| description abstract | The findings of the Multiple Doppler Radar Workshop are summarized by a series of six papers. Part I of this series briefly reviews the history of multiple Doppler experimentation, fundamental concepts of Doppler signal theory, and organization and objectives of the Workshop. Invited presentations by dynamicists and cloud physicists are also summarized. Experimental design and procedures (Part II) are shown to be of critical importance. Well-defined and limited experimental objectives are necessary in view of technological limitations. Specified radar scanning procedures that balance temporal and spatial resolution considerations are discussed in detail. Improved siting for suppression of ground clutter as well as scanning procedures to minimize errors at echo boundaries are discussed. The need for accelerated research using numerically simulated proxy data sets is emphasized. New technology to eliminate various sampling limitations is cited as an eventual solution to many current problems in Part III. Ground clutter contamination may be curtailed by means of full spectral processing, digital filters in real time, and/or variable pulse repetition frequency. Range and velocity ambiguities also may be minimized by various pulsing options as well as random phase transmission. Sidelobe contamination can be reduced through improvements in radomes, illumination patterns, and antenna feed types. Radar volume-scan time can be sharply reduced by means of wideband transmission, phased array antennas, multiple beam antennas, and frequency agility. Part IV deals with synthesis of data from several radars in the context of scientific requirements in cumulus clouds, widespread precipitation, and severe convective storms. The important temporal and spatial scales are examined together with the accuracy required for vertical air motion in each phenomenon. Factors that introduce errors in the vertical velocity field are identified and synthesis techniques are discussed separately for the dual Doppler and multiple Doppler cases. Various filters and techniques, including statistical and variational approaches, are mentioned. Emphasis is placed on the importance of experiment design and procedures, technological improvements, incorporation of all information from supporting sensors, and analysis priority for physically simple cases. Integrated reliability is proposed as an objective tool for radar siting. Verification of multiple Doppler-derived vertical velocity is discussed in Part V. Three categories of verification are defined as direct, deductive, and theoretical/numerical. Direct verification consists of zenith-pointing radar measurements (from either airborne or ground-based systems), air motion sensing aircraft, instrumented towers, and tracking of radar chaff. Deductive sources include mesonetworks, aircraft (thermodynamic and microphysical) measurements, satellite observations, radar reflectivity, multiple Doppler consistency, and atmospheric soundings. Theoretical/numerical sources of verification include proxy data simulation, momentum checking, and numerical cloud models. New technology, principally in the form of wide bandwidth radars, is seen as a development that may reduce the need for extensive verification of multiple Doppler-derived vertical air motions. Airborne Doppler radar is perceived as the single most important source of verification within the bounds of existing technology. Nine stages of data processing and display are identified in Part VI. The stages are identified as field checks, archival, selection, editing, coordinate transformation, synthesis of Cartesian fields, filtering, display, and physical analysis. Display of data is considered to be a problem critical to assimilation of data at all stages. Interactive computing systems and software are concluded to be very important, particularly for the editing stage. Three- and 4-dimensional displays are considered essential for data assimilation, particularly at the physical analysis stage. The concept of common data tape formats is approved both for data in radar spherical space as well as for synthesized Cartesian output. 1169 | |
