Chapter 2:

Applications of Doppler Weather Radar

 
James W. Wilson
 

2.1. Introduction

This chapter reviews applications of Doppler weather radar for weather analysis and forecasting. As a result of extensive research with Doppler radar during the 1970's and 1980's a variety of Doppler radar applications that could significantly improve weather warnings and very-short period forecasting were demonstrated. In 1991 these results motivated the National Weather Service (NWS), Federal Aviation Agency (FAA) and the Air Force to begin the installation across the United States of an operational network of about 160 Doppler radars, called the WSR-88D. In 1992 the FAA began to install another network of about 45 Doppler radars called the Terminal Doppler Weather Radar (TDWR) in the vicinity of major airports.

Researchers use Doppler radars to obtain detailed understanding of storm and clear-air boundary layer wind fields. Using techniques originally proposed by Lhermitte (1968) and refined by Armijo (1969), detailed three-dimensional fields of air motion are obtained by combining the Doppler radial wind components from two or more radars. Using multiple Doppler techniques to obtain the full wind field would be impractical for operational applications because of the large number of radars required. Fortunately, experiences with single Doppler radar in research and quasi-operational experiments has demonstrated numerous applications for warning and short-period forecasting.

A Doppler radar directly measures only the wind component along the beam axis; interpretation of Doppler velocity displays is not straightforward and requires training and experience using mostly pattern recognition techniques. The identification of tornadoes, supercell storms, hail storms, gust fronts, downbursts and fronts is presently accomplished by computer algorithms and by the forecaster monitoring and interpreting displays. The computer algorithms are far from mature and the forecaster plays a critical interpretation role. Doppler radar products such as divergence, vertical and horizontal wind velocities, microbursts, storm movement vectors and precipitation amounts are more successfully obtained by computer but still benefit substantially from human input. Effective Doppler interpretation in real-time requires a human-machine mix, with the computer expected to take a greater share of the load as automatic interpretation techniques evolve.

Much of the material presented here on Doppler radar applications has been taken from earlier publications (Wilson et al. 1980; Wilson and Wilk 1982; Wilson and Carbone 1984; Wilson and Roesli 1985). Because of high publication costs for color pictures it will be necessary to show examples of the radar display in black and white. This makes interpretation of the radar displays considerably more difficult. The interested reader is encouraged to look at the above references where many of the same displays are presented in color. The reader is also referred to a very comprehensive discussion of the history and applications of weather radar that was written by a large number of experts and edited by Atlas (1990).

2.2. Applications

2.2.1 Wind versus height

Meteorological Doppler radars generally detect precipitation particles, insects and refractive index gradients. These scatterers move with the wind. Thus Doppler radar can measure the wind component in the direction in which the radar beam is pointing. Lhermitte and Atlas (1961) first described how a single Doppler radar could be used to measure vertical profiles of wind velocity in widespread precipitation; it can also be used in clear-air. Their technique, which produces the Velocity-Azimuth Display, or VAD, involves rotation of the antenna while it is directed at a constant elevation angle to record radial velocity at a fixed range versus azimuth.

Browning and Wexler (1968) presented a through analysis of the VAD technique showing how wind velocity, divergence and deformation of the wind field can be obtained via harmonic analysis. Baynton et al. (1977) and Wood and Brown (1986) have described how wind information can be readily obtained visually from the Doppler radar color display. With experience, many wind features can be identified visually from color displays Time-lapsing greatly facilitates this process. Much of the success in identifying important weather phenomena lies in the ability to observe strong velocity gradients in range and azimuth, allowing for the identification and quantification of confluent, diffluent and rotational wind features.

Figure 2.1 is used to help explain simple visual interpretation procedures for the radial velocity field and to describe the VAD technique. The Doppler velocity pattern in Fig. 2.1 was obtained by pointing the antenna at an elevation angle of 7o and rotating 360o in azimuth. The zero Doppler velocity contour is highlighted as white in Fig. 2.1; receding velocities are lightly shaded and approaching velocities darkly shaded. The data were taken during a widespread precipitation event. Because the antenna is elevated, the height of the radar beam increases with range; e.g., at 30 km range the beam height is 3.7 km. The radial velocity (Vr) is a function of the horizontal wind speed (Vh), wind direction (), the fall velocity of the particles (Vf) and the antenna azimuth () and elevation () and is given by

Vr = Vh cos cos(-) -Vf sin                                         (2.1)

In a uniform wind field, the radial velocity for a fixed range will vary sinusoidally with azimuth. Figure 2.2 is an example of a VAD display for the 10 km range (height=1.2 km) in Fig. 2.1 showing this sinusoidal variation. The minimum Doppler velocity occurs where the radar beam is pointing upwind (=210o) and a maximum when pointing downwind. The horizontal wind speed can be closely approximated at two points around the circle: the maximum and minimum. Direction estimates can be estimated at four points: the maximum, the minimum and two zero Doppler velocity points. The wind will be normal to the radar beam at the zero points. From Fig. 2.2 or Fig. 2.1 (10 km range) all four estimates indicate a wind direction of roughly 210o and the minimum and maximum points indicate a speed of about 33 m/s. When these points give differing values this indicates the presence of non-uniformities in the wind field which can be utilized to determine other important wind features. Wind direction as a function of height can most easily be determined from the zero-velocity contour. In Fig 2.1, the zero contour (white) rotates clockwise with height, indicating veering winds and warm advection. A low-level jet or wind maximum exists at a height of ~1.3 km (11 km range).

Figures 2.3a-2.3d illustrate some of the quantities (wind speed, wind direction, divergence and vertical velocity) that can be derived from a harmonic analysis of VAD-type data. Details of the technique is given by Browning and Wexler (1968). These data can be produced in near real-time by a computer. The six hour analysis period in Fig. 2.3 includes the time period of Fig 2.1.

2.2.2 Front location

Sharp wind shifts, which are frequently associated with frontal boundaries, are usually easily identified on the Doppler velocity display. Thus, fronts can be precisely located and their movement closely monitored. Figure 2.4 shows the low-level wind shift associated with a cold front approaching the coast of Washington state. In this geographic region accurate frontal locations are rarely known because data is sparse. Note that west of the radar the zero velocity band makes a sharp bend from about 290o to 360o. Along the sharp bend and extending to the southwest there is a very close packing of the velocity contours. This indicates a sharp change in wind direction from about 200o to 280o. From further inspection of the contours, an estimation of wind speeds of about 24 m/s and 15 m/s ahead of and behind the front, respectively are obtained. Assuming constant advection speed, the forecaster can accurately predict the time of frontal passage and wind velocity after passage.

2.2.3 Clear air features

Sensitive research radars and the operational WSR-88D and TDWR are able to routinely observe winds in the clear-air boundary layer particularly during the warm season. The principal sources of clear-air return, particularly at radar wavelengths less than 10 cm, is from particulates (Hardy and Katz 1969; Gossard 1990). The particulate scatterers in most cases are insects. Refractive index inhomogeneities (Bragg scattering), particularly at longer (10 cm) wavelengths, contribute to clear air return in frontal zones. Knight and Miller (1993) have also shown that cumulus clouds can be observed even when they first become visible because of Bragg scattering along the edges of the cloud. The refractive index inhomogeneities result primarily from large local gradients in water vapor on the scale of one-half the radar wavelength.
 

Similar to the examples for widespread precipitation in Fig. 2.3 the VAD technique can be used to monitor the vertical wind profile in the clear-air boundary layer.

Boundary layer convergence lines such as synoptic fronts, gust fronts, sea-breeze fronts and horizontal convective rolls are usually visible on sensitive Doppler radars as lines of enhanced reflectivity factor and/or a line of radial velocity convergence. Figure 2.5 shows several thin-line echoes associated with convergence lines from Colorado. Two gust fronts and three horizontal convective rolls are marked. Figure 2.6 shows a field of thin lines caused by horizontal convective rolls observed in Kansas. The thin lines are associated with the low-level convergent regions between oppositely rotating rolls. A detailed description of this commonly observed feature is given by Christian and Wakimoto (1989).

2.2.4 Forecasting thunderstorm initiation

Satellite cloud imagery shows that convective storms are often triggered at intersecting arc cloud lines generated by the outflow of pre-existing convective storms (Purdom 1982). Numerous other investigators using dense networks of surface stations, radar and numerical models have also shown the importance of near-ground convergence for the initiation of convective storms (e.g., Byers and Braham 1949; Watson and Blanchard; 1984, Rabin and Doviak 1982; Crook et al. 1991).

Wilson and Schreiber (1986) have shown that thunderstorms typically initiate along boundary layer convergence lines that are visible on Doppler radars. Wilson and Mueller (1993) have demonstrated how the monitoring of these boundary-layer convergence lines can be used to successfully prepare very short period forecasts of thunderstorm initiation.

Figure 2.7 is an example of a line of thunderstorms that resulted from the collision of two convergence lines in Colorado. The boundary labeled 1 is moving from the northwest and was the result of cool outflow from thunderstorms over the mountains. The boundary moving from the southeast (labeled 2) is of unknown origin but was intensified by outflow from the line of thunderstorms immediately to its southeast. Echoes >30 dBZ are shown in black. The boundaries first collide at location A (Fig. 2.7c) and they continue to collide both northeast and southwest from this point for the next 20 min. The boundaries then become one and move toward the northwest. A new line of echoes can be seen in Fig. 2.7f along the line of collision (labeled B). The old line of echoes has almost dissipated by this time. Besides boundaries A and B, other thin-lines oriented north-south are visible; these appear to be horizontal rolls. Boundary 1 intersects the most pronounced roll (labeled 3) and a short line of thunderstorms is initiated (labeled C).

Figure 2.8 is an example of a 30 min nowcast of thunderstorm location based on the forecast techniques described by Wilson and Mueller (1993). These experimental nowcasts are being sent to Denver's Stapleton Airport Control Tower. The forecaster in this example was able to nowcast the initiation of a line of thunderstorms where there was no previous precipitation echo by monitoring the position of a convergence line and the growth of cloud echo in its vicinity. About 10 min after the validation time almost the entire nowcast polygon was full of echo.
 

2.2.5 Severe Storms

2.2.5.1 Tornadoes

Non-coherent radars have long been used to identify storms that are likely to produce severe weather. The maximum reflectivity, reflectivity structure and height of the echo are all indicators of severity (Burgess and Ray 1986). Once a storm is identified as severe, its tracking and extrapolation become very important for warning purposes. Doppler velocity data has significantly improved the radar meteorologist's ability to identify severe storms. In particular, mesoscale vorticity signatures in the Doppler radial velocity field can be used to identify supercell storms. By monitoring the magnitude of this mesocyclone, tornado warnings can be issued (JDOP staff 1979; Burgess and Ray 1986). The mesocyclone signature is a couplet of closed Doppler velocity contours of opposite sign spaced in azimuth by about 4-15 km. In some conditions a tornado may produce a very large beam-to-beam shear called the Tornado Vortex Signature (TVS) (Brown et al. 1978).

Figure 2.9 is an example of two tornado-producing supercells that were observed in central Oklahoma. The large mesocyclone centered at 28o, 60 km, (Fig. 2.9b) has maximum approaching velocities of 32 m/s on the west side, with receding velocities of 42 m/s 8 km to the east (shear 1.1x10-2s-1). At the mesocyclone center, there is a TVS (not obvious in black and white version). Adjoining azimuths indicate speed change from -32 to +42 m/s. The other tornado-producing mesocyclone, centered at 5o and 45 km, shows a velocity couplet of -28 and +21 m/s separated by 5 km. Each mesocyclone produced a well-defined hook echo in the reflectivity display (Fig. 2.9a). Hook echoes by themselves are not reliable signatures for the presence of tornadoes because these echoes are often difficult to detect and may occur without mesocyclones and tornadoes.

Small, short duration tornadoes often occur in the absence of a supercell storm, these non-supercell tornadoes can only be detected within about 40 km of the radar. These tornadoes discussed by Wakimoto and Wilson (1989) and Brady and Szoke (1989) can occur when pre-existing boundary layer misocyclones (diameter <4 km) along convergence lines rapidly intensify when they become co-located with an intense updraft of a rapidly developing convective storm. Forecasts of non-supercell tornadoes are possible by monitoring boundary layer convergence lines for misocyclones and rapidly developing cells. A warning can be issued when the misocyclone begins to strengthen and grow vertically when colocated with a cell rapidly developing overhead.

2.2.5.2 Macrobursts

Occasionally a convective storm will initiate a large-scale diverging outflow that produces damaging surface winds over a large area. Fujita (1981) has named these "macrobursts." Figure 2.10 shows an example of such an event occurring in eastern Colorado. The reflectivity display (Fig. 10a) shows the southern end of a band of 55 dBZe thunderstorms. The Doppler-velocity display (Fig. 2.10b) shows a large, strong outflow region at the eastern portion of the storm. An area 15 km long has radial velocities in access of 30 m/s. Maximum Doppler velocities of 44 m/s occurred within the dark black region.
 

2.2.5.3 Hail

A variety of techniques have been tested that use radar to detect hail. At first, measurements of radar reflectivity alone were used as a hail detector (Ward et al. 1965). The success of this technique has been varied (Dye and Martner 1978; Waldvogel et al. 1978). Petrocchi (1982) and Smart and Alberty (1984) successfully developed and tested, a computer algorithm for hail identification. This algorithm applied Lemon's (1978) three-dimensional reflectivity structure model for a severe hailstorm.

Witt and Nelson (1984) utilized single Doppler radial-velocity measurements of divergent outflow at storm top to predict, with encouraging success, maximum hail size.

A promising technique for identifying hail regions within storms uses dual-polarization or differential-reflectivity (ZDR) measurements. This technique subtracts the vertically polarized reflectivity factor from the horizontal reflectivity factor. Values greater than zero correspond to rain while frozen precipitation will generally give values near 0 dB. Dual polarization is only available on research radars at this time.

2.2.6 Aviation Weather

2.2.6.1 Microburst warnings

Wind shear in the lowest few hundred meters of the atmosphere has been responsible for a number of aircraft accidents (Fujita, 1980, National Research Council 1983; NTSB 1983). The most dangerous situation is when an aircraft encounters a microburst when landing or taking off. The aircraft will experience a sudden increase in head wind, quickly followed by a strong tailwind and possible catastrophic loss of air speed. Doppler radar has been found to be a very effective means for identifying microbursts (Wilson et al. 1984). In fact, the TDWR Doppler radars are presently being installed in the vicinity of major airports within the U.S. to warn aircraft of dangerous windshear conditions.

Figure 2.11 is an example of the evolution of a typical microburst as observed by Doppler radar while scanning at an elevation angle of 0o. The presence of a negative velocity gradient with increasing range is an indication of divergence. The signature for a microburst is typically closed Doppler velocity contours of opposite sign, spaced in range along the same azimuth less than 4 km apart. The microburst center in Fig 2.11 is at a radar range of ~27 km (range marks at 20 and 30 km) and the effective height of the radar beam is ~100m. The four displays cover a 7 min time period. During this time, the microburst is born and reaches a peak difference of 25 m/s between maximum approaching (~19 m/s) and receding (6 m/s) radial velocities. Two minutes after Fig. 2.11d, the flow weakened and spread horizontally. Twelve minutes after initiation, the event was finished.
 

Figure 2.12 is a black and white representation of displays used by aircraft controllers at Denver's Stapleton airport. The display in Fig. 2.12a, which is based primarily on Doppler radar data, is used by FAA supervisors to plan airport operations. This display which is updated at 1-5 min intervals provides information on precipitation location and movement, hazardous windshear locations and intensities, and windshift line locations and forecast locations. Figure 2.12b is used by controllers to provide pilots with windshear advisories, for example the top line would be read to a pilot on approach to runway 26 as microburst alert expect 35 knots loss at 3 miles distance on final approach; threshold winds are 110o at 6 knots.

2.2.6.2 Windshift forecasts

Particularly important to the efficient operation of airports is the anticipation of wind changes that require reconfiguration of landing and takeoff patterns. Forecasts of windshifts with lead times of about 20 min can greatly improve efficiency. Operational computer algorithms have been prepared which search Doppler radar data for wind shift lines based on thin-line reflectivity echoes as observed in Figs. 2.5 and 2.7 and Doppler velocity convergent features like shown in Fig 2.4. By monitoring and linearly extrapolating these features, accurate windshift forecasts are automatically made and provided to the controllers. The arching line northeast of the airport in Fig. 2.12a is an example of a computer detected windshift line; the dashed lines north of the detected position is its forecast position every 15 min.

2.2.6.3 Weather-impacted airspace

An additional product is under development for traffic controllers which is called weather impacted airspace. It will be based primarily on Doppler radar data from both the TDWR and WSR-88D's. The product will be used by controllers and pilots to determine which airspace would be hazardous to aircraft operations. Input to the product would include radar detection of tornadoes, supercell storms, hail, heavy rain and strong windshears.
 

2.3. Conclusions

The early experience of operational forecasters as the WSR-88D's are being installed at Weather Forecast Offices across the U.S is that they are able to improve their abilities to understand and forecast severe thunderstorms, blizzards, windshifts, thunderstorm initiation and to monitor precipitation accumulations. They are also benefiting by the ability to observe and understand how small scale wind features affect precipitation events. It is expected that as more and more operational Doppler radars are installed and a large number of forecasters routinely examine the data that additional applications will emerge.

This review has emphasized analysis, forecasting and warning applications possible with Doppler radar, however, the radar should not be considered as a stand alone tool. It must be integrated with data from surface stations, human observers, radiosondes, profilers, satellite, lightning detectors and numerical weather computer forecasts.

Not discussed in this review are a number radar limitations that in certain meteorological situations can seriously effect the utility of the data. The most serious of these are range folded echoes, ground clutter, velocity folding and side lobe echoes. These radar limitations are discussed in chapter ???. These data problems seriously affect the accuracy of algorithms to detect severe storms, estimate rainfall amount and detect windshift lines. The full utilization of the Doppler radar data is also limited by the inability of the human to continuously observe and extract pertinent information from radar displays. Thus, it is essential that effective computer algorithms are available that will aid the forecaster. This includes the detection and extrapolation of severe storms, tornadoes, windshift lines and the preparation of rainfall accumulation maps and time histories of wind profiles. To mitigate the effect of data limitations and to improve algorithms, it is essential that a significant ongoing engineering and research effort be in place.
 

Acknowledgments. The author would like to especially thank Dan Megenhardt of NCAR who converted the numerous figures from color to black and white. Tammy Weckwerth is thanked for reviewing the manuscript.
 
 

2.4 References

Armijo, L., 1969: A theory for the determination of wind and precipitation velocities with Doppler radars. J. Atmos. Sci., 25, 570-573.

Atlas D., 1990: Radar In Meteorology. American Meteorology Society, Boston, 806 pp.

Baynton H.W., R.J. Serafin, C.L. Frush, and G.R. Gray, 1977: Real-time wind measurements in extratropical cyclones by means of Doppler radar. J.Appl. Meteor., 16, 1022-1028.

Brady R.H., and E.Szoke, 1989: A case study of nonmesocyclone tornado development in northeast Colorado: similarities to waterspout formation. Mon. Wea. Rev, 117, 843-856.

Bromn R.A., L.R. Lemon and D.W. Burgess, 1978: Tornado detection by pulsed Doppler radar, Mon. Wea. Rev., 106, 29-38.

Browning, K.A. and R. Wexler 1968: A determination of kinematic properties of a wind field using Doppler radar. J. Appl. Meteor., 7, 105-113.

Burgess, D.W. and P.S.Ray, 1986: Principles of radar. Mesoscale Meteorology and Forecasting, P.Ray, Ed., Amer. Meteor. Soc., 85-117.

Byers H.R. and R.R.Jr. Braham, 1949: The thunderstorm, U.S. Government Printing Office, Washington, D.C. 287pp.

Christian and Wakimoto, 1989: The relationship between radar reflectivities and clouds associated with horizontal roll convection on 8 August 1982. Mon. Wea. Rev., 117, 1530-1544.

Crook, N.A. T.L.Clark and M.W.Moncrieff, 1991: The Denver cyclone, Part II: Interaction with the convective boundary layer. J. Atmos. Sci., 44, 2109-2126.

Dye J.E. and B.E. Martner, 1978: The relationship between radar reflectivity factor and hail at the ground for northeast Colorado thunderstorms, J.Appl. Meteoro.l, 17, 1335-1341

Fujita, T.T., 1980: Downburst and microbursts - an aviation hazard, Preprints, 19th Conf. on Radar Meteoro., AMS, Boston 94-101.

Gossard E.E., 1990: Radar Research on the atmospheric boundary layer. Radar In Meteorology, D.Atlas Ed., Amer. Meteor. Soc. 477-533.

Hardy, K.R. and I. Katz, 1969: Probing the atmosphere with high power, high resolution radars. Atmospheric exploration by remote probes. Vol 2, pp. 217-243. Washington, D.C., Nat. Acad. Sci.

Knight C.A. and L.J.Miller, 1993: First radar echoes from cumulus clouds. BAMS, 74, 179-188.

Lemon L.R., 1978: On the use of storm structure for hail identification, Preprints, 18th Conf. on Radar Meteorology, AMS, Boston, 203-206.

Lhermitte, R.M., 1968: New developments in Doppler radar methods. Preprints, 13th Conf. on Radar Meteorology (Montreal), AMS, Boston 14-17.

Lhermitte, R.M. and D.Atlas 1961: Precipitation motion by pulse-Doppler radar. Preprints, 9th Conf. on Radar Meteorology (Norman), AMS, Boston, 218-223.

National Research Council, 1983: Low-Altitude Wind Shear and its Hazard to Aviation, National Academy Press, 112 pp.

NTSB, 1983: Aircraft accident report - Pan American World Airways, Clipper 759, N4737, Boeing 727-235, New Orleans International Airport, Kenner, Louisiana, 9 July 1982, NTSB/AAR-83/82, National Transportation Safety Board, Washington DC, 115 pp.

Petrocchi P.F., 1982: Automatic detection of hail by radar, Air Force Geophysics Lab., Tech Report 82-0277, Hanscom Air Force Base, Maryland.

Purdom J.F.W. 1982: Subjective interpretation of geostationary satellite data for nowcasting. Nowcasting, Editor K.Browning, Academic Press, London, 149-166.

Rabin R.M. and R.J. Doviak, 1982: Prestorm observations in the clear air boundary layer with a Doppler radar, Preprints, 12th Local Severe Storms Conf., AMS, Boston, 425-429.

Smart J.R. and R.L. Alberty 1984: An evaluation of the performance of the NEXRAD hail algorithm, Preprints, 22nd Conf. on Radar Meteoro., AMS, Boston 202-207.

JDOP Staff, 1979: Final report on the Joint Doppler Operational Project (JDOP) 1976-1978. NOAA Tech Memo. ERL NSSL-86, Norman Okla., 84 pp.

Waldvogel A., W.Schmid and B. Federer, 1978: The kinematic energy of hailfalls. Part I: Hailstone spectre. J.Appl. Meteorol., 17, 515-520

Ward N.B., K.E. Wilk, and W.C. Herman, 1965: WSR-57 reflectivity measurements and hail observations, ESSA Tech Note 3, NSSL-24, Norman OK, 1-8

Watson A.I. and D.O.Blanchard, 1984: The relationship between total area divergence and convective precipitation in south Florida, Mon. Wea. Rev., 112, 673-685

Wakimoto R.M. and J.W. Wilson, 1989: Non-supercell tornadoes, Mon. Wea. Rev, 117, 1113-1140.

Wilson J.W., R.E. Carbone, H. Baynton and R. Serafin, 1980: Operational applications of meteorological Doppler radar. Bull. Amer. Metero. Soc., 61, 1154-1168.

Wilson J.W. and R.E. Carbone 1984: Nowcasting with Doppler radar: The forecaster-computer relationship, Proc. Nowcasting-II Symposium, Norrkoping, Sweden, ESA SP-208, pp 177.

Wilson J.W., R.D. Roberts, C.K. Kessinger and J. McCarthy, 1984: Microburst wind structure and evaluation of Doppler radar for airport wind shear detection, J.Climate Appl. Meteorol., 23, 898-915.

Wilson J.W. and C.K.Mueller, 1993: Nowcasts of thunderstorm initiation and evolution, Wea. Forecasting, 8, xxx

Wilson J.W. and H.P. Roesli, 1985: Use of Doppler radar and radar networks in mesoscale analysis and forecasting, ESA Journal, 9, 125-145.

Wilson J.W. and W.E. Schreiber, 1986: Initiation of convective storms by radar-observed boundary layer convergent lines. Mon. Wea. Rev., 114, 2516-2536.

Wilson J.W. and K.E. Wilk, 1982: Nowcasting applications of Doppler radar. Nowcasting, K.Browning, Ed., Academic Press, London, 87-105.

Witt A. and S.P. Nelson, 1984: The relationship between upper-level divergent outflow magnitude as measured by Doppler radar and hailstorm intensity, Preprints, 22nd Conf on Radar Meteorol. AMS, boston, 108-111.

Wood, V.T. and R.A.Brown, 1986: Single Doppler velocity signature interpretation of nondivergent environmental winds. J.Atmos. Oceanic Technol., 3, 114-128.
 
 
 
 
 
 
 
 

Figure captions
 

Fig 2.1. Doppler velocity display for a widespread precipitation event along the Washington state coastline on 1316 PST 4 March 1979. The antenna elevation angle is 7o and the range marks, labeled toward the southwest, are at 10 km intervals. Approaching velocities (negative values) are darkly shaded while receding velocities are lightly shaded. The zero velocity contour is in white. The velocities are contoured in 5 m/s intervals and labeled every 10 m/s.
 

Fig 2.2. Velocity-Azimuth Display (VAD) for the data at the 10 km range ring in Fig. 2.1.
 

Fig 2.3. Time-height displays of wind features derived from harmonic analysis of VAD data for a 6 hr period which contains the time period in Fig. 1. a) wind speed, b) wind direction, c) divergence and d) vertical velocity.
 

Fig 2.4. Doppler velocity display of a cold front approaching the Washington state coastline. The display shows a near surface (elevation angle 0o) sharp wind shift line 30 km west of the radar. Note the sharp bend in the zero velocity contour and close packing of the contours. Winds east of the front about 150 m above the surface are 200o at 24 m/s and west of the front 280o at 15 m/s.
 

Fig 2.5. Radar reflectivity factor display showing clear-air enhanced thin-line echoes associated with two gust fronts and horizontal convective rolls. The data were collected near Denver, Colorado on 5 September 1990. Reflectivity factors in dBZ are given by the scale at the bottom. The antenna elevation angle is 1.2o. Range marks are at 20 km intervals.
 

Fig 2.6. Radar reflectivity factor display of thin line clear-air echoes associated with horizontal convective rolls. The thin lines are the updraft regions between oppositely rotating convective rolls. Data were collected 1 March 1991 in Kansas. The antenna elevation angle is 1o. Range marks are at 20 km intervals.

 

Fig 2.7. Radar reflectivity factor display showing the collision of two wind shift lines (labeled 1 and 2) and resultant line of initiated thunderstorms (labeled B). Another short line of thunderstorms (labeled C) is initiated when boundary 1 intersects a horizontal roll (labeled 3). Echoes > 30 dbZ are black. Boundaries 1 and 2 have reflectivities of about 15 dBZ. The antenna elevation is 0.9o and the range marks are at 20 km intervals. A time period of 74 min is shown. Times are (a) 2226, (b) 2242, (c) 2252, (d) 2302, (e) 2317, (f) 2333 UTC.
 

Fig 2.8. Sample display showing precipitation echo and nowcast polygon for forecast and valid times. This is an example of experimental 30-min nowcasts of thunderstorm location being sent to Denver's Stapleton Airport Control Tower. The large polygon is the nowcast region. The smaller thick lined polygon is the region where 30 dBZ echo was forecast. Echo intensity levels are shown at 10, 30 and 45 dBZ. Range marks are at 20 km intervals. The two intersecting lines represent the runways at Stapleton Airport. (a) Precipitation echo and nowcast polygon location at forecast time, (b) precipitation echo 30 min later than (a) at validation time.
 

Fig 2.9. Doppler radar display of two tornado-producing supercell storms in north-central Oklahoma. The elevation angle is 1.5o and range marks are at 40, 60 and 80 km, azimuth marks are at 0, 20 and 40o. a) Radar reflectivity display showing two hook type echoes associated with the mesocyclones and b) Doppler velocity display of two tornadic mesocyclones at the same time and location as a). The dark shades represent Doppler velocities toward the radar and the light shades away from the radar.
 
 

Fig 2.10. Display of a Colorado macroburst-producing storm. The antenna elevation angle is 0.2o and the two range marks are at 60 and 80 km. a) Radar reflectivity factor, b) Doppler velocity display at same time as a). Maximum velocities within the black region are 44 m/s.
 
 

Fig 2.11. Doppler velocity display showing the evolution of a Colorado microburst. The antenna elevation angle is 0o and the white range marks are at 20 and 30 km. Approaching velocities are darkly shaded and receding velocities lightly shaded. A 7 min time period is represented. The maximum approaching velocities in d) are 19 m/s and the maximum receding are 6 m/s over a distance of 3.3 km. a) 1641, b) 1643, c) 1646, and d) 1648 MDT.
 

Fig 2.12. a) Black and white representation of the color display called the TDWR Geographic Situation Display. It is used by FAA supervisors for tactical planning in the airport terminal area. Precipitation intensity is represented here by shades of gray (actual display is in color). Windshear and microburst warnings are shown by open and closed ovals, respectively. The number in the center of the oval is the radar estimated maximum head to tailwind difference. The runways are shown by the thick solid lines in the center of the display. If a runway is shaded solid black it is affected by a windshear event. Range rings are shown in 5 nm intervals from the airport center. b) alphanumeric display corresponding to a) which displays wind shear alerts for each runway for the local controllers to relay to pilots.