LEE, James (4930 Research Drive, Huntsville, AL, 35805, US)
WALKER, William (4930 Research Drive, Huntsville, AL, 35805, US)
BALAJI, Mrinal (4930 Research Drive, Huntsville, AL, 35805, US)
BRASWELL, Rick (4930 Research Drive, Huntsville, AL, 35805, US)
CHANDRASEKAR, V. (4930 Research Drive, Huntsville, AL, 35805, US)
LEE, James (4930 Research Drive, Huntsville, AL, 35805, US)
WALKER, William (4930 Research Drive, Huntsville, AL, 35805, US)
BALAJI, Mrinal (4930 Research Drive, Huntsville, AL, 35805, US)
BRASWELL, Rick (4930 Research Drive, Huntsville, AL, 35805, US)
| CLAIMS What is claimed is: 1 . An apparatus for calibrating a dual polarimetric NEXRAD radar system, comprising: an antenna/radome that includes an antenna mounted electronics (AME) unit with built in test equipment (BITE) for generating and processing a calibration signal; a transmitter circuit; a receiver circuit; and where the AME unit calibrates the system by, determining a reflectivity factor for the system directly from a transmitter pulse, determining a system bias by calculating a vertical polarization path gain from a transmitter, receiver and antenna path, and a horizontal polarization path gain from a transmitter, receiver and antenna path, and determining a system differential phase with a signal processor for the radar system. 2. The apparatus of Claim 1 , where the system bias is determine by calculating the bias in a sun power measurements, the bias in a transmitter power measurements, the bias in a test signal power measurement, the bias in a BITE test signal, and a bias at a transmitter power path. 3. The apparatus of Claim 2, where all of the bias measurements are made during installation of the radar system. 4. The apparatus of Claim 1 , where the bias is determined to within 0.1 dB. 5. The apparatus of Claim 1 , where the reflectivity factor is by determining the combination of a gain of the antenna, the loss of the radome, and the loss of the wave guide, where the combination referenced to an anchor point for the measurement of transmitter power and receiver sensitivity. 6. The apparatus of Claim 5, where the anchor point are horizontal and vertical couplers. 7. The apparatus of Claim 1 , where the reflectivity is determined to within 1 dB. 8. The apparatus of Claim 1 , where the AME is enclosed self enclosed on the antenna. 9. The apparatus of Claim 8, where the AME enclosure contains a heater and cooler. 10. The apparatus of Claim 9, where the AME enclosure maintains an internal temperature of about 25° C. 1 1. An apparatus for calibrating a dual polarimetrie NEXRAD radar system, comprising: an antenna/radome; a transmitter circuit; a receiver circuit; and means for automatically calibrating the radar system that is mounted on the antenna, where the system is calibrated by, determining a reflectivity factor for the system, determining a system bias, and determining a system differential phase. |
FOR
UNITED STATES LETTERS PATENT
METHOD AND APPARATUS FOR CALIBRATION AND ACCURACY ANALYSIS OF A DUAL POLARIMETRIC NEXRAD RADAR SYSTEM
APPLICANT: V. CHA RASEKAR, JAMES LEE, WILLIAM
WALKER, MRINAL BALAJI and RICK BRASWELL
METHOD AND APPARATUS FOR CALIBRATION AND ACCURACY ANALYSIS OF A DUAL POLARIMETRIC NEXRAD RADAR SYSTEM
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application claims priority from U.S. Provisional Patent Application No.
61/371 ,999 for "Method and Apparatus for Calibration and Accuracy Analysis of a Dual Polarimetric Nexrad Radar System" that was filed on 8/9/2010.
FIELD OF THE INVENTION
[0002] The invention relates generally to a method of calibration for a dual polarimetric
Nexrad weather radar system.
BACKGROUN D ART
[0003] Most dual polarization weather radars in the research community such as CSU-
CFIILL, and SPOL calibrate their system by pointing the antenna vertically during conditions of light rain over the radar and rotating it about its own axis (vertical) by integral multiples of 360 degrees. This is the most simple and straight- orward way of calibrating a radar for differential reflectivity, because it calibrates the complete system and provides a reference using polarization symmetry principles. However, this operation cannot be carried out by the Next Generation Radar (Nexrad) WSR-88D system simply because its antenna elevation cannot be increased above 60 degrees. In addition, the WSR-88D needs to calibrate the system Zu, offset once every volume scan. In normal operation, it is not guaranteed to be raining at the radar site. Therefore, an engineering scheme had to be developed to calibrate the WSR-88D system Za r offset after every volume scan, within an accuracy goal of 0.1 dB. SUMMARY OF THE INVENTION
[0004] In some aspects, the invention relates to an apparatus for calibrating a dual polarimetric NEXRAD radar system, comprising: an antenna/radome that includes an antenna mounted electronics (AME) unit with built in test equipment (BITE) for generating and processing a calibration signal; a transmitter circuit; a receiver circuit; and where the AME unit calibrates the system by, determinin a reflectivity factor for the system directly from a transmitter pulse, determining a system bias by calculating a vertical polarization path gain from a transmitter, receiver and antenna path, and a horizontal polarization path gain from a transmitter, receiver and antenna path, and determining a system differential phase with a signal processor for the radar system.
[0005] In other aspects, the invention relates to an apparatus for calibrating a dual polarimetric NEXRAD radar system, comprising: an antenna radome; a transmitter circuit; a receiver circuit; and means for automatically calibrating the radar system that is mounted on the antenna, where the system is calibrated by, determining a reflectivity factor for the system, determining a system bias, and determining a system differential phase.
[0006] Other aspects and advantages of the invention will be apparent from the following description and the appended claims.
BRIEF DESCRIPTION OF DRAWINGS
[0007] It should be noted that identical features in different drawings are shown with the same reference numeral.
[0008] Figure 1 shows the basic functional block diagram of the dual-polarization upgrade subsystem in accordance with one embodiment of the present invention.
[0009] Figure 2 shows circuit diagram for a horizontal test signal measurement in accordance with one embodiment of the present invention.
[0010] Figure 3 shows circuit diagram for a vertical test signal measurement in accordance with one embodiment of the present invention. [001 1] Figure 4 shows a circuit diagram for a transmitter power reading through the horizontal transmitter measurement path in accordance with one embodiment of the present invention.
[0012] Figure 5 shows a circuit diagram for a transmitter power reading through the vertical transmitter measurement path in accordance with one embodiment of the present invention.
[0013] Figure 6 shows a system calibration sequence in accordance with one embodiment of the present invention.
[0014] Figure 7 shows a performance check calibration sequence in accordance with one embodiment of the present invention.
DETAILED DESCRIPTION
[0015] A method and apparatus to calibrate the Nexrad WSR-88D weather radar system has been developed. The initial upgrade will utilize the existing WSR88D routines for calibrating reflectivity. The new methods are fairly detailed and have not been operationally implemented before. The methods are based upon the characterization of the transmitter and the receive path through the receiver and the RVP8 signal processor. This method is quite unique and leads to further refinement of how reflectivity calibration is done.
[0016] The NEXRAD Dual Polarization Upgrade is implemented such that the capability to obtain dual polarization data is introduced using as much of the existing WSR-S8D hardware as possible. The existing hardware includes the WSR8 reflector, pedestal, waveguide, cabling, transmitter, radio frequency (RF) Generator, built in test equipment (BITE) circuitry, pedestal control electronics, Radar Control Processor, and Radar Video Processor. In the WSR-88D Dual Polarization Upgrade there are three important system calibrations which must be performed: namely, the calibration of the reflectivity factor, also mapped as the determination of dBZO for reflectivity (Vaisala- Sigmet notation), determination of system Z dr bias, and determination of initial system differential phase. All three calibrations are challenging in their own way. The present invention uses this measurement of the system Ζ& bias and reflectivity calibration principles and evaluation.
[0017] The WSR-88D Radar Data Acquisition (RDA) system with dual polarization upgrade consists of a dual polarization antenna, transmitter, RF Generator with CO HO Output, duplexer (RF Pallet) for splitting the transmitter pulse, receiver front end including LNAs, Antenna Mounted Electronics (AME) unit, dual polarization BITE circuitry, Sigmet RVP8 signal processor with IF digitizer (IFD), and Sigmet RCP8 radar control processor.
[0018] Figure 1 shows the basic functional block diagram of the dual-polarization upgrade subsystem. The signal flow diagram is also shown in Figures 2-5 for test and power signals for vertical and horizontal orientations. The AME houses a receiver down converter, as well as BITE circuitry for calibration of the radar. The AME is a self- enclosed environmentally controlled system which has a heater and cooler to control its internal temperature to 25°C. This helps to maintain the gain in the down converter module as well as the test signals necessary for calibration.
[0019] The radar can be subdivided into three separate sections: the antenna/radome, transmit path, and receive path. To calibrate reflectivity and differential reflectivity all three sections should be characterized. During operation, two engineering paths are used. Test Signals can be injected into the horizontal and vertical receiver channels to calibrate the receiver. This path is shown in Figures 2 and 3 respectively. The horizontal and vertical transmitter pulses are measured using the horizontal and vertical transmitter measurement paths shown in Figures 2 and 3 respectively. A third measurement path is used, and this is the sun measurement path. The antenna is pointed towards the sun and power samples are made until the maximum powers in the horizontal and vertical channels are measured. The sun measurement signal path is in both Figures 4 and 5. Using these signal paths, both and reflectivity can be calibrated.
[0020] Starting from the radar equation, the bias in Ζ& can be expressed in terms of the various differences between the two polarization channels as: Zdr Bias (dB) = [G h (dB) - G v (dB)] + l[o* (dB) - G V A (dB)} ( 1)
Where Zdr bias is the total system bias, G h and G v are the horizontal and vertical transmit and receive path gains, and and G are the antenna gains in the horizontal and vertical polarizations respectively. Similarly, we define G h and G v
G h (dB) = G[ (dB) + Gl (dB)
G v (dB) = G v ' (dB) + G (dB)
Where G[ and G v ' are the total horizontal and vertical transmit path gains, G^ and G * are the horizontal and vertical receive path gains. Using the measurement paths in Figures 2 and 3, equation 1 is rewritten in terms of system measurements:
Zdr Bias (dB) = 2Zd m (dB) + Zdr n (dB) ~Zdr Tx Sense (dB) (2)
Where Zdr Sun is the bias in sun power measurements made at the digital receiver, RVP8, Zdr Tx is the bias in the transmitter power measurements at the RVP8 during an online calibration, Zdr Rx ' est is the bias in test signal power measurements made at the RVP8 during online calibration, Zdr B,TE is the bias in the BITE Test Signal at the horizontal and vertical front end couplers, and Zdr Tx Seme is the bias in the transmitter power measurement paths.
There are two measurements made at installation time, which are made for Z d , bias determination, namely - the bias in the test signals at the front end couplers and also the bias in the transmitter measurement path. The bias in the test signals is measured by connecting the cables in normal configuration and taking a measurement in the horizontal receiver channel using the RVP8, and then swapping the cables and taking another measurement at the RVP8. The difference in the two measurements is then the bias in the test signals.
[0022] To measure the bias in the transmitter measurement path, a transmitter pulse reading is first taken in the horizontal channel with the cables in normal configuration. The cables are then swapped and another reading is taken at the RVP8 for the vertical channel. The difference in the two measurements is the transmitter measurement path bias.
[0023] The procedures mentioned above do introduce additional biases due to differences in cable torquing. The solution to this is a live-monitoring of the test signal while the cables are being tightened down. Upon installation, the cables are first tightened down and the test signal is turned on and injected into the horizontal and vertical receiver channels. A measurement is made at the RVP8 for both channels. For all other cable connections in normal configuration, we have observed through data that they can be torqued down until the same readings are made at the RVP8, within 0.02 dB.
[0024] There are three online measurements needed for equation 2. All online measurements are differential measurements made by the RVP8. It has been shown that the rms fluctuation of a set of RVP8 differential power measurements is 0.02 dB. Zdr Sun is measured while tracking the sun over a period f up to 30 minutes (this is currently 5 minutes). The solar variance can be reduced to the order of (0.01) 2 by doing solar scan. This measurement is made during NEXRAD's Sun Check and is an off-line calibration. Zdr Tx is measured once every eight hours during a performance check in a calibration routine called "Power Sense". The third measurement Zdr^ 5 ' is measured once before every volume scan and is made at the same time as the receiver calibration. The system calibration and performance check sequences are shown in Figures 6 and 7 respectively.
[0025] A variance analysis is performed using equation 2 from above. The variance for is given by the equation: VAR[ Zdr s ] = 4 x VAR[ Zdr Sm ] + VAR[ Zclr Tx ] +
VAR[ Zdr Rx Test ] + VAR[ Zdr BITE } + VAR[ Zdr Tx Sense ]
= 8 x VAR[RVP8] + 8 x VAR[Solar] + 2 x VAR[RVP8] + 2 x VAR[RVP8] + 2 x VAR[RVP8] + 2 x VAR[RVP8]
Using RVP8 to make all differential measurements, the resulting accuracy for Ζ & calibration is 0.097 dB which is within the 0.1 dB specification.
[0026] Reflectivity may be calibrated by using the test signal path and the transmitter measurement paths as shown in Figures 2-5. Antenna Gain, radome loss, and waveguide losses are measured at installation using the NEXAD sun check and are all "lumped" into a single gain number. This gain number is referenced to an anchor point, which are the horizontal and vertical front end couplers in the system.
[0027] The transmitter power is measured at our anchor point, and so is receiver sensitivity (I 0 ). Thus, all measurements necessary for reflectivity calibration are referenced back to a single point, the front end couplers as shown in Figures 1 -5. The NEXRAD linearity routines use these measurements to calculate the reflectivity calibration factor dBZ 0 . The following provides a brief explanation f how these measurements are made.
[0028] At installation, the losses in the transmitter measurement paths are measured. The losses are measured by first using a calibrated power meter to measure the peak power at the horizontal and vertical front end couplers just before the transmitter pulse enters the antenna. Then, the same transmitter samples are measured through transmitter measurement paths in Figures 4 and 5. The difference between the power meter measurement and the RVP8 measurement is the loss in the transmitter measurement path. Using this loss, the transmitter peak power can now be measured during online operation using the transmitter measurement path and the RVP8. This measurement is currently made once every eight hours because of limitations in the software architecture. However, this could be measured very frequently , say once every VCP for a much more accurate transmitter peak power measurement.
[0029] To measure I 0 , the full power of the test signal is measured using a calibrated power meter at installation of the dual polarization upgrade. Next, using existing NEXRAD routines, the step attenuator is calibrated. During operation, the step attenuator is varied through the linear range of the receiver and a best fit curve is plotted. The noise floor is then measured. The point of intersection between the best fit curve and the noise floor is I 0 .
[0030] The calibration constant dBZO is defined as the minimum detectable reflectivity at a range of 1 km. The standard NEXRAD equation used to calculate dBZ 0 is as follows, where dBZO is defined as:
where λ is the wavelength in centimeters, G is the antenna gain, 0 is the antenna beamwidth in degrees, τ is the pulsewidth, K is the dielectric of water, P T is the transmitter peak power in kW, L t is the transmit path loss, L r is the path loss in front of the first active component of the receiver to the antenna, L d is the detection loss through the matched filter, N is the noise level measured at the RVP8, and g is the receiver gain. This equation can be simplified to: dBZ 0 (dB) = C{dB) - P Tx (dB) + / 0 (dB) Where C is the log form of all the constants, P ,x is the transmitter peak power in dBkW, and I 0 is the MDS of the receiver. The transmitter peak power and the receiver MDS were measured using calibrations described in the previous sections.
[0031] The accuracy budget for reflectivity is 0.528 dB, and with a requirement of 1 dB, there is a margin of 0.472 dB. Instead of running a linearity check after every VCP, the NEXRAD routines could be modified to adjust dBZ 0 only using the transmitter measurement path. The reason behind this is that the transmitter measurement path itself includes all of the transmitter path and the active components of the receiver. Thus, if any losses or gains along the either the transmitter path or the receiver path changes, the power of the measured transmitter pulse will change at the RVP8. Therefore, dBZ n can be adjusted according to changes in the transmitter power measurement. This measurement will increase or decrease regardless of where the gain/loss fluctuations are in the radar system. The existing NEXRAD linearity routines could be run every eight hours and determine a baseline dBZ u . From then on the transmitter measurement routine can be run to adjust this baseline. This would also eliminate the dependence on the step attenuator that is now being used in the BITE/Cal Test Signal Generator.
[0032] Another alternative is that dBZo can be measured directly using the transmitter pulse, without having to use the test signal step attenuator at all. Again, this is the result of the fact that the transmitter measurement path includes both the transmitter and receiver paths. Equation (3) can be modified for this measurement as: dBZ 0 = C(dB)- P P (dB) + 60 + L c ° 0 upler (dB)- L b (dB) + N(dB)
Where Ρ Ί νρ is the transmitter pulse measurement at the RVP8, L™' pkr is the coupling value (dB) of the front end coupler, L b is the loss t hrough the transmitter measurement path to the horizontal channel LNA, and TV is the noise floor measured at the RVP8. This method would allow for a much faster and more accurate calibration of dBZo . The step attenuator is not required for this calibration.
[0033] The calibration methods described here is based on the principles of basic calibration procedures researched over years at the various research institutions such as the CSU-CI I ILL radar, NCAR SPOL radar and the NSSL, and the NWS ROC, except this is implemented automatically using the Antenna Mounted BITE circuitry. These engineering calibrations as presented meet specifications of the z c i r bias estimation within 0.1 dB and reflectivity within 1 dB and we believe they are a good engineering approach to the calibration process. While the invention has been described with respect to a limited number of embodiments, those skilled in the art, having benefit of this disclosure, will appreciate that other embodiments can be devised which do not depart from the scope of the invention as disclosed here.