussed by Armi and Williams (1993) shed light on the mechanism of the continuously stratified hydraulic flow through a gap. Although the channel contains only a lateral contraction over level ground, the resulting flow (Fig. 2-10) has a striking similarity with that over a mountain ridge as derived, for example, by Smith, 1985 (Fig. 2-11). A low-level jet of supercritical speed forms downstream of the narrowest, section topped by a stagnant fluid mass overhead. For realistic flows over mountain passes the two models must be combined. As Fig. 2-10 indicates, the flow profiles induced by the contraction are "self-similar". Hydraulic jumps must be expected to form farther downstream.
Scientific Questions
4.1 To determine the relative importance of gap width versus terrain elevation changes along the floor axis on deep, continuously stratified flow through realistic topography.
4.2 A related question is to determine the relationship between the gap flow and the flow above mountain-top level; in particular, whether the gap flow is reinforced by flow aloft along the axis of the gap or by a mean-state critical level which caps the low-level cross mountain flow.
4.3 To study the vertical and cross-gap distribution of wind speed and thermodynamic properties. These are controlled by inviscid stratified dynamics together with surface friction along the valley floor and side walls. The frictional effects as well as dissipation and mixing of the low-level high speed flow need to be included in realistic models.
Observational Requirements
It is proposed to examine the flow over the Brenner Pass in the central eastern Alps. The Brenner represents the most pronounced and deepest gap through the Alps with an elevation of only 1,370 m, while neighbouring mountains reach heights of over 3,000 m. Cross-sections of the valley between Innsbruck and the pass, along the Wipptal, are plotted at several locations in Fig. 2-12.
Figure 2-12. View of several cross-sections of the Wipptal valley.
Figure 2-13. Schematic view of the deployment of additional instrumentation along the
Brenner cross section.
Automatic weather stations will continuously measure the mass and wind field along the valley floor and also vertically along the steep slopes of the valley. The vertical structure of the atmosphere will also be sampled by a Doppler sodar at the Brenner pass, and by radiosondes upstream and downstream of the pass. A chain of microbarographs along the axis of the valley will be used to identify the positions of hydraulic jumps.
The axis of the Wipptal is relatively straight, and as a consequence, the flow down the Wipptal is well-suited for study by Doppler lidar. The scanning Doppler lidar proposed in this study has been used successfully to reveal the detailed temporal and spatial structure of downslope winds, topographically channelled jets and mountain lee waves (Banta et al. 1995). Here it will be used to map the response of the along-valley velocity field to the changes in the width and elevation of the valley. In addition it can be used to study time varying processes associated with mixing of the strong flow that develops downstream.
The layers flowing through the pass will be probed by aircraft. The Merlin is particularly suited to the study of the low-level flow and the NCAR Electra and the NOAA P-3 to flows aloft. The downward looking lidar on the Electra can be used to map the low-level inversion. Clouds are likely to be present over the windward slopes in which case the upstream low-level flow field can also be efficiently mapped using the Doppler radars on the Electra and the P-3. The LEANDRE 2 Lidar on the French Fokker27 can also be used to obtain a cross section of the flow in the Brenner Pass.
Introduction
All prominent locations in the Alps with frequent Foehn conditions lie either in or close to a cross barrier valley which originates at a pass in the main crest. This means that the small scale (e.g. meso-gamma) structure of Foehn is strongly influenced by such valleys and the investigations during MAP should hence concentrate on the related processes. A better understanding, modelling, and forecasting of Foehn flow in large valleys is highly desirable for many reasons. Air traffic, especially slowly moving vehicles like paragliders, hang gliders, manned balloons, but also air liners may encounter dangerous conditions in areas of severe turbulence commonly observed in Foehn flows below crest height. Furthermore air pollution is very sensitive to Foehn conditions. A penetration of the Foehn down to the ground may bring a sudden relief to some valley segments whilst others remain within a shallow inversion without significant air mass exchange. The channelling and local forcing of Foehn flows in valleys may even cause damage to the human and natural environment by extreme wind gusts. Broken trees, blown off roofs and chimneys and even vehicles blown off the road are occasionally being reported. And finally the whole complex of biometeorological and -climatological impacts of Foehn necessitates a more detailed knowledge of its meso- and microscale structure.
Figure 2-14 Profile of the Alpine divide from the Colle di Cadibona, which is the border
to the Apennine, to the Danube at Vienna, which is the border to the Carpathians. The
Brenner Gap is indicated as BG and the relatively low ridge height in the area of the
origin of the Rhein and its tributaries is indicated as "Rhein window" (RW). The
prominent Alpine Passes are plotted as: CL: Col de Larche; GV: Col de Montgenevre; CE: Col
de l Echelle; ME: Col du Mont Cenis; PB: Col du Petit Saint Bernard; GB: Col du Grand
Saint Bernard (Grosser St. Bernhard Pass); SI: Simplon Pass (Passo Sempione); GT: Passo
del San Gottardo (St. Gotthard Pass); LK: Passo del Lucomagno (Lukmanier Pass); SB: Passo
del San Bernardino; ML: Passo del Maloja (Maloja Pass); OF: Passo dal Fuorn (Ofenpass);
RS: Reschenpass (Passo di Resia); BR: Brennerpass (Passo di Brennero); RT: Radstaetter
Tauern Pass; HT: Hoher Tauern Pass; SR: Schober Pass.
Figure 2-15 High resolution surface potential temperature distribution during a
Foehn event on December 18, 1997, 1200 UTC. The contour interval is 2 K with dashed 16 C
and lower isolines. The plot is based on a hand analysis of the station data (dots),
digitized with a 5 km regular grid spacing. The border between the warm Foehn air and the
shallow cold air north of the Alps is concentrated to a narrow band. Furthermore the
penetration of the cold blocked air through the lowest passes into the northern valleys is
very pronounced, e. g. in the Hinterrhein-valley (lower left side). Especially in areas of
low station density, the temperature distribution is purely hypothetic, taking into
account topographic features only.
Scientific Questions
5.1 The dynamics of that part of the blocked, potential cooler air mass, reaching typically up to mean crest height on the windward side of the main ridge, which is flowing through deep Alpine passes towards the lee-side valleys (shallow Foehn).
5.2 The interaction between low-level and mid-tropospheric Foehn flows on the scale of large alpine valleys.
5.3 The mechanism of temporal and spatial evolution and cessation of Foehn flows in complex valley systems on a local scale.
As several aspects require consideration of boundary layer processes, this project will be conducted in close cooperation with the PBL project P8 (see below).
Figure 2-16. Schematic tracks for the Merlin aircraft above the lower Rhine valley. Each
track will be flown at three different levels.
Observational Requirements and Measurement Strategies
The valley selected for this study should be such that segments of the main valley and/or junctions to side valleys exist which are parallel to the direction of the main crest. Several low passes in the main crest of the Alps should allow the development of shallow Foehn, see Fig. 2-14. The spatial extent of the target area should include the crest of the Alps and include parts of the Alpine forelands. The Swiss-Austrian-German part of the upper Rhine valley from the springs to the area north of Lake Constance has been proposed. In Fig. 2-15 this target area is shown with isotherms of the surface potential temperature of the operational meteorological network during an intense Foehn episode.
Various measurements are needed to capture the above mentioned phenomena. A dense network of surface stations (Mesonet for temperature, humidity, pressure, wind speed and velocity and a few for global radiation) must be set up along specific valleys and adjacent slopes and mountain tops to study the local Foehn penetration and retreat. Some surface energy balance stations are necessary to observe the behaviour of cold air pools and to test the role of the solar radiation on the Foehn penetration.
As the knowledge of the vertical structure is vital for the investigation of Foehn a dense profile of upper air stations should be positioned across the section of the Alps where the Foehn valley is located, including the windward side. The horizontal resolution shall be in the order of a few 10 km, the temporal resolution ideally quasi-continuous but not less than approximately 3 hours during IOPs. Besides RAWIN stations some sodars and especially UHF/VHF-wind-profilers (RASS) will be highly desirable to study the temporal development of the vertical structures with high accuracy. In areas of relatively weak winds (cold air pools) tethered balloon soundings can be used for the measurement of temperature, humidity and wind profiles whereas in regions with strong winds instrumented kites represent a valuable means of profiling. Video cameras will help to follow the behaviour of cold air pools, the top of which are usually being visible by haze or fog horizons.
During IOPs airborne measurements with high-level research aircraft should be carried out with dropsonde releases. The tracks should not only cover sections across the main ridge but also some along barrier segments and at different vertical levels to study the three dimensional structure of the flow. The CNRM Merlin, which can also operate in the valleys, should take in situ measurements of the lower level channelled Foehn flow. These will also be complemented by a Motorglider (Metair-Switzerland). The flight plan presented here includes a large axis oriented along the valley axis and four cross tracks. This flight plan will be performed at three or four different levels above the valley. They will have to cover various stages of the Foehn life cycle (formation, established Foehn, dissipation).
Finally, the LA constant level balloons will be used to reconstitute the flow trajectory at the mesoscale. Cross-alpine trajectories will be made possible by successive tracking of the balloons by LA and MeteoSwiss.
For a detection of the complete flow across or along a valley segment a Doppler-lidar with high spatial (1 m x 1 m x 300 m, typically) and temporal resolution seems to be a promising tool. The facilities under consideration are the scanning lidar of Laboratoire de Météorologie Dynamique and the upward pointing lidar of the Observatory of Neuchâtel.
Introduction
The basic theory for the vertical propagation of internal gravity waves forced by airflow over idealized two-dimensional topography has been well established for several decades (e.g. Smith, 1989). As these upward propagating mountain waves amplify, turbulent breakdown may occur. Recently, simple two- and three-dimensional models have been able to successfully simulate gravity-wave generation and evolution (Olafsson and Bougeault, 1996; SchŠr and Smith, 1993; SchŠr and Durran, 1997). However, it has not been conclusively demonstrated that numerical models can accurately predict the occurrence of wave breaking over three-dimensional complex topography, such as the Alps. In spite of significant advances in the last several decades in our understanding of airflow around and over topography, relatively little is known about three-dimensional wave breaking in nature.
One of the major factors limiting our understanding of the important wave breaking processes and characteristics can be attributed to the lack of observational data on the necessary temporal and spatial scales. As gravity waves amplify and wave-wave interference occurs between vertically propagating waves generated from complex topography, breaking may occur in localized patches and seemingly random episodes (Bacmeister and Schoeberl, 1989). These characteristics contribute to difficulties in measurement by conventional observational platforms and underscore the need for a systematic program of wave breaking observation and verification of numerical models.
Gravity-wave amplification and subsequent wave breakdown is known to play a critical role in several aspects of atmospheric science. The deposition of wave momentum flux associated with the breaking process has an important influence on the momentum budgets of the troposphere, stratosphere and mesosphere that requires proper parametrization in numerical weather prediction and climate models (Bessemoulin et al., 1993). Large-amplitude gravity waves and wave breaking have important implications for clear-air turbulence generation and the vertical transport of water, aerosols and chemical species. It is not known how reliably existing models treat such processes.
With the accomplishment of following scientific questions, our increased understanding of three-dimensional gravity wave breaking and wave drag will directly lead to necessary improvements in the parametrization of gravity wave drag effects in numerical weather prediction and climate models and to better warnings of CAT for aviation.
Scientific Questions
6.1 What is the predictability of large-amplitude gravity waves and wave breaking and the sensitivity to the synoptic-scale flow, critical levels, tropopause effects, diurnal cycle forcing, and 3-D aspects of the Alpine topography?
6.2 What is the vertical distribution of the two-dimensional momentum flux in the presence of strong cross Alpine flow and wave breaking?
6.3 What is the significance of PV generation in wave breaking?
6.4 What is the nature and amplitude of gravity waves that continue to propagate aloft following a wave breaking event?
6.5 What are the dominant scales over which scalars (such as ozone, water vapour or aerosols) are mixed during wave breaking events, and how efficient is the mixing?
6.6 What are the necessary combination of remote sensing and in situ techniques for the observation and measurement of gravity waves breaking over complex terrain, often associated with clear air turbulence?
Observational Requirements
The development of improved high-resolution models and new remote sensing techniques should make it possible in MAP to predict, observe and verify the occurrence of upper-level wave breaking. Most of the breaking of interest to MAP will occur in the layer extending from the middle troposphere up to the lower stratosphere at altitudes between 5 and 20 kilometres. Two types of observational platforms are suited to observe this phenomenon: high-flying aircraft and ground-based platforms (VHF radar and lidar). The DLR Falcon can be equipped with a Doppler lidar system WIND (under development; most likely looking downward), and an upward or downward pointing differential absorption lidar (DIAL) tuned for water vapour, ozone measurements or light-weight dropsondes with GPS (a joint venture between DLR, NCAR, and Vaisala). It is currently under investigation whether DIAL and dropsondes can be made available for the same mission. The WIND lidar system can be used to document the vertical distribution of the two-dimensional momentum flux during strong cross mountain flow and wave breaking. The dropsondes can be used with 25 km spacing to depict the three-dimensional mesoscale structure of the wave breaking regions in the middle to upper troposphere as well as the upstream conditions. Additionally, in-situ flight data will be important to provide further depiction of the local conditions associated with wave breaking including turbulence measurements.
In certain flow situations, there may be strong mountain wave amplification or wave breaking in the middle and upper troposphere that could be observed with the NCAR Electra. The Electra, operating by itself or with another aircraft further aloft, could gather flight level and lidar data using the SABL system, which would test model predictions of tropospheric wave breaking. These missions should be designed flexibly to respond to forecasts of wave breaking and to maximize coordination with PV banner measurement missions. Schematic flight track segments, with the purpose of identifying intense breaking regions, can be performed with an initial leg (about 150 km) perpendicular to the flow just downwind of the main Alpine ridge, as depicted in Fig. 2-17 a. Flight tracks can be performed parallel to the flow across the mountain ridge dissecting the three dimensional breaking area (Fig. 2-17 a). The Electra, flying stacks at low levels, can be coordinated with the DLR Falcon to perform simultaneous stacks in the upper troposphere and lower stratosphere enabling a more complete description of the structure of the breaking region. Multiple levels of wave breaking may occur as suggested in the numerical simulation results in Fig. 2-17 b. Under some conditions, gravity wave breaking and PV banner measurement missions can be combined into a single flight track constructed along a rectangular path in a box that will sample the low-level breaking and presumably PV-generation regions, as well as the topographically forced downstream effects.
 |
 |
Figure 2-17. Numerical simulation of wave breaking using NRL's COAMPS (horizontal resolution of 5 km with 60 vertical levels) for a southerly Foehn case valid at 1200 UTC 12 January 1996 (24 h simulation time). (a) 10-km Richardson number (color shading less than 1.0) and (b) vertical section of cross mountain wind (color; m/s) and potential temperature (K) along the axis A-B displayed in (a). Commercial aircraft reports of moderate turbulence (L) between 11-15 UTC for flight levels 8000-9500 m are displayed in (a). Terrain elevation of 1000 m (green contour) and 2000 m (bold green contour) are shown in (a). Schematic flight transects through the wave breaking regions are shown by the black lines.
Ground-based wind profilers and scanning Doppler lidars could also be used to measure wave breaking activity if suitable concentrations of aerosols exist in the upper troposphere and stratosphere. However, gravity wave breaking regions are thought to be transient and geographically varying. Although no specific target area on the ground can be identified for these investigations, promising areas of wave breaking are the climatologically well established south-Foehn and lee wave areas over Switzerland and Austria (e.g., see Fig. 2-17 a). Corresponding areas also exist on the south side of the Alps for North Foehn flows. Additionally, aircraft turbulence climatological data may enable identification of preferred regions of wave breaking that can be used for flight track strategies and placements of supplemental radiosonde observations, which may include multiple ascents.
The properties of gravity wave breaking in the upper-troposphere and stratosphere are influenced by ambient conditions associated with broad-band travelling wave breaking and jet stream turbulence over the Alps that are not directly connected with orographically forced waves. We anticipate in a typical flight encountering both breaking and non-breaking regions that should enable the ambient conditions to be characterized. Fixed platform instruments (i.e. ground-based VHF radars or lidar) may help define the climatology of turbulence in specific locations over the Alps. These instruments could be measuring the turbulence conditions well in advance of the MAP field phase so that normal levels of activity (as function of wind direction and speed) are well known.
Guidance from high-resolution numerical models to indicate likely wave breaking regions in real-time will be essential for a successful measurement effort. An example of a numerical simulation, performed using NRL's COAMPS, of gravity wave breaking during the south-Foehn event of 12 January 1996 is shown in Fig. 2-17. The simulated wave breaking regions at 10 km are nearly coincident with several commercial aircraft turbulence reports. These results and other recent simulations (e.g., DLR using MM5 see e.g. Dšrnbrack et al., 1998, LA and CNRM using MESO-NH, and EC-RPN using MC2) indicate promise for the capability of numerical models to predict the occurrence of breaking and to identify favourable breaking regions in real time in order to design aircraft observational flight strategies. Further confidence in these model simulations must be established in the "pre-MAP" period through a vigorous program of model intercomparison and development of diagnostic techniques for wave breaking. A hierarchy of high-resolution model intercomparison studies ranging from simple two- and three-dimensional idealized flows over realistic Alpine topography to a selection of relevant real-data cases involving upper-level wave breaking should be performed. These simulations should be run using research and operational mesoscale models, including those targeted for real-time forecasts during MAP, having mesh sizes of 15 km or less. Aircraft turbulence reports or occasional radiosondes ascents that penetrate breaking regions in these real-data cases presumably will be sparse and incomplete, making model validation difficult. However, the model intercomparisons should lead to an improved understanding of the mesoscale predictive skill as well as characteristics of the wave breaking, such as the temporal and spatial scales and preferred locations relative to local topographic features. Careful evaluation of model output will be required to identify variables (e.g., turbulent kinetic energy, momentum flux profiles, winds) that are best suited to depict three-dimensional wave breaking regions in real time. The results from the numerical simulations will lead to further refinement of our wave breaking hypotheses, conceptualizations and experimental design, including flight observation strategies.
Introduction
Idealized and real-case numerical experiments suggest that flow past three- dimensional topography is often accompanied by the formation of elongated filaments of vorticity and potential vorticity (so-called PV banners or shearlines) which trail downstream from topographic obstacles (Smolarkiewicz and Rotunno 1989, SchŠr and Durran 1997, Smith et al. 1997). These results are consistent with earlier theoretical ideas about stratified flow past isolated mountains (Smith 1989). The PV banners usually occur in pairs of positive and negative filaments, and their formation is associated with irreversible topographic processes such as flow-splitting, boundary layer separation and gravity-wave breaking. Analysis of operational high-resolution NWP runs over the Alps suggest that PV banners are abundant and occur whenever there is appreciable flow past the Alps (Aebischer and SchŠr 1998). The banners often attain a length of several hundred kilometers, and extend on occasion as far north as Hamburg (see Fig. 2-18).
The presence of PV banners is significant for several reasons. First, the PV banners can be considered as a key-ingredient of the wake which is related to a range of mesoscale processes and features (i.e. flow splitting, boundary-layer separation, gravity-wave breaking, wake stability, vortex shedding, dissipation, drag, boundary layer processes). Second, the PV invertibility principle implies that the PV anomalies are Ð together with surface thermal anomalies Ð the key-elements of the topographic interaction with the larger-scale approximately balanced flow. Such interactions can take the form of interaction with synoptic scale circulation anomalies (e.g. lee cyclogenesis) but are also relevant for the deceleration of the mid-latitude westerly flow (e.g. orographic drag).
Figure 2-18. View of model generated PV banners in northerly (top panel) and southerly
(bottom panel) flow past the Alps. The diagrams depict the wind vectors and the PV
distribution on the 850 hPa level (from Aebischer and Schär).
Over the last three years, as NWP models have increased their spatial resolution, PV banners have been routinely simulated. Satellite observations of island wakes have begun to confirm these model predictions, but PV banners have never been systematically studied in major mountain ranges. Associated with the lack of appropriate computational resolution and operational data coverage, the value of current and future high-resolution numerical model results is thus difficult to assess.
Scientific Questions
7.1 What is the high-resolution structure of the Alpine wake? What is the width of the PV banners? What is their vertical extent and downstream development?
7.2 Is the Alpine wake quasi steady as suggested by current numerical models, or can it on occasion become turbulent? What determines the banner's stability with respect to barotropic shedding instabilities, and with respect to inertial instability in regions of negative PV? What is the statistical nature of the turbulence associated with the banners?
7.3 Which processes are responsible for the generation of the banners? How do the banners interact with the planetary boundary layer, and with meso- and synoptic scale flow features?
7.4 Do current high-resolution atmospheric models appropriately predict PV banners and their associated three-dimensional circulations? Can we use model results to resolve scientific issues related to the structure and generation of PV banners?
Observational Requirements
Alpine PV banners occur during a wide range of weather situations, including in particular deep southerly flow (South Foehn) and strong northerly or north-westerly flow (Mistral and North Foehn). The frequency of both these circulation patterns has two maxima in the seasonal cycle, one in spring and the other in autumn. For the investigation of PV banners within MAP, priority should be given to northerly and north-westerly flow situations for the following reasons: First, for these wind directions the upstream conditions are well monitored by operational soundings over the European continent. Thus it will not be necessary to probe the upstream flow by additional observations. Second, the PV banner to the south-western Alpine tip (separating the Mistral to the west from the wake to the east) appears to be the most pronounced and most frequent of its kind, and appears to interact on occasions with Alpine lee cyclogenesis. This PV banner might thus be of direct interest and relevance to weather forecasting in the region. Third, from a practical point of view within MAP, the focus on northerly flow will reduce the competition for observing facilities since most of the other questions are related to southerly flow. Nevertheless, should facilities become available during episodes of deep southerly flow, the associated PV banners are equally attractive for investigation.
During the field phase, background and upstream coverage will be provided through operational soundings, profilers and surface networks running at increased temporal resolution. The current layout of the airborne observations is based on the NCAR Electra, but the NOAA P-3 could be utilized in a similar fashion, or could be used to complement the survey. The CNRM Merlin could be used for in situ measurements within the banners, and the LEANDRE 2 lidar on the French Fokker27 could be used to map the differences in boundary layer height associated with the banners. The use of airborne observing systems in conjunction with the high-resolution background coverage present in the Alpine region will substantially contribute towards answering the scientific questions.
Figure 2-19. Schematic flight tracks for the PV-banner project on (a) the meso-beta-scale
and (b) the meso-gamma-scale.
The two flight tracks in Fig. 2-19 are generic examples of flight tracks to be designed in detail at a later time. They relate to meso-beta and meso- gamma-scale aspects of the flow, respectively. The meso-beta-scale track is at one single mid-tropospheric level, while the meso-gamma-scale track contains low-level trajectories at several flight levels. The two flight- paths require about 3 h of flight time each, and could be flown on the same mission. The flight paths will have to be adjusted before takeoff by using information from numerical model results and available observations. The aircraft would be instrumented as follows:
The refinement of the observational strategy will try to exploit synergies with other MAP projects. For instance, to the extent that some of the PV banners might be associated or occur simultaneously with tropospheric gravity wave breaking, coordination with that project is particularly attractive. Similarly, during episodes of deep southerly flow, the experimental activities related to PV banners might provide some useful background information for the questions addressed in sections P4 and P5 of this document. There will also be intensive interaction with the numerical modelling groups, since the detailed planning of the flight track will be guided by real-time numerical forecasts.
The boundary layer subprogramme is intended to answer some basic scientific question on the structure of the low-level flow and of turbulent exchanges over the highly complex alpine terrain. As such, it has its own objectives. However, it is also considered to contribute to the two primary scientific objectives of MAP, by supplying key measurements of turbulent fluxes of moisture for the precipitation objective, and turbulent fluxes of momentum for the dry dynamics objective.
Introduction
Very little is known at present concerning the turbulence structure and exchange processes in the boundary layer (BL) over mountainous terrain. In fact, the sparse available observations are point measurements of little general meaning while numerical models use parametrizations which are based on flat, homogeneous terrain. We may assume that the modification of the BL turbulence structure as compared to flat homogenous terrain is determined by the underlying topography. Therefore, various specific entities of the topography such as 'a valley' (e.g. u-shaped or v-shaped) should be investigated in order to identify the basic features of these modifications.
Although there is some theoretical understanding on the flow over/around low gentle hills or through idealized valleys, there is a need to extend these ideas to much steeper and rougher mountainous terrain. The starting point of any data analysis may then be to establish the main differences between such idealized theories and the observations, i.e. the influence of the complex orography.
Due to mechanically and thermally enhanced vertical motions as well as to advective enhancement, the net (effective) vertical transport is expected to be greatly modified (often: larger) with respect to flat terrain (under comparable atmospheric conditions).
It should also be kept in mind that the interaction of the BL and the free troposphere is determined to a large extent by the dynamics of the large scale flow and only partially by the characteristics of the local surface.
Scientific Questions
8.1 Development of mixing height (including its definition) in the Alpine region as well as upstream and downstream of the Alps. Determination of the vertical extent of the BL and the spatial/temporal distribution of turbulence statistics within the BL.
8.2 Description, characteristics and parametrization of effective vertical turbulent fluxes of momentum, heat, moisture, and other atmospheric constituents over complex terrain. Note that the turbulent moisture fluxes define an interrelation to hydrology. The knowledge of physically consistent area-averaged turbulent fluxes is mandatory for weather prediction models. Furthermore, the interaction of mesoscale flow with the alpine complex orography and the thermal forcing must be investigated in order to better understand the structure and organization of convective systems and heavy precipitation.
8.3 Interaction between local winds and (valley) BLs/ Erosion of (valley) BLs by Foehn. Ambient winds have an effect on the development of thermally forced local wind systems. Their influence is certainly instrumental and depends on the characteristics (stability, forcing, ...) of both systems.
8.4 Exchange of air masses and atmospheric constituents between the BL and lower free troposphere. The preceding points highlight possible modifications of the BL over complex terrain as compared to flat, homogeneous surfaces. Consequently, also the exchange processes at the BL's top are likely to be enhanced (or at least different) with respect to current understanding.
8.5 The impact of boundary layer, through momentum, moisture and heat exchanges, on the formation of heavy orographic precipitation, distribution of precipitation particles, upstream flow blocking, lower tropospheric wave breaking, and PV banners.
Observational Requirements and Measurement Strategies
As essentially no reliable turbulence data for mountainous terrain can be found in the literature, the observations will have to be as complete as possible with respect to spatial and vertical resolution. Different examples of, e.g the topographic entity 'valley' should be probed: varying shapes, orientation, aspect ratio, etc. In agreement with the various target areas of MAP, it is likely that at least two different valleys can be included in the measurement plan: one valley north of the Alps and one valley within the Ticino basin in connection with the hydrological project.
This will be achieved by close coordination with project P5 on the Rhine valley (for the northern side), and by special measurements on the Riviera valley, situated within the Lago Maggiore target area (for the southern side).
Observations at various locations within a given topographic entity will have to include a) mean variables (wind vector, temperature, humidity, pressure), b) turbulence statistics including vertical turbulent fluxes (and in particular surface fluxes) and c) the mean state of the lower free troposphere in the vicinity of the topographic feature under consideration.
The observational strategy will include the following measurements:
A) near surface data:
i) A cross section of surface-based towers through the valley with various levels of
instruments (mean variables and turbulence statistics). These observations will be
conducted continuously during the SOP.
ii) Scintillometry for average fluxes (or statistics) over a certain distance, along
topography or cross-wise (by episodes only).
B) Valley atmosphere/ free troposphere:
i) aircraft, equipped with instruments for in situ measurement of turbulence/ mean
variables. Flight legs within the valley (along valley transects at different heights and
both sides) and crossing the valley/free atmosphere intersect. This type of flight pattern
requires a small, light aircraft (such as the motorglider of METAIR). By episodes only.
ii) aircraft with airborne DIAL. 'Instantaneous' cross-section, e.g. along a valley. By
episodes only.
iii) tethered balloons and radio soundings will probe regularly the mean vertical
structure (e.g., every three hours, for certain episodes).
iv) sodar/lidar/radar. Vertical profiles at different locations. Regularly or episodes.
All the episodic measurements (such as aircraft, soundings, ...) will be scheduled simultaneously and measurement locations will be chosen with respect to an optimal spatial representation and sufficient overlap for intercomparison.
