Meteorological radars are divided into two groups, WEATHER RADARS AND WINDFINDING RADARS, the first group designed to detect precipitation such as rain, snow and hail, the second aimed at the measurement of wind velocity, direction and altitude.
Development of radar for use in the field of Meteorology began in the Decca Radar Company in 1952. Unwanted ‘clutter’ from precipitation, obscuring echoes from aircraft and shipping had been a problem for the Marine Radar Type 159, also for Airfield Approach Radars then under development. Seeing these phenomena in a positive light the Storm Warning (Weather) Radar was conceived.
The 159 Marine Radar was modified to have longer pulse duration, a lower prf (pulse repetition frequency) and a much slower antenna rotation rate, thus allowing improved detection and unambiguous display of precipitation echoes out to a much greater range.
The Antenna construction for the Type 40 utilised that of the type 159, a ‘Double Cheese’ with a relatively wide vertical beam facilitating the excellent display of short-range precipitation. But its weakness was, that at the longer ranges only precipitation areas of considerable vertical extent and intensity were detected and displayed consistently. The equipment was then marketed as a ‘Storm Warning’ Radar, mainly for service in the tropics.
X-band radars provided the most efficient detection of precipitation, so new equipments in the band were soon to be developed. The first of these new radars was the Type 41 which employed a variant of the 14ft. single curvature antenna designed for the Type 424 (Airfield Control Radar) radiating a horizontal beam-width of 0.6 degrees and a vertical beam-width of 4 degrees – this allowing improved detection of precipitation of reduced intensity and a vertical extent to greater ranges than the Type 40. The display was provided with a (slightly optimistic) maximum range scale of 250nm. The aerial had a tilt facility that would permit some investigation of the vertical extent of the areas of precipitation. This facility was controllable and observable by the PPI operator.
Product improvement of the Type 41 was to see the incorporation of the later ‘orange peel’ antenna reflector (from the 424) radiating a 2.8-degree vertical beam with an increased tilt range, further improving echo height investigation by the PPI operator.
A new 75KW Transmitter/Receiver was developed for weather radar use and the increased power output facilitated improved detection performance. This Tx/Rx was too large to mount on the antenna and was supplied in a wall-mounted cabinet for installation in the antenna support building with a waveguide connection to the rotating joint of the antenna/turning unit. The new equipments were designated TYPES 42 and 42A. The Type 42 was designed with an antenna rotation rate of 10rpm and a tilt range of –2 to +28 degrees. The display was a 305mm diameter rotating-coil PPI. The Type 42A incorporated a 305mm fixed coil PPI with full off-centering facility. The maximum display range of both types was 400km. A data recording system (‘Radargraph’) could also be provided to maintain a permanent record of echo distribution on electro-sensitive paper.
The Type 43X was designed to provide both PPI and HRI presentation. To present clear and accurate information of the vertical structure of precipitation echoes on the HRI it was necessary to have the narrowest possible vertical beam and to achieve this the antenna was mounted with its long axis in the vertical, providing a 0.6degree beamwidth in that plane. Vertical positioning or scanning over a range of 35degrees was driven by a hydraulic actuator with operation control from the master control unit. Continuous rotation at 10 or 20rpm or manual azimuth setting was under operator control.
CAPPI – Constant Altitude Plan Position Indication
The Radargraph and Cappigraph were recording systems developed for use with weather surveillance radars to provide a permanent visual record based on the data shown on the PPI. The Cappigraph would simultaneously record information from four different altitudes (5,10,18 and 30,000 feet), employing 7”diameter synchronously revolving charts.
The excellent reflectivity of X-band radiation by precipitation also leads to an attenuation of the radar signal as it passes through areas of precipitation. For this reason the intensity of precipitation lying beyond other areas of heavy precipitation become unreliable and serious quantitive measurement of intensity is almost impossible. Increasingly, meteorologists wished to use weather radars for the assessment of precipitation intensity and accumulations over wide areas. To avoid the reduction of accuracy, due to signal attenuation, it was necessary to use a longer wavelength and the S-band TYPE 43S was developed to meet the new requirements. The lower reflectivity of precipitation at S-band requires a considerable increase in transmitter power and the need to radiate an acceptable beam-width required a larger antenna. The Type 43S used a 750kw Transmitter (a version of ‘LOTUS’) and a low noise factor receiver to achieve its excellent level of performance with the 3.66m diameter dish radiating a 2degree beam. The antenna drive system used electric motors to provide rotation speeds of up to 6rpm and the antenna could be set to any selected azimuth angle for elevation scanning over angles up to 90degrees in 6degree steps, using an hydraulic drive.
The virtual freedom from signal attenuation at S-band made the 43S particularly suitable for use in regions subject to severe tropical cyclones. The 43S was also used in the UK’s River Dee Weather Radar Project, which was collaboration between the Meteorological Office, Water Resources Board, the Dee/Clwyd River Authority and Plessey Radar. This project combined data from the radar and a network of rain gauges to investigate the accuracy of radar for the determination of precipitation rates and the calculation of precipitation amounts over large areas.
The River Dee Project demonstrated the value of radar measurements, but also indicated that improvements could be possible if a narrower vertical beam-width was used. It was expected to reduce the errors introduced by echoes from the higher ground in the catchments area and the inclusion of echoes from higher levels at which precipitation in the form of sleet or snow contaminated the echoes from rain. To advance the investigation Plessey Radar undertook a conversion of the radar system to C-band operation. This reduced the radar beam-width to 1 degree and led to a considerable improvement in overall accuracy. The higher reflectivity of precipitation at the shorter wavelength meant that excellent detection performance was achieved with a transmitter power of only 250KW.
The success of the Dee Weather Radar project helped to prove the value of radar as a means of determining precipitation amounts over extensive areas, also the suitability of C-band for this application. A decision was made within Plessey Radar to develop two new C-band weather radars, building on the experience gained. The two new radars were the TYPE 45C, with a 3.66m diameter dish and the TYPE 46C with a 2.5m diameter dish. The radiated beam-widths were 1 degree and 1.5 degree respectively. The Type 45C antenna assembly incorporated balancing counter weights and was protected from extreme wind loadings by a radome. For the Type 46C,balancing fins were fitted allowing the system to operate safely without radome protection. Both equipments included a 250KW transmitter and maximum use was made of solid-state technology.
The first equipment in the new C-band range incorporated a Fixed-Coil Display, which could be used in either PPI or HRI format. A new microprocessor based display system (Colour Scan) was also developed to provide plan and vertical cross section presentation of precipitation intensity using six colours on a 500mm monitor. Alternative computer-based data processing systems also became available. Type 45C and 46C weather radars were delivered worldwide to Meteorological and National Defence Services. In the U.K. radar coverage for the London/Thames area was provided, also for the North West Water Authority.
Subsequently, starting from the late 1980’s, the type 45C based UK weather radar network was progressively updated and new radar systems added to increase coverage of the British Isles. Twelve radars formed the national network as shown on the following map. Data from the remote radar sites was fed to the UK Meteorological Office Centre (Braknell) for integration into its ‘Frontiers’ processing system for analysis. Weather products were then provided on a commercial basis to a wide range of end users. (e.g. BBC, Water and River Authorities, major sporting and other events, airports, local authorities etc.)The integrated data was also provided to other users in Europe, as part of the E.E.C. ‘COST73’ programme, to develop a European wide integrated weather radar network. The company was an industrial member of this programme and the achievements were published (see ‘Weather Radar Networking, Kluwer Academic Publishers Ref. EUR 12414 EN-FR; 1990).
The main features of the updating were to maximise radar stability, provide a Doppler facility and add a new generation of signal and data processing, called RADEX, at each radar site. (see pages 121 and 122). RADEX pre-processed the raw weather data (which can be immense in quantity) before onward transmission (using encription and high speed lines) to the Meteorological Centre. RADEX also provided colour displays of weather analysis at the radar site and, where required, for local users. In applications outside the UK, Plessey Radar offered a system under the name of SUNrise, this to undertake the processing and provision of weather analysis products on colour displays at the Meteorological Centre.
The comprehensive range of weather radar equipments supplied by the Radar Company was completed by a new S-band system; the TYPE 45S. This equipment combined the design and control capabilities of the Type 45C and 46C systems with the latest development in S-band technology, replacing the type 43S.
Windfinding Radars operate by tracking a Radar Reflector carried aloft by a balloon filled with hydrogen or helium. If the slant range, elevation and azimuth angles obtained from the radar are recorded at regular intervals the speed and direction of the wind at all levels throughout the ascent can be calculated. The first radar equipments used for this purpose were modified military tracking or gun laying systems.
In 1956 Decca Radar developed the Type WF 1, the first commercial radar designed specifically for the windfinding task. It was an X-band system with a 20KW transmitter. The cabin housed the operator(s) and the electronic equipment. The 2.4m diameter antenna system was mounted on the front of the cabin with its elevation angle controlled manually by the operator using a hand wheel within the cabin. The azimuth angle of the assembly was also controlled manually by the operator using a second hand-wheel within the cabin. The antenna was fitted with a ‘shepherd’s crook’ feed and radiated an offset 1 degree beam. The feed was rotated to provide a conical scan. The signals returned from the balloon-borne reflector were displayed on a novel display; the I-scope. When the antenna was correctly aligned to the target the display showed an annular echo at a distance from the display center corresponding to the target range. When the antenna was not correctly aligned the echo trace showed a break. The task of the operator was to steer the system back on target using the two hand-wheels, so as to restore a complete annular trace. The elevation and azimuth angles were shown on dials in the cabin. The operator also had to determine the range of the target by aligning a range marker to the annular echo trace using another (smaller) hand-wheel. The I-scope could be switched to display a range of +/- 2000m, with the range ring (or strobe) appearing half way along the displayed radius. The strobe range was indicated on another dial. The cabin accommodated the tracking operator and an assistant to record indicated values at regular intervals.
The prototype of this new radar was first tested at the Meteorological Office, Upper Air Station at Hemsby, Norfolk and was then operational at the ‘Second International Comparison of Radio Sondes’ at Payerne, Switzerland, successfully tracking the long train of sondes launched over the period.
The WF 1 was a very successful design and equipments were supplied to Meteorological services throughout the world. After some years a requirement existed for a system capable of obtaining upper wind data to greater altitudes, to meet the increasing demands of aviation. To address this the Type WF 2 was developed, in 1960, by the Decca Radar Company. This equipment would be powered by a 75KW transmitter with a considerably increased range over the WF 1. The Transmitter/Receiver was mounted on the back support of the antenna, so reducing the waveguide run losses. Operator comforts were improved (glass-fiber insulation and air conditioning) and their task simplified. The indicators could be stilled temporarily to simplify data recording, with automatic update resuming immediately. The ‘shepherd’s crook’ antenna feed was replaced by a more efficient modified Cutler feed.
For both the WF 1 and WF 2 it was necessary to visually align the antenna to the target using a site mounted on the antenna framework, until the target could be identified on the I-Scope. The task was made easier on the WF-2 by providing a central section of the antenna dish, which could be moved forward toward the feed to radiate a wider (4.5 degree) beam during the initial search phase. Having gained target identification on the I-Scope the central section would then be moved into the profile of the main dish to restore the 1-degree tracking beam.
The increasing cost of supporting staff at remote sites and the requirement of additional sites to extend the international network of upper air observation stations showed the need for a system capable of tracking balloon-borne reflectors automatically and with the capability of adding fully automatic data processing. Low initial costs and minimum installation, operational and maintenance costs were also important. To meet these requirements the Type WF 3 windfinding radar was developed, incorporating a number of new design concepts. The initial cost objective was to produce an automatic tracking system at or below that of the optical Pilot Balloon Theodolite System used at many remote upper air stations. Although this objective proved to be unachievable, the attention paid to all aspects of the design resulted in the development of a relatively low cost radar system, which fully met all the other requirements.
RADIOSONDES - Additionally, a weather balloon was able to carry a ‘radiosonde’, which transmitted temperature, pressure and humidity data to the ground station for the duration of the ascent. A ground processing system, such as RADTRACK (see 11.11), would then produce messages in civil meteorological formats. WF-3M was a mobile military version of the product, which produced messages for the artillery and when fed into the ballistic equation they facilitated missile ranging. Project ‘Firefly’ was a major contract for 10 such mobile convoys with a further system for a missile test site. In this case the ground processing was produced in conjunction with the Beukers Corporation of the USA. As customers undertook several launches per day it resulted in a significant business supplying balloons, reflectors and radiosondes over many years. The radiosondes were designed and manufactured by the Viz Corporation of Philadelphia USA.
The WF 3 consisted of two basic units; the Antenna Unit and the Control and Data Unit, these two units were interconnected by cable.
The antenna unit contained a 0.9m parabolic dish antenna with an offset rotating Cassegrain sub-reflector to provide a 2.3degree conically scanning beam. The transmitter/receiver was mounted on the rear of the main antenna to minimise the waveguide run. The complete assembly was protected by two hemispherical radomes to balance the wind loading. A hand-held control unit was provided to enable the operator to align the antenna to the target, using a sighting tube mounted on top of the antenna unit.
The Control and Data Unit contained simple operator controls and displays for use to complete the initial target acquisition and during subsequent automatic tracking. All the electronic control and tracking units were mounted on printed circuit boards housed within the unit. A servicing kit, containing spare circuit boards was provided with the equipment to allow first line servicing to be undertaken by board replacement on site, without the need to call on the services of a skilled technician. Faulty boards could then be sent for repair at a central base or returned to the factory.
Various system configurations could be provided. In the simplest version the antenna unit was mounted on tripod legs of sufficient length to support a tent, which housed the Control and Data Unit fitted immediately below the antenna. The tent provided shelter for the system operator. No other infrastructure was required apart from a 1kVA power source, which could be a portable generator at sites where mains power was not available. In the more widely used configuration the antenna was mounted on the roof of a suitable building. The control and data unit could be mounted in the same or another building.
The final development in the X-band Windfinding Radar Series was the WF-33. This resulted from the ‘Aladdin’ project (supported by the Met. Office) under which early WF-3 equipments were shipped back to the factory for refurbishment and at the same time underwent the up grading of the antenna unit and replacement of the control unit. The new RADTRACK micro-processor based system then provided data that could be recorded on a printer or fed to a computer for calculation of the wind speed and direction at all levels of the ascent. The optional WINDPROC PT system was also available to meet this requirement and to generate pilot messages.
The RADTRACK and WINDPROCT PT systems were designed and built for Plessey Radar by ASMAP Electronics, an Isle of Wight based company.
It was in the early years of MET RADAR development that the Australian Bureau of Meteorology sought tenders for the design and supply of a sophisticated product that would operate as both a WINDFINDER & WEATHER RADAR – The Decca Radar Company were awarded the contract.
The product entered the market as the WF-44. While the system was to comprise many proven elements its integration was not straightforward. In 1962 the Heavy Radar Group were operating in part at Davis Road and in part on the Isle of Wight. Control and Processing circuitry was to be designed by a ‘Radar Dev’ team based in Surrey (Davis Road No.1) but none of those engineers were prepared to move to the Island. To resolve this problem a team of IOW based engineers took on the task of meeting the terms and conditions of the contract. This was further exacerbated because, in 1964, Decca were in the middle of selling-off their Heavy Radar Group (HRG) to the Plessey Company and the AR-1 programme was in full swing.
It was 1965 when the first of twelve WF-44 equipments went into service with the Australian Met. Bureau, the product having the ability to successfully operate in all the harsh conditions that a country as vast as Australia could present.System Parameters:
The 650Kw Tx/Rx was to be a version of the LOTUS transmitter. It would operate in the band 2800-2900MHz at a prf of 400pps (wind-finding) or 275pps (in the weather radar role). Pulse length could be set to 1.5 or 0.35 microseconds. A travelling wave tube amplifier gave a receiver noise factor of 6dB. Automatic frequency control and swept gain (with appropriate laws for wind-finding and weather radar applications) were provided.
The antenna was a 2.4m.dish fed by an offset ’Cutler’ feed radiating a 3degree conical beam offset by 1. In the wind-finding mode the aerial feed was spun at 1400rpm to provide the required conical scan with the aerial positioning being performed by 2 reversible motors. A number of preset automatic scanning modes were available together with an automatic target location facility providing automatic switch-over to the tracking mode. The Master Control Unit incorporated all necessary controls and indicators together with a special display that could be switched to A-scope, A/R scope or 4R-scope for monitoring or manual tracking, should such assistance be required at any stage.
System Details: - In the weather radar mode the feed was locked in a pre-determined position, with aerial rotation at 10rpm driven by a more powerful motor, allowing precipitation echo presentation on the PPI. The aerial could also be set to any azimuth angle with continuous nodding at 10 degrees per second over an elevation angle range from 0 up to 90degrees for echo presentation on the HRI (Height Range Indicator). The PPI and HRI were fixed coil displays of the MK 5 type with full off-centering and electronic range and bearing or height line presentation.
The radar had several search programmes for locating the target/balloon and automatically locking onto it. The product had digital readout of range, azimuth and elevation with a highly accurate method of target range measurement. Automatic punched tape recording of the process was part of the sub-system. The balloon tracking capability was in excess of 200KM.
- Frequency – in the Band 2800-2900MHz
- Aerial – 2.4m double skinned parabolic dish
- Tilt – -2degrees to +95degrees
- Azimuth Rotation – 10rpm
- Conical Beam – 3degrees
- Spinning Rate – 1400rpm
- Peak Power – 650Kw
- Pulse length 0.35 and 1.5µsec.
- PRF – 400/pps in Windfinding 275/pps in Weather role
- Ranges – 300m to 200Km
- Windfinding – 50,100,200,400 K/h PPI/Weather role
- Displays – 127mm 4A/R scope in Windfinding role
- 305mm PPI/ in Weather Role (optional)
- 305mm HRI/Weather Role (optional)
COLOURSCAN (A Colour Display and Processing System for Meteorological Radars)
Weather radars had been in use since the 1950’s for quantitive assessment of meteorological phenomena measurable by radars. In the late 1970’s research had been directed towards providing, in a clear manner, quantitive analysis of rain echoes from radar for short-term forecasting and hydrological data. There were three main applications for this information. The first and most frequent application was for storm warning, particularly regarding aircraft safety. The second was short term predictive forecasting, involving the projection of current measurements for the immediate following hours and, in conjunction with satellite data, over wide areas not accessible by radar signals. The third application was in hydrology, involving the measurement of water content reaching the ground, and control of the water resources thereafter.
The purpose of the COLOURSCAN equipment can best be summarised as providing a clear and unambiguous visual display of the size, shape, intensity and location of rainstorms, which could be assessed by meteorologists without recourse to interpretation by skilled radar operators. At the summary level this was achieved by taking the raw video from the weather radar together with range, azimuth and elevation data, extracting the data specific to rain and integrating it to produce the average rain density in individual cell areas. The size of these cells was operator selectable and six levels of colour imaging were available for display on a standard television colour monitor.
Until the late 1970’s the only method of displaying the radar picture was using conventional circular CRT displays, which required skilled interpretation. Archiving and dissemination of the data required either photography and facsimile equipment or verbal reporting procedures, thus severely limiting the analysis that could be applied.
The following simple block diagram describes the configured Colourscan equipment.
The Colourscan equipment replaced the CRT display within a weather radar installation. Colourscan allowed a more flexible approach to the archiving and dissemination of the weather data. The equipment comprised two main units, a Processor and a Raster Scan Colour Monitor. The Colour Monitor was a conventional 625 line high quality studio monitor whose inputs were typically, Blue for the lowest density of precipitation, the next was Green, then Orange and Red for the very strongest precipitation along with a non-interlaced TV sync-pulse. (The Met. Office eventually took the design over and ever since has displayed their weather forecasts based on our Colourscan.) The system provided 256 active TV display lines enabling the construction of processed pictures containing 256 x 256 of colours representing the average intensity of the rainfall at the location on the screen.
The Processor Unit comprised four sub-systems (a). Radar interface (b) Processing sub-system (c) Colour Monitor Drive Interface (d) Operators Control Panel. Signal processing was based on the Plessey MIPROC 16bit microprocessor which analysed radar signals to provide the data for colour presentation. The Processor facilitated the formatting of archived recordings and data dissemination over telephone lines, enabling a flexible approach to user configurations. The programme for Colourscan was stored in PROM, thereby eliminating costly RAM storage and disc/tape readers. The programme contained both Diagnostic and Operational software.
The Radar Interface accepted radar antenna position data, transmitter trigger timing pulses and raw video. The Interface performed video signal conditioning and digitisation followed by pre-averaging in azimuth, this to reduce the data rate to proportions manageable by the processor. The Processor controlled the operation of the interface and accepted the data transfers from the interface to the processor via a 16bit parallel interface. The Interface also contained built-in test and calibration functions controlled by The Processor during operational use. The colour monitor drive sub-system comprised two 16K x 16K dynamic RAM stores; one was dedicated to continuous refreshing of the colour monitor and the second was used as a back-up store to the processor during the operational cycle. The sub-system provided synchronous control of the stores and TV sync-pulse generation, providing control of address and data highways under instruction from the processor. The operator could communicate with the system via a dedicated control panel, which had been laid out to conform with the familiar equipment practices present in the radar. This eliminated the reluctance of a traditional operator confronted with conventional computing input/output protocol.
The Colourscan display was to replace the conventional CRT display. Three-dimensional presentation was not possible at the same time so two options were made available. The display could either provide the equivalent of a PPI for azimuth scan display, or as an HRI elevation data. PPI presentation provided 256 x 256 cells with each cell representing the average rainfall rate over a square area of 0.5, 1, 2, or 5Km. This defined the range displayed according to the cell size chosen by the operator. When operating as an HRI the cell size was fixed at 0.5Km in range and height and the display consisted of two horizontal strips across the screen, each representing 25Km height. The top strip showed ranges from 0 to 128Km and the bottom strip showed ranges from 128 to 256Km maximum range.
The product was typically used with the 45 and 46 C-band and S -band Radars throughout the U.K., Saudi Arabia and Brazil where master and remote stations were configured.
RADEX AND SUNrise
As the technology advanced and customer requirements for weather forecasting products increased, in both volume and complexity, the Company licensed a digital processing and display system, (complete with application software) from Lassen Research (California USA). Lassen Research were recognised as world leaders in the development and implementation of Doppler weather radar signal and data processing. The system, located at a radar site, was designated ‘RADEX’ and that at a Meteorological Centre, where all remote radar site data was received and integrated into ‘Network Data’, was designated SUNrise. All the UK weather radars (Type 45C) were updated with RADEX, which compressed the vast amount of raw weather radar returns available at each radar site for transmission to the Meteorological Centre and its ‘Frontiers system’, as well as providing weather forecasting products for marketing to users local to the radar coverage. RADEX was also exported to update overseas weather radar networks. (e.g. Italy and Indonesia).
More details about the RADEX and SUNrise sub-systems.
By converting the 45C into a Doppler weather radar it met the requirements of the next generation of weather radars as demanded by international requirement specifications. This enabled the RADEX/SUNrise processing system to produce the full range of real-time weather analysis products to service the ever-increasing users of meteorological information.
Doppler weather radars provide data on the velocity as well as the intensity of precipitation, enabling processing systems to identify, for example, areas of windshear and to distinguish moving precipitation echoes from static ground echoes.
In a period of widespread precipitation, a Doppler weather radar collects a massive amount of return echoes for signal and data processing. This becomes a large three-dimension database, as weather radar can be controlled to varying scan patterns (by sectors and elevation of particular interest) mixed in with routine 360degree scans. There is, as an example, an analysis product called CAPPI (Constant Altitude Plan Position Indicator). This provides a colour display of precipitation, at any selected altitude, by taking a cut through the system data base.
The RADEX system was specifically designed to be located at remote, unattended, radar sites, as the norm. It therefore had full system remote control. SUNrise performed the same functions as in RADEX and provided an extended range of analysis products. Also, in a weather radar network it would integrate the databases from individual radars and provide regional or national analysis for the forecasters.