Figure 3-1. Timing of the phase of MAP.
The Mesoscale Alpine Programme is structured in three phases (see Fig. 3-1). Phase I of the programme is dedicated to modelling activities, theoretical studies, improvement of existing Alpine climatologies, and to the preparation of the field phase. A distinctive feature of the programme is that a number of these activities will continue throughout its duration. Phase II is the field phase referred to as General Observing Period (GOP). It is a 13 months period where the MAP Data Centre devotes maximum effort to collect data in near real-time without additional measurements. For most of its duration, it will be primarily based on operational observing systems, but during a 3-month Special Observing Period (SOP), dedicated observing systems such as research aircraft and temporarily installed observational platforms will be deployed, and activated during a number of Intensive Observing Periods (IOPs). In Phase III of the programme the field phase results will be analysed and used for modelling studies. This scheme assures that
In the following subsections further information is provided on the studies and activities to be undertaken in the respective phases of MAP.
Figure 3-1. Timing of the phase of MAP.
The last decade has seen the development of many high-resolution numerical weather prediction models (see overview in Wergen and Majewski 1993, du Vachat 1994). Presently, at least four centres in Europe provide operational NWP models for the Alpine region resolving the meso-b scale with mesh sizes between 15 and 35km (i.e. Deutscher Wetterdienst, Météo France, Regional Meteorological Service Bologna and MeteoSwiss). Moreover research models with even much higher resolutions are now being tested.
The MAP goals for the numerical experimentation during Phase I are the following:
To achieve these goals, the following steps are planned:
As a prerequisite for high-resolution numerical modelling, a high-resolution (~1km) physiographical data set including orography, land-sea mask, soil parameters, albedo, and preferably some description of vegetation covering the entire Alpine region need to be available for all groups involved in MAP. MAP seeks to coordinate and facilitate the use of such data for numerical experimentation. As a first step, the participating groups should make available their own high-resolution topographical data sets covering different parts of the Alps. These data can then be combined to derive a unified data set for the whole region to be used by all groups. A data set including topography and land-sea mask at a resolution of 1km and covering the whole Alpine region exists at the UK Met. Office.
Satellite observations can be used to determine some of the additional external parameter fields (vegetation, snow cover, etc.) required for the specification of the lower boundary condition in numerical models.
The MAP Data Centre (see Section 4)
will be responsible for the collection of the data sets and making
them available to the MAP community.
To meet the hydrological requirements of MAP several research groups will have to take up coordinated efforts in model experiments and development. It is suggested to delineate the hydrological parts of the modelling in a first step to the river basins in the southern and central Alps. The participating research groups will apply their models each to particular basins as parts of the whole region.
In order to establish and improve the coupling of hydrological and atmospheric models and to provide hydrological model components for the adequate representation of the land surface and sub-surface water processes the following steps have to be taken in Phase I:
The present composition of hydrological research groups contributing to MAP has expertise in distributed catchment modelling of small to large river basins, flood forecasting, modelling of soil, snow-cover and vegetation effects, and experience with various remote sensing techniques.
In Phase I of MAP a climatological survey is needed for the detailed planning of the field phase. The survey will aid the definition of location, time and duration of the Special Observing Period. This task requires statistics of the frequency of occurrence, duration and spatial extension of the mesoscale features relevant for the programme, and should provide average values and variability on a monthly base. Typical examples of statistics (e.g. for cold fronts) compiled in advance of a field phase are published by Monk (1986, `FRONTS 87') and Hoinka (1985, `German Fronts Experiment 1987').
Furthermore, statistics of meteorological parameters are required in order to describe the variability, intensity and other important characteristics of the phenomena under consideration. A desirable objective is to identify within the sample of available parameters, those key parameters that play a dominant role in the relevant mesoscale process.
In addition there is a need for statistics that describe the link between different scales, from the small scale up to the synoptic scale. Numerical models can help to supply a climatology of particular mesoscale situations, e.g. events favourable to wave breaking. The resulting statistics can describe the link between the synoptic scale situation and the smaller-scale phenomenon and can be very helpful in guiding the planning of intense measurements during the field phase.This is particularly true for phenomena which are not currently detectable with the routine observational networks. Wave breaking is a striking example of such a flow feature. Numerical simulations (Scinocca and Peltier 1989), observations (Smith 1987) and measurements (Neiman et al. 1988) have demonstrated the importance of wave breaking. Indeed lack of routine observations implies that climatological statistics are required of their temporal and spatial occurrence.
Several mesoscale features such as gravity wave breaking and shear lines elude detection by the conventional observing network. During Phase I, numerical simulations should be used to establish preferred regions in the Alps where these phenomena can be found in order to guide the installation of wind profilers or other remote sensing systems.
For the first time, the inner-Alpine atmosphere will be probed regularly with the new radiosonde station in Innsbruck. Data from wind profilers and the new radiosonde should be evaluated in Phase I to aid in
The testing of new airborne remote sensing systems over the Alps should be encouraged during Phase I. The results of such tests will be useful in planning the field phase.
A working group of operational forecasters of all Alpine countries
has been initiated in order to prepare them as specialists of
MAP-related subjects. Interaction of MAP research scientists with
operational forecasters will be of mutual benefit. It will serve
to pinpoint the key forecasting issues, facilitate the introduction
of better forecasting methods, and help establish decision trees
for the forecasting issues of the field phase. The fruitful way
to initiate contacts is the distribution of a questionnaire on
specific difficulties of analysis and/or prognosis of Alpine events
such as fronts, deep convection, Foehn events (see e.g. Hoinka
and Smith 1986). A similar action has been taken by MAP. To maintain
contacts and to train a group of motivated forecasters for the
field phase activities it is possible to organize regular workshops
where case studies of typical MAP events can be investigated and
discussed. Those forecasters who develop special links to the
project should ideally be freed of other operational duties during
the field phase to enable them to focus on MAP-related forecasting
tasks and to interact with the research scientists.
The MAP field experiment will take place during a 13-month period, (the so-called General Observing Period, GOP) during which the normal station network will be upgraded and mesoscale forecasting efforts will be intensified. This period encompasses an entire annual cycle in which the seasonal variability of Alpine weather systems can be investigated and includes an extra month at the beginning of the period to solve operational problems and assure that all routine data collection systems are fully operational.
Within the GOP, the 3-month period from mid August to mid November is designated as the Special Observing Period (SOP) for which the time and space resolution of the routine observational networks will be enhanced. This 3-month period appears climatologically to be the most suitable for measurements documenting the phenomena related to the primary objectives of the programme, i.e. convection, gravity waves and Foehn. The early part of this period usually has deep convection events over the northern and central Alps, and is followed by a period with typically a number of Foehn events, along with significant convection and upslope precipitation events on the southern side of the Alps (cf. the cases cited in Section 2.1). Specific climatological analyses will be carried out during Phase I of MAP to determine the final dates of the SOP.
Within the SOP a number of Intensive Observing Periods IOPs, say 10 to 15 each of 1 to 3 days duration, will be defined. During these IOPs, whose selection depend on the actual meteorological situation, all available measuring platforms, including research aircraft will be deployed.
Phase I studies will help to define the optimal experimental target areas and to refine suitable observational strategies. It is anticipated that the main study area for precipitation and convection will be located on the southern side and slope of the Alps, whereas gravity wave and Foehn phenomena will be investigated on the northern side of the Alps. For ground-based observing systems fixed target areas on both sides of the Alps as well as a number of Alpine transects are envisaged. The missions of aircraft and mobile ground-based measuring platforms will be designed to complement the fixed installations. Nevertheless, aircraft and mobile stations offer the opportunity to significantly extend the domain of action.
For the field experiment MAP aims at an observational strategy that makes optimal use of an efficient and economical combination of operationally-available measurements and a set of special equipment. The complete observing system constitutes in essence a composite system of three components:
Figure 3-2. Schematic view of the composite observing systems during the Special Observing Pe
riod (SOP).
A sketch of this composite observing system is provided in Fig. 3-2 where an imaginary north-south cross-section perpendicular to the Alpine chain is depicted. Convection on the southern side of the Alps is probed by aircraft equipped with Doppler radar (box-pattern flight track), ground-based radar, radiosoundings and light-weight dropsondes. Prior to the major convective event, the low-level meso-b-scale and upstream environment will have been observed by mesonetworks including Doppler sodar, radiosondes, wind profilers and low-flying aircraft. River runoff measurements provide integral and filtered information on the precipitation input into the hydrological catchment.
Lee-side flow characteristics are investigated by aircraft at lower levels (zig-zag pattern), mesonets, wind profilers, Doppler sodars and curtains of light-weight dropsondes released from high-flying aircraft. Gravity waves are observed by ST-radars, aircraft and Doppler lidar mounted on aircraft at upper-level. This is just a rough guide of the major systems and more detailed account is given in the following sections.
The finer details of the observing system depends upon several factors: decisions by the participating nations; competing field projects; funding situations; technological developments; and a refinement of the project requirements. However, already at this point in time it appears reasonable to have a station layout comprising of both several transects and chain-like set-ups parallel to the main Alpine ridge. Carefully selected areas will have to be especially heavily instrumented in order to yield the data necessary for tackling the objectives.
A. Mesoscale Environment
In comparison to the other major mountain ranges of the earth, the Alpine area is covered with a relatively dense routine meteorological observational network. Figure 3-3 shows the operational radiosonde and weather radar network. Recently the upper-air network has been complemented by the radiosonde station at Innsbruck.
Nevertheless, the present density of these networks is far from adequate to describe the meso-b-scale environment above and around the Alps (in contrast to the operational meso-b-scale numerical weather prediction models!). For this purpose substantial temporary enhancements of the upper-air network will have to be realized during the field phase. As an example of such an exercise the set-up of a forerunner field experiment `MiniMAP' is shown (see detail in Fig. 3-3, Richner 1994). All supplementary upper-air observations will be fed into the GTS in order to enable their utilization in real-time numerical modelling, in particular in the assimilation cycle of ECMWF.
In addition, there exists a dense network of automatic surface stations reporting a rich set of meteorological parameters at 10 minute intervals (Fig. 3-4).
B. Orographically Influenced Precipitation
Precipitation measured operationally by surface rain gauges is the meteorological quantity which is observed with the highest spatial resolution in the Alpine area (Fig. 3-7). However, the high-resolution measurements are only available as daily precipitation sums. Furthermore, these observations are usually not readily accessible beyond the boundaries of the individual Alpine countries or even provinces, and certainly not in real time. In addition to the surface rain gauge network there also exists a number of ground-based weather radar stations (some of them with Doppler capabilities). However they do not constitute a real network, because the data exchange and integration has yet to be fully realized.
Figure 3-3. Networks of radiosonde (stars) and weather
radar stations (open triangles) in the Alpine region. The shaded
area depicts the mesoscale array of sounding stations (filled
circles) operated during MiniMAP (Richner 1994).
Modern operational mesoscale numerical prediction models with a horizontal resolution on the order of 10km call for a comparable spacing of observations for verification purposes (Binder 1992). Furthermore, process studies of orographic precipitation enhancement need observational data on a similar or even more refined scale. Adequate temporal resolution is on the order of an hour to a few hours.
For the field phase (GOP) it is essential to integrate the existing precipitation observing networks, gauges and weather radar, that include numerous non-GTS surface stations. Recognizing the inability of covering the entire Alpine region with a surface network of the required resolution, it is necessary to select target areas for the installation of mesonets. For the investigation of extreme gradients of orographically influenced precipitation (altitude dependency, shielding effects) it is appropriate to establish a small number of densely instrumented Alpine transects. Radar complements the high-quality surface point measurements from rain gauges by providing area-covering observations with high spatial and temporal resolution. However use of radar data requires consideration of shielding by orography, ground clutter effects as well as the height-dependency of the radar reflectivity profile (Joss and Waldvogel 1990; Joss and Pittini 1991).
Figure 3-4. Network of automatic surface stations in
Austria and Switzerland. Other Alpine countries also operate automatic
surface networks. Open circle: station height below 1000m; filled
quadrangle: station height between 1000m and 2000m; open triangle:
station height above 2000m.
C. SOP Observing Systems for Objective 1 (Convection)
The MAP proposes to investigate several aspects related to convection and these pose different observational requirements. The bulk effect of convection involves a transition from the pre- to the post-convective atmospheric state (vertical thermodynamic, humidity and wind profiles). There are indications that these atmospheric characteristics depend to some extent on environmental conditions such as the surface texture and state, latitude and the terrain geometry (e.g. Binder 1990). Determination of the atmospheric state can be achieved by radiosonde ascents, dropsondes and wind profilers.
A careful description of the planetary boundary layer prior to convective activity is needed to contribute to the identification of the trigger mechanisms for convection. Better knowledge of these mechanisms would help to improve their description in numerical prediction models. At present the representation of the initiation of convection originating in the boundary layer is quite crudely and schematically described, although model predictions prove to be sensitive to these processes (Kain and Fritsch 1991). To this end, surface mesonets, clear-air measurements with Doppler radar and wind profilers, arrays of Doppler sodar and low-level research flights can all yield useful observational data. With the exception of low-flying aircraft these facilities are also appropriate for observing the progression of gust fronts and downdraft outflows of convective systems.
Valuable information on the development and propagation of convective systems is provided by Doppler radar devices. The airborne instrument ELDORA is a particularly notable observation system. Remote sensing techniques, both airborne and ground-based, can contribute to the determination of the heat, moisture and momentum budgets of convective systems as well as to the detection of convective-scale structures, e.g. low-level convergence, high-level divergence, up- and downdrafts, and mesocyclones associated with supercell storms.
Figure 3-5. Rain gauge network in the Alpine region. Dots indicate locations of stations with reg
ular reports of daily precipitation. The display currently comprises a total of about 5000 observ
ing stations from various national institutions. Contacts to additional institutions are underway
to further improve the data coverage. The topography is shown at an altitude of 800 m. The pre
cipitation data will become available to MAP scientists for selected periods through the MAP data
center. (Courtesy of Christoph Frei, Atmospheric Science ETH)
In-situ measurements by research aircraft are required to elucidate the microphysical and electrical processes associated with convective clouds. The armoured T28 aircraft of the South Dakota School of Mines and Technology operated in Switzerland in 1982/83 (Waldvogel et al. 1987), and its use enables the acquisition of in-situ measurements even in the core of deep convective clouds.
For studies of deep convection there is a need to explore the potential of combined measuring systems early on in the preparatory phase of MAP. Examples of field programmes of deep convection include the CLEOPATRA experiment (carried out in southern Germany in summer 1992, Meischner et al. 1993) and the SETEX experiment (initiated in 1994 in southern Germany and northern Switzerland). Both programmes led to a close cooperation of German and Swiss scientists (Haase-Straub et al. 1994). These and similar field experiments can be viewed as preparatory activities for the intense observing periods of MAP.
D. SOP Observing Systems for Objective 2 (Gravity Waves and Foehn)
To observe the characteristics of Foehn flow in vertical cross-sections, upper-level aircraft traverses will be utilized. The aircraft should be equipped with a light-weight dropsonde system in order to construct sections from dropsonde curtains. Also an upward pointing lidar device will be of particular value in detecting patches of gravity wave breaking above flight levels. Across- and along-barrier flight tracks at different heights are needed to describe the three-dimensional structure of these phenomena (Fig. 3-6)
To observe the characteristics of Foehn flow at lower levels and its interaction with the planetary boundary layer, suitable instrumentation are wind profilers (1000MHz; 400MHz), preferably equipped with radio acoustic sounding systems (RASS), and Doppler sodar instruments. In addition, a cross-Alpine transect of surface stations, including microbarographs, has to be established. Experience from earlier projects (cf. Davies and Phillips 1985; Richner 1987; Bessemoulin et al. 1993) will facilitate the design of such a network.
The three-dimensional structure of the lower-level wake-flow (shear lines, convergence lines, turbulence characteristics) can be probed by zig-zag research aircraft patterns, much like the strategy followed in the wake-subprogramme of the Hawaiian Rainband Project (Smith and Grubisic 1993). In addition to in situ registered quantities, remote sensing techniques like sideward pointing lidar devices open new perspectives.
Upstream and downstream conditions of cross-barrier flow are to be sampled by an enhanced network of radiosonde stations, as described in sub-section A above.
E. Other Observing Systems
In addition to the observing systems immediately related to the main objectives of MAP, it is planned to install further systems that will be deployed by individual participating groups. In this section a list is provided of such possibilities.
Research aircraft can measure profiles of mean and turbulent fluxes of mass, momentum, and heat above the Alpine topography and thereby support observational and modelling studies of the evolving Alpine boundary layer. Aircraft with differential Global Positioning System (GPS) capabilities are now thought to provide suitable positioning accuracy to make direct measurements of the horizontal pressure gradients that are responsible for driving synoptic and inner-Alpine wind systems, including even the thermally-driven wind systems.
Lidar-equipped aircraft could be used to support the evaluation of hypotheses regarding the vertical transport and diffusion of aerosols, water vapour and ozone in the Alpine atmosphere. Hypotheses regarding the transport and diffusion of air pollutants within and across the Alpine chain can be evaluated also using regional and local scale air motion tracers (perfluorocarbons and sulfurhexafluoride).
Other networks of instruments can be helpful in meeting the objectives of the planetary boundary layer task including Doppler sodar networks, radar profiler/RASS networks within and surrounding the Alps, mesonets, surface flux stations, slope profiles, and tethered, pilot, and manned balloon sounding systems.
Frequent (3h), narrowly spaced (~50km) ground-released radiosonde ascents and dropsonde curtains from aircraft as well as aircraft flying at about tropopause height level are best suited for observing upper-level features.
For the study of additional low level features such as Bora, Mistral, etc., the special equipment of different research institutions should be combined in order to meet the high space-time density required for detailed observation. Appropriate observing means comprise surface mesonetworks, minisondes, pilot balloons, tethered balloons, constant level balloons, sodars, boundary-layer wind profilers, etc.
Figure 3-6. Schematic of tentative cross-Alpine and along-barrier multi-level flight tracks serving
i) the study of Foehn, gravity waves (including wave breaking) and boundary layer processes, and
ii) the study of the lower-level wake-flow features. Top: Plan view. Note that cross-barrier flight
tracks can be rotated in azimuth and shifted in east-west direction to best match wind fields.
Bottom: Vertical cross-sections of flight tracks indicated in the top panel. The displayed aircraft
and flight level assignments are merely indicative rather than definitive (cf. Table 3-2).
F. Targeted Major Observing Systems
The following tables, one for ground-based and aerological systems (Table 3-1) and one for aircraft (Table 3-2), summarize the key-observing systems for the SOP of MAP. These tables do not list operational devices and are not complete in the sense that all systems which will be deployed during the field phase are listed. Emphasis is placed on sophisticated systems with special capabilities of non-operational nature which are particularly suited and desirable for the purposes of MAP. During the ongoing planning process specifications will be formulated in more detail.
| mesonets | 10 min | 10 km | target areas; transects | |
| microbarographs | 10 min | 10 km | abs. pressure: 0.3 hPa | |
| Doppler sodar | 1 min | 10 km | 10 m to 40 m | typical height range: 500 m |
| Doppler radar | 10 min | 1 km | ||
| Doppler lidar | 30 min | horizontal and vertical range very dependent on atmospheric conditions, up to 20km | ||
| Wind profiler/RASS
VHF (50mhz) Low UHF (400 MHz) UHF (1200 MHz) | 10 min 10 min 1 min | 375 m 150 m 75 m | Height range of profiler (associated RASS) 3 km to 30 km (7 km) 1 km to 16 km (3 km) 150 m to 4 km (1 km) | |
| additional radiosondes | 3 h | 40 km | 50 m | some mobile |
| light-weight dropsondes | curtains | 25 km | 50 m | see aircraft |
G. Observing Systems for Hydrology
Data sets used in hydrological modelling are precipitation, meteorological parameters such as 10m wind, surface and 2m temperature, dew point temperature, snow cover, runoff, evapotranspiration, soil water content and ground water levels. Precipitation, meteorological parameters and snow cover are observed by the dense network of the meteorological surface stations (Fig. 3-4) and by the rain gauge network (Fig. 3-5). The discharge in diversions and the volume of water stored in the major reservoirs must also be monitored.
For hydrological purposes, areal precipitation has to be calculated with suitable methods from point measurements. As an example, Cemagref in Lyon is ready to supply precipitation fields for single events based on rainfall measured on the ground. This information would be provided with a high spatio-temporal resolution for the northern part of the Alps: The one dimensional network from Lyon to the Belledonne massif in the French Alps, called the "TPG" (Transect of Pluviographs for analysis and modelling of rainfall Gradients), exists since 1987 and consists in three important assembly lines crossing the mountain ranges with increasing elevations: 1000m, 2000m and 3000m (Desurosne et al. 1994). The network density is about one gauge per 2.3km2.
| A. High Altitude (above 10000 m) | ||||
| Falcon-20 | DLR | 3 h/ 2000 km | 12000 m | lidar (DIAL); Doppler lidar (#), lightweight dropsondes (#), aerosol and droplet spectrometers |
| Gulfstream IV | NOAA | 9 h / 7500 km | 13500m | light-weight dropsondes; Doppler radar |
| WB57 | NCAR | 7 h / 5000 km | 20000m | light-weight dropsondes; lidar; atmos. chemistry |
| B. Medium Altitude (5000 m - 10000 m) | ||||
| C130 | NCAR | 10 h / 5500 km | 9000 m | light-weight dropsondes; Doppler radar (ELDORA); lidar; atmos. chemistry |
| C130 | UKMO | 10 h / 5500 km | 10000 m | light-weight dropsondes; microphysics |
| P3 | NOAA | 10 h / 5500 km | 7500 m | light-weight dropsondes; Doppler radar |
| Merlin IV | Météo France | 4 h | 8000 m | microphysics; turbulence |
| Fokker 27 ARAT | CNES, MF, CNRS, IGN | 3 h | 7000 m | microphysics; turbulence; lidar |
| Dornier-228 | DLR | 8 h / 2600 km | 8000 m | atmos. chemistry (trace gases), droplet spectrometer (#) |
| C. Low Altitude (below 5000 m) | ||||
| Stemme S10 | Met Air | 8 h / 1500 km | 5000 m | differential GPS for mesoscale pressure field; atm. Chemistry incl. NMHC |
For the estimation of areal precipitation in hydrological basins the network of weather radar will be an important complement to the traditional rain gauge network (Joss and Waldvogel 1990). In particular in high Alpine regions, where the ground based rain gauge network is less dense, radar provides improved resolution of rainfall measurements both in space and time. Radar data will help for the better understanding of the variability of the precipitation-elevation gradients in a mountainous environment
Figure 3-7. Part of the river runoff measuring station network in the Alpine region.
Work is underway to further complete the map. (Courtesy of P. Schädler)
Evaporation is either measured by agrometeorological stations in class-A-pans, or by other instruments that give not the real evaporation data but rather some kind of evaporation index. Reliable evapotranspiration measurements are available from weighing lysimeters. Only a few instruments are in operational use, by some agrometeorological stations or in research projects (Arbeitsgruppe Lysimeter 1989). To observe soil water content suitable instruments are weighing lysimeters, Time-Domain-Reflectometry (TDR) instruments or neutron probes. These point measurements are frequently not very representative - especially in an alpine environment - due to the high variability of soil types and vegetation cover.
Snow cover, which is important not only for hydrological purposes but also for energy balance considerations, is observed by different networks. Snow depth is usually observed daily, snow water equivalent is measured periodically by hand or monitored by automatic equipment. In addition, snow cover extension may be monitored by remote sensing techniques (Baumgartner and Rango 1991).
Runoff is measured by dense networks (Fig. 3-7) operated by different state hydrological services as well as by private companies (e.g. hydropower companies) and sometimes also by research institutes. Basin size varies between some hectares and several thousand square-kilometres. The stations are spread all over the Alps and can be found at very different altitudes. The time resolution of these data records is usually very high (some minutes). Unfortunately many of these runoff observations are distorted by anthropogenic effects: Diversions of water from one basin into another, artificial reservoirs (dams), drinking water supply etc. Therefore suitable basins have to be carefully selected.
International networks and data bases have been established in the framework of the International Hydrological Programme (IHP) of UNESCO: One is AMHY the "Alpine Mediterranean HYdrology" programme (Lama 1994), a subproject of the project FRIEND the "Flow Regimes from International Experimental and Network Data" (Roald et al. 1994). Another programme is ERB, the "Euromediterranean network of experimental and Representative Basins" (Barbet 1993).
In France discharge information is available from some small adjacent basins (ten or so) to the "TPG"-region and from fitted-together watersheds (about five) of the Guiers river (north- western part of the Isère department). This information together with the rainfall information is suitable to be taken into account in rainfall/runoff modelling.
Other important data for hydrological purposes are parameters describing basin characteristics such as topographic, morphologic, geologic, pedologic and land-use parameters. Such data can be found or assembled in Geographic Information Systems (GIS). Parameters for the ERB-basins can be found in the ICARE-database, the "Inventory of the CAtchments for Research in Europe" (Barbet et al. 1995). For the Swiss research basins they are published in Aschwanden (1996) and HADES (1992).
After the concerted observational effort of the field phase two major tasks need to be tackled as an integral part of MAP: Field data have to be assembled and analysed and the scientific hypotheses set up in Phase I need to be tested.
As outlined in Section 3.2 observational data from the field phase will originate from a variety of measuring platforms. Individual data sets may need completion, calibration and correction. Intercomparison and intercalibration studies (e.g. Richner 1985, for aircraft data) will be necessary in order to ensure and to increase the quality and consistency of the overall MAP field phase data. The creation of reference data sets, cf. the ALPEX level-IIB and ALPEX level-IIIB data sets in the past, will greatly increase the value of the data collected for subsequent studies.
Experience from ALPEX has shown that re-examination and analysis of observational data can be a tedious and long-lasting process. For instance, a documentation of errors for ALPEX level-IIB SYNOP and radiosonde data was not published until 5 years after the field experiment (Steinacker et al. 1987). The generation of the ALPEX level-IIIB data set is still in progress (Paul 1994). Thus, learning from this experience, it seems necessary that such initiatives form an integral part of MAP.
Ultimately, field observations will serve to meet the scientific objectives of MAP:
It is planned to make observational data available in real time. Such an effort will enable real-time evaluation of the collected data in the data assimilation cycle of ECMWF, and other prediction centres. It will be of special interest to monitor the impact of the extra data on the forecast quality.
The unique density and variety of measurements will render feasible the testing and improvement of objective mesoscale analysis techniques (in particular on the meso-b scale). The influence can be investigated of the individual kinds of data on the specification of the initial conditions for high-resolution numerical prediction models. Results from such studies will feedback on the design and equipment of future observing networks.
The assembled - and analysed - field data will be of particular value for case studies. They will help in the assessment of the quality of model forecasts in intercomparison studies and contribute to ascertaining the most effective version of an individual model. It is already proposed to include a future MAP case (or cases) in the catalogue of events to be studied within the frame of the COMPARE project (COMPARE Newsletter No. 6, 1995). Forecast verification is most important in view of further improvement of high-resolution numerical weather prediction models. It is well recognized, that present day observing networks do not meet the requirements of a serious verification on the meso-b scale in several respects. The MAP data set will improve this situation considerably.
The MAP data set, unique with respect to spatial and temporal resolution and to the selection of observed parameters, will help also in the validation of individual model components such as the parameterization schemes for cloud physics, convection and turbulence. Thus, new developments can be expected from the insights gained by consideration of the observations. Such considerations will relate to numerical, theoretical and conceptual studies.
Experience in utilizing field data to test the scientific hypotheses and to validate numerical models will help to improve the quality of the final data set. Data post-processing efforts and scientific investigation will benefit optimally, if they are closely tied together in Phase III.
The strength of MAP is that both scientific issues and practical applications will benefit. The MAP data set will be made available to interested institutions beyond those directly involved in the programme to enable a broad use of the data. It is believed, that a careful preparation of the data base will guarantee that the MAP data set becomes a long-term reference for studies in mountain meteorology.
