P1.2 NEXRAD OPEN RADAR DATA ACQUISITION (ORDA
Contents
1. 10 3 CW Substitution CW Substitution is used to verify the receiver gain using a calibrated external signal generator The signal generator injects CW at a known power level into the receiver at 2 points allowing comparisons with the internal RF Generator This checks accuracy of the test path the front end and the receiver cabinet path 11 CONCLUSION Calibration provides verification that reflectivity accurately compensates for detectable system changes It also provides insight into hardware behavior and improves troubleshooting capability Automating as many tasks as possible reduces the chance for error from human input and speeds test completion improving concentration on the problem instead of the test The Off Line tests build upon the foundation formed by the On Line tests allowing improved analysis and increased data for diagnosing problems 12 ACKNOWLEGEMENTS The authors would like to thank Dale Sirmans and William Urell for their support in writing this paper Note The views expressed are those of the author s and do not necessarily represent those of the National Weather Service 13 REFERENCES Free A Patel N Heck A 2004 ORDA Internal Report ORDA System Calibration Free A 2004 ORDA Internal Report Legacy vs SIGMET dBZ Calculation Rinehart Ronald E 1997 Meteorologists 3 edition Publications Columbia MO Radar for Rinehart Ice R McGehee T Rhoton R2. This Noise Temperature represents the contribution of noise in the receiver signal path from the receiver protector to the IFD The noise added by waveguide components feedhorn antenna side lobes and the radome are all taken into account in the cold temperature Te Noise Temperature is extremely sensitive to receiver changes therefore it is an ideal parameter for use in early detection of failing components 6 REFLECTIVITY CALIBRATION CONSTANT dBZ The ORDA system outputs 1km reflectivity moment data that is computed using equation 3 P dBZ Ole a 20log R AxR dBZ 9 Where Ppr is the return signal power N is the Noise value corrected for elevation R is range A is the two way atmospheric loss and dB is the system calibration constant computed using equation 4 dBZ represents the reflectivity of a OdB Signal to Noise target at a range of 1km and includes all the constants in the radar equation Rinehart 1997 10 2 8 3 6 ET 2 xIn 2 x 22 x10 x10 x10 aa T xP xG X0 XcxTX K xL xL xL 8 4 Where is the wavelength in centimeters P is the received power in milliwatts G is the antenna gain is the beamwidth t is the pulsewidth in microseconds K is refractivity L is the receiver waveguide loss in dB L is the transmitter loss in dB Lg is the receiver detection loss in dB N is the system Noise in milliwatts and g is the receiver gain 6 1 Transmitter Power P The tr
3. Saxion D Warde D Guenther R Sirmans D Rachel D 2005 Radar Operations Center ROC Evaluation of New Signal Processing Techniques for the WSR 88D 21 International AMS Conference on Interactive Information and Processing Systems for Meteorology Oceanography and Hydrology Heiss W McGrew D Sirmans D 1990 NEXRAD Next Generation Weather Radar WSR 88D Microwave Journal Patel N and Macemon B 2003 NEXRAD Open Radar Data Acquisition ORDA Signal Processing amp Signal Path 20 International AMS Conference on Interactive Information and Processing Systems for Oceanography and Hydrology Meteorology Patel N Free A Jim G 2005 NEXRAD Open Radar Data Acquisition ORDA Receiver Characteristics 21 International AMS Conference on Interactive Information and Processing Systems for Meteorology Oceanography and Hydrology Sirmans D Urell W 2001 Radar Operations Center On Measuring WSR 88D Antenna Gain Using Solar Flux SIGMET 2004 RVP8 User s Manual
4. are 80dB and 85dB respectively Legacy short long pulse Noise measurements were approximately 57dB Depending on system waveguide configurations and other receiver path components Noise measurements from system to system will vary 1 2dB 5 NOISE TEMPERATURE Noise Temperature is related to the system noise figure see Figure 1 Noise Temperature is best measured by injecting different noise source levels at the input of the receiver Figure 1 ORDA Noise Figure The NOISE test signal is injected at the front end to compute a noise power level Noise with a calibrated signal Another measurement Noise st with no injected signal is also taken The ratio of Noiseo to Noise along with the known Hot and Cold source temperatures gives the noise temperature as shown in equation 1 The cold temperature Tc value is the temperature of the antenna The hot temperature Ty is given by equation 2 using the computed Excess Noise Ratio ENR of the NOISE test signal and is calculated at the receiver front end i e the receiver protector Noise pn Ty T Noise syy NoiSe remp 1 Noise gn 1 Noise of 5 Tz 290 10 1 2 ORDA Noise Temperature is 250K to 270K Legacy measurements gave values of 550K to 600K due to an error in the Legacy equation However the absolute values are not important in themselves Noise Temperature is an excellent indicator of changes in component performance and or system calibration
5. 4 89 60 40 20 PON POr PAOA T Frnt EPO ptr Fan Dre Figure 2 Receiver Transfer Curve Problems in Dynamic Range can indicate saturation problems in the active receiver components Noise problems will indicate problems in the lower end of the receiver curve 8 VELOCITY SPECTRUM WIDTH In addition to varying the CW test signal for measurement of system Linearity and the Reflectivity Calibration the CW test signal is phase shifted in reference to the COHO to simulate Doppler shifts on the test target returns Using this technique the following velocities are sequentially simulated 0 velocity 1 4 V Nyquist 3 8 VNyquist 5 8 VNyquist The velocities are simulated by progressively phase stepping the test targets produced by the frequency generator Applying a dither to the successive phase simulates spectrum width A dither of 13 is applied to give a constant spectrum width of 3 55 m s with a Vnyquist 28 m s The ORDA signal processor uses the input test signal to verify RF generator phase shifter operation and to verify signal processor phase reference and processing consistency With ORDA two types of velocity and spectrum width computations are performed The phase shifted CW test signal is used in off line calibration During on line calibration a predefined I Q data stream is input to the processor for verification of the velocity and spectrum width algorithm Velocity Spectrum Width problems are rare and
6. P1 2 NEXRAD OPEN RADAR DATA ACQUISITION ORDA RECEIVER CALIBRATION Alan D Free SI International Norman Oklahoma Adam K Heck and Nita K Patel RS Information Systems Inc Norman Oklahoma 1 ABSTRACT The Open Radar Data Acquisition ORDA System will replace the Legacy Radar Data Acquisition RDA system that is currently part of the NEXRAD fleet Many Legacy calibration parameters e g SYSCAL have been replaced with new system calibration parameters e g dBZo with the ORDA system This paper provides a description of test methodology for automatic on line calibration and off line calibration algorithms When appropriate differences to Legacy parameters will be identified and explained 2 INTRODUCTION The NEXRAD Qpen Radar Data Acquisition ORDA enhancement program provides automatic on line calibration to maintain radar performance and provides enhanced off line calibration routines to support system maintenance alignment and troubleshooting Patel et al 2003 Calibration is performed during system startup periodically in standby and during Volume Coverage Pattern VCP retrace i e transition from end of a particular VCP to the start of the next VCP Most calibration routines use one of several test signals available in the WSR 88D architecture The four test signals shown in Figure 1 are Continuous Wave CW constant level test signal at IF frequency RF noise NOISE broadband noise so
7. a position and check antenna gain This external source s position relative to the earth is well known and accurate and observatories like Penticton calibrate its noise output in the WSR 88D frequency band measured in solar flux To check antenna gain we first use the solar flux data and convert it to sun noise temperature referenced to the receiver front end This conversion accounts for beamwidth polarization assumed antenna gain wavelength sun earth distance and path loss from antenna port to receiver front end Next we must measure the sun temperature as seen by the receiver We use the following equations to calculate the sun temperature then Where Tc is the Blue Sky Antenna temperature measured at the receiver front end Pc is the Antenna power level is the Noise Source temperature at the receiver front end Py is the Blue Sky with Noise Source on power level Ts is the Sun temperature Ps is the Sun power level and Pye is Sun with Noise Source on power level Tr is the receiver Noise temperature and g is the receiver gain from the receiver protector to the IFD The calculated to measured ratio gives an estimate of the antenna gain correction and is dependent on correct system calibration 10 2 Error Estimate The error estimate is an off line procedure that can quickly pinpoint dBZO problems It is completely automated and gives information on changes in noise gain linearity and transmitter power
8. ansmitted power directly affects system calibration Along with Noise N and receiver gain g transmitted power can fluctuate during system operation therefore P is measured periodically Transmitter Power is measured during the surveillance cut of a VCP and also during the Performance Check During a VCP cut the system measures the average power of the transmitter once a second The samples are averaged together and corrected for the duty cycle to obtain the transmitter peak power During the Performance Check a similar procedure is used however samples are obtained at one second intervals with the antenna parked The transmitter has not changed in the ORDA architecture and is the same as that available in the Legacy WSR 88D configuration Nominal transmitter peak power is 700kW Because the WSR 88D system uses an average power meter to measure power the zero level is important Power Meter Zero is done during on line calibration to measure the slight positive bias on the power meter thereby ensuring accuracy of power measurements when radiating 6 2 Sensitivity lo System sensitivity N g represents the power level of a OdB Signal to Noise target measured at the receiver input This term is called b and also represents the system s Minimum Discernible Signal MDS b and Transmit Power are the only variables of the dB equation that change regularly and need to be updated bis calculated by the Linearity routine and pro
9. e IFD through the OaB test attenuator setting to obtain a reference point Based on the reference value the 103dB steps of the test attenuator are calibrated using the internal CW test signal and by varying the injection point between the cabinet and the front end 7 2 Dynamic Range ORDA software automatically executes Dynamic Range measurements every retrace period a capability not available in Legacy RDA software Using the built in CW test signal stepped through all 103dB steps of the test attenuator the system s 1dB compression point is identified The Dynamic Range is then computed as the delta between b and the 1dB compression point Figure 2 Free 2005 shows the ORDA off line linearity and reflectivity test measurement display This calibration test computes the system noise floor compression point minimum detectable signal linearity and dynamic range and shows the results in a graphical window This test run was done in short pulse at the KCRI channel 2 test bed system in Norman OK As shown here the Dynamic Range from OdB S N to 1dB compression is given as 95dB The nominal legacy value for Dynamic Range is 93dB Lineerity b Metiectinity Eixis Aiie Mom Stor Pace Paraereters Rache Constant Owns Cw Freee Loss CW Obret Lact Matched Neer bna Wein Armei Referrers Poin basg Puse Frot Ero fo Aroeiver Curve Resets Coevgre st ier Dyrare Range Notes Mare Devete Update atipun Dwa
10. tes Short Pulse Long Pulse KD Check KD power ine z z e a m D emerse y fowsubsituion f fa outer Suppressor a Suncheck Az El Offsets Antenna Gain 4 SYSTEM NOISE System Noise measurements provide a measure of the receiver path Noise is measured with the transmitter off i e non radiating When in standby Noise is measured with the antenna in parked position 0 azimuth and 23 elevation Operationally the measurement is only made with the antenna above 3 5 in elevation to avoid bias due to ground noise No test signals are injected during noise measurements In ORDA the System Noise is measured in both short and long pulse to obtain accurate system calibration for both short and long pulse VCPs This is different from the Legacy computation in that an offset was added to the short pulse measurement to obtain the long pulse noise value ORDA uses digital matched filtering therefore the filter parameters are specified for both short and long pulse independently Since the matched filter loss is a contributor to the system noise measurement independent computations are made for the differing pulse widths System Noise is obtained by computing signal the power of approximately 5000 I Q samples along a span of 10 radials The System Noise parameter is adjusted by an elevation scale factor representing the influence of ground noise on the reading ORDA Noise measurements for short and long pulse
11. urce RF test signal RFD sample of RF drive signal input to the Klystron Klystron Delayed KD 10 us delayed sample of transmitter output Corresponding author address Alan D Free Sl International 2227 W Lindsey Ave Suite 1500 Norman OK 73069 e mail Alan D Free noaa gov 3 CALIBRATIONS Table 1 details which procedures are run during the on line calibration during VCP retrace and periodically during standby Free et al 2004 The table further details the tests available through the off line System Test Software STS Calibration of the WSR 88D from transmitter to receiver consists of measuring transmitter power measuring system noise floor measuring gain loss of all components in the receiver signal path computing an error offset for reflectivity and measuring antenna system parameters namely antenna gain and pointing accuracy Parameters contributing to reflectivity accuracy transmitter power noise level and linearity are measured during the on line calibration each VCP Off line STS calibration functions provide additional detail and also assist in system verification maintenance and troubleshooting Table 1 On line and Off line Calibration System Noise Short Pulse Long Pulse Noise Temperature Reflectivity Correction dBZ System linearity Linearity Test Attenuator Calibration Velocity Spectrum Width VQ Processing External Phase Shifter Transmitter Power Error estima
12. vides a measure of the linearity of the system as well as the Noise floor of the system Nominal ORDA sensitivity is measured to be 114dBm Correlating Legacy sensitivity is nominally 113dBm 7 SYSTEM LINEARITY This test verifies the linear receiver response Data points along the linear graph are used to determine the calibration constant b used in the dBZ calculation The ORDA design incorporates a staggered linearity test performed during VCP retrace to verify the entire linear operating range That is each VCP will measure 10 points on the linear curve The 10 points will be staggered with points used on subsequent VCPs in order to eventually test the entire linear range For example the first VCP will use points 21 27 33 39 etc the second VCP will use points 22 28 34 40 etc During every Performance Check the ORDA calibration procedures will test the entire receiver transfer curve Deviations from linear response indicate a failing system components or non linear response from the active components The ability to view data points at different attenuation levels is one advantage over the Legacy architecture that only provided data at four discrete test signal levels 7 1 Test Attenuator The test attenuator is used to vary the injected signal level from 0 to 103dqB therefore it is critical each attenuation step is accurately calibrated An automated off line routine allows the technician to inject a signal into th
13. would indicate a phase unstable receiver component This typically only happens when the component becomes non linear and its noise increases 9 KD CHECK The KD Check provides information about the receiver protector and compares the cabinet versus front end signal levels to help isolate path gain loss errors in the front end components versus the cabinet components KD pulse level checks help differentiate between transmitter or receiver problems 9 1 Clutter Suppression Since the KD pulse is a sample of the klystron output it is an ideal test target for testing Clutter Suppression The level of the KD pulse is measured with and without Clutter Suppression enabled in the signal processor The level of suppression and consistency in the measurement gives a high level of confidence in the phase stability of the system Problems here typically indicate problems with transmitter stability Typical Clutter Suppression levels for ORDA is around 55dB Legacy Clutter Suppression was nominally 50dB Ice 2005 Heiss 1990 10 CALIBRATION VERIFICATION Off Line procedures to verify calibration accuracy are integral to the maintenance strategy There are three major procedures used to verify reflectivity calibration Suncheck Error Estimate and CW Substitution Each procedure has a different focus and exposes different calibration problems 10 1 Suncheck Suncheck uses the sun as a noise source to do 2 things calibrate antenn
Download Pdf Manuals
Related Search