Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
3 Observational Strategy To accomplish the stated scientific objectives the observational network must be capable of 1. defining the synoptic-scale fields over the entire tropical Atlantic and neighboring land masses with sufficient accuracy and resolution to define the major synoptic systems; 2. determining the bulk of properties of convective ensembles; and 3. intensive sampling of the internal structure of the convective ensembles whose bulk properties are being monitored, in order to obtain representative statistics. Each of the above functions will require a special combination of observa- tional tools and techniques. The total observational program can be viewed as consisting of the sum of these three subsystems, plus whatever is needed to coordinate them in order to ensure that the observational requirements of the experiment are met. 3.1 Definition of the Synoptic-Scale Fields Definition of the synoptic-scale fields will be accomplished primarily by satellite observations, which must be routinely available by the time of the experiment. These will be supplemented by radiosonde observations from land and ship stations in the region of the experiment together with reports from commercial aircraft and ships of opportunity (in addition to the ordinary net- 13
work, a number of additional stations are being established as part of the World Weather Watch). It is also expected that a number of oceanographic vessels will be making regular traverses of the ITCZ. A certain number of dropsonde or ship-based radiosonde observations may be required to provide additional definition over the remote midoceanic regions. The number of ships that can be used for this purpose will be limited to those that are available over and above what is needed for monitoring bulk properties of convective ensembles (see Sec. 3.2). The extent of the midocean synoptic network is one of the key remaining questions to be settled. This synoptic network will operate on a fixed schedule in real time. The processing, analysis, and archiving of these data can be accomplished largely through the use of existing operational facilities. 3.2 Determination of the Bulk Properties of Convective Ensembles Determination of the bulk properties of convective ensembles will require 1. detailed vertical profiles of the convergence of mass, heat, moisture, and momentum into the area enclosing one or more such ensembles (radiosonde data); 2. mapping of the distribution of active convective elements within the area (satellite reconnaissance and radar); and 3. quantitative estimates of total precipitation falling within the area (radar and some rain gauges). In order to obtain statistics with sufficient accuracy and representativeness it will actually be necessary to sample the bulk properties of a large number of convective ensembles. This can conveniently be accomplished by establishing one or more ship-based networks. The operation of this network would not be confined to periods of dis- turbed weather, since it will also be necessary to compile statistics on the bulk properties of the undisturbed tropical atmosphere. At synoptic observa- tion times, the ship-based network will also function as an integral part of the synoptic network. 3.2.1 Location of the ship-based networks must satisfy the following require- ments: (a) high probability of encountering active convection associated with synoptic disturbances in varying stages of development; and (b) proximity of bases for aircraft operations (see Sec. 3.3), supplies, communications, and other necessities. 14
The only regions that meet these criteria are the extreme western Atlantic in the vicinity of the Leeward Islands and the extreme eastern Atlantic be- tween 5N and 20N. The latter region experiences a much higher frequency of active convection. The decision of the Tropical Experiment Board (see GARP Special Publication Nos. 1 and 2) that the experiment should be conducted in the eastern Atlantic is endorsed provided that the necessary components of the observing system can be made available in this region. 3.2.2 Spacing of ships in the network will be governed by the following considerations: (a) The probability of encountering active convection is pro- portional to the area of the network (i.e., the square of the station spacing). (b) It would be highly desirable to have overlapping radar coverage. This limits the spacing to approximately 250 km. (c) Spacings of on the order of 500 km appear to be most efficient from the point of view of monitoring bulk properties. It would appear that the effectiveness of the ship-based network can be maximized by clustering the ships with radar in the inner part of the network, with spacing of about, or less than, 250 km, and by positioning the ships without radar around the periphery of the network on the order of about 500-km spacing. 3.2.3 The success of the experiment depends critically upon our ability to monitor processes occurring within the subcloud layer. In order to obtain the necessary vertical and time resolution it will be necessary to supplement the radiosonde data with soundings made with tethered balloons. 3.2.4 The network should be fixed in space and should operate on a regu- lar schedule for radiosonde ascents (6-hour intervals during undisturbed periods and 2- or 3-hour intervals during disturbed periods). 3.2.5 With the exception of certain radar products required for real-time operation of the aircraft (see Sec. 3.3), the data from this network can be processed in the constant-lag mode (see Chapter 5). Special centralized facilities will be required for data transmission, reduction, and archiving. 3.3 Intensive Sampling of Convective Ensembles Intensive sampling of the internal structure of convective ensembles should aim at providing representative statistics on the following items, ranked in order of priority: 15
1. the horizontal distribution of divergence and vorticity at various levels in the vicinity of mesoscale convective elements; 2. heights, vertical profiles, and internal structure of deep, convective clouds; 3. typical updraft and downdraft velocities and areal coverage at various levels; and 4. lifetimes and typical life-cycles of mesoscale convective systems. In addition to the above items, it would be desirable to obtain information on (a) the rate of sensible heat flux and evaporation from the sea surface under disturbed conditions; (b) typical distribution of drop sizes, liquid water content, and relative humidity in and around deep convective clouds; (c) relations between positions of mesoscale convective elements and sea-surface temperature anomalies; (d) radiative heating rates under various cloud conditions; and (e) effective freezing level for cloud droplets in deep convection. This subprogram of the experiment will be mainly based on the use of instrumented aircraft, which can be viewed as high-speed observing platforms, operating on a flexible schedule determined on the basis of real-time satellite and radar information. This will be supplemented by ground- or ship-based Doppler radar. The aircraft operations will take place within the ship-based network discussed in Sec. 3.2, but the flight paths and schedules need not be coordinated with the radiosonde subprogram. The data will be processed in the constant-lag time mode. The success of this part of the observational program will depend critically upon the number and type of instrumented aircraft that are made available by the participating nations (see Sec. 4.1.4). 3.4 Other Observations Other observations may include oceanographic measurements of various types as suggested by the Joint (ICSU/WMO) Organizing Committee (JOC), aerosol sampling, and measurements related to cloud microphysics. In general, these observations will be planned, implemented, and processed by individual investigators and, therefore, fall outside the range of responsibility of the planning committees, except in cases where they interact with the other subprograms (e.g., in use of facilities). Every effort should be made to accom- 16
modate these peripheral observational programs, as long as they do not inter- fere with the observational plan stated above. 3.5 An Overview We have seen that the overwhelming task of formulating a comprehensive observational program can be simplified to some extent by breaking it down into three subprograms, each of which is designed to accomplish a limited set of objectives. Quite fortuitously, it happens that each of the subprograms described above requires a different set of observational tools; apart from the satellite, there is relatively little overlap between the three observing sub- systems. It also happens that there exists a precedent for each of the sub- programs: (a) Definition of the synoptic-scale fields will be based on a refined version of current operational analysis techniques, which already make extensive use of a mix of satellite, radiosonde, commercial aircraft, and surface observa- tions. (b) Several groups working independently have had some degree of success in deducing the bulk properties of deep cumulus convection on the basis of the island-based radiosonde network in the western Pacific. These studies provide a basis for specifying the number of observations, the accuracy, and resolution required to obtain adequate definition of the quantities to be measured. (c) Research aircraft have already seen extensive use in past field pro- grams. Published studies based on analysis of the data from these programs have already shown that most of the measurements listed in Sec. 3.3 can be made successfully with aircraft. Thus, there is a rational basis for the design of each of the three subpro- grams. During the planning and implementation phases of the experiment, only a minimal amount of coordination between subprograms may be re- quired. It is in the analysis phase that the parts of the experiment should merge into a coherent whole. For the purpose of clarifying differences between the observational strategy outlined above and those presented in previous reports, it may be helpful to mention a number of objectives that the proposed experiment will not accomplish. (a) It will not be possible to attain a clear definition of equatorial wave modes as envisioned in plans for the western Pacific experiment. Only in the 17
equatorial mid-Atlantic would there be any possibility of studying wave dis- turbances that are not strongly affected by the land-sea distribution. This area has relatively little active convection and is not conveniently located from a logistical point of view. (b) It will not be possible to resolve individual mesoscale convective sys- tems with the radiosonde network. It is doubtful that this could be done even with a ship spacing an order of magnitude smaller than contemplated in Sec. 3.2.2. A spacing this small would not be compatible with the objec- tive of obtaining definitive statistics on the bulk properties of convective ensembles, since, with the greatly reduced area of the network, it would not be possible to sample a sufficient number of systems during the course of the experiment. The observational plan outlined above will rely on radar and satellite data for determining the position, size, shape, and intensity of the mesoscale con- vective systems within the network. Since both vorticity and divergence of the horizontal wind field tend to be concentrated within the mesoscale con- vective elements, there is reason to hope that the area integrals obtained from rather widely spaced stations, if carefully interpreted, can serve to define the gross features of the mesoscale wind field. (c) The experiment is not specifically designed to monitor planetary-scale features or long-term periodicities. The synoptic subprogram may provide some useful information on the largest scales, but it should be borne in mind that this will not be much different from the type of information that will be routinely available when World Weather Watch is fully implemented. The experiment will not be of sufficient duration for studying long-term period- icities. (d) A repetition of the detailed study of sea-air exchange processes, which was the central focus of BOMEX, is not recommended. The sea-air transfer measurements mentioned in Sec. 3.3(a) are for the purpose of estimating the typical heat and moisture fluxes from the sea surface during disturbed condi- tions. 18
