1. Introduction
Field experiments focusing on meteorological phenomena commonly have as their objective the three- dimensional sampling of a volume of the atmosphere as the system under investigation evolves. To achieve this objective, measurements from a number of sites and platforms at a variety of scales are utilized. Typically, remote observations from space, such as those available from weather satellites, are used to provide information on the synoptic-scale environment in which the system develops. In situ measurement from an aircraft platform can provide, for a limited period of time, more detailed information in a targeted area of interest. In addition, ground-based instruments can provide high-resolution time series at a limited number of sites. Though satellite images are often archived during such field experiments for later analysis, the availability of real-time imagery opens the door to redirecting an aircraft measurement platform to sample an area of interest in a timely and spatially meaningful fashion. Imagery can be provided directly to research personnel on an aircraft either through direct reception or through the use of a ground- based receiving facility and a data link between the ground and the aircraft.
The NRC Convair 580 research aircraft has been instrumented to carry out in situ measurements (Jordan et al. 1993) of meteorological phenomena and has been used in organized field experiments in collaboration with agencies such as the Atmospheric Environment Service (AES). The challenge in carrying out such measurements has often been compounded by operation in remote areas such as in the Arctic or over the ocean. Satellite imagery and weather information have normally been available from a flight planning or field experiment operations center, where go–no-go decisions are made for experiment flights. However, this information may be many hours old by the time the aircraft arrives “on station” in the experiment area. In remote areas, the possibility of communicating with the aircraft after it leaves the immediate area of the airport has been very limited or nonexistent. As well as meeting the research objectives the ability to provide updated weather information in flight was considered important for operational safety. These factors led to the consideration of systems using satellite data and images that could provide updated wide area weather information in remote areas to augment the aircraft data gathering systems.
Various options were considered in the development of such a capability. When the project was initiated, direct reception of satellite imagery was considered the best option since viable data link communications technologies to relay the imagery from a ground station were either too limited in range, too low in bandwidth, or too expensive. In addition, this option allows the aircraft to operate in a stand-alone mode independent of a ground- based receiving station. A number of weather satellite systems provide direct transmission of images including geostationary [NOAA GOES series (U.S. National Oceanic and Atmospheric Administration Geostationary Operational Environment Satellite)] and polar-orbiting [NOAA POES series (Polar Operational Environment Satellite) and Russian Meteor series] satellites. The various transmission formats available are summarized in Table 1. The most desirable high-resolution digital transmission formats use the microwave S band at approximately 1700 MHz but require a relatively large (1-m diameter or larger parabolic dish) antenna, which for a moving platform requires the use of a sophisticated tracking mechanism. The design of such an antenna was considered very difficult and costly for an aircraft installation. The alternative low-resolution Automatic Picture Transmission (APT) format from polar-orbiting satellites utilizes the 137-MHz VHF band and requires the use of a much simpler, low-gain, fixed omnidirectional antenna, providing full hemispheric coverage without tracking. In the case of the NOAA polar-orbiting weather satellites, the images are generated in the onboard processor by subsampling the data stream of the Advanced Very High Resolution Radiometer (AVHRR) instrument, providing a resolution of approximately 4 km pixel−1 (at nadir) for the analog format APT transmission of 120 lines min−1. In comparison, the High Resolution Picture Transmission (HRPT) digital format used by many earth receiving stations provides the full 1.1 km pixel−1 (at nadir), 10 bit pixel−1 resolution of the AVHRR instrument. A visible and an infrared (IR) channel are selected by the onboard processor from the five available AVHRR channels (see Table 2) for the APT transmissions. For the visible, channel 2 is normally provided and for the infrared, channel 4 is selected in the daytime and channel 3 in the nighttime. The thermal IR channel provides the capability of deriving scene temperatures with proper calibration, although the use of APT precludes derivation of sea surface temperatures using the more accurate multichannel SST algorithms (e.g., McClain et al. 1985) used with the full AVHRR datasets. The choice of direct reception of APT imagery seemed consistent with the desire to equip the aircraft with a “quick look” imaging system at a relatively low cost to provide the latest remote synoptic and mesoscale observations of weather systems, particularly at higher latitudes where polar satellite passes are frequent.
2. APT system implementation
a. Electronics and computer system
The APT system installed on the aircraft is illustrated in the block diagram in Fig. 1. This system was derived from a commercial ground-based APT system using an industrial-grade rack-mounted PC-compatible computer system and an uninterruptible power supply (UPS). A 486-66MHz processor board is mounted in the ISA- compatible backplane along with commercial APT receiver and decoder boards (Quorum PC-137 and Wefax- II). The workstation user interface consists of a rack- mounted 17-in. multisynch monitor operating at 1024 × 786 8-bit pixels (providing 256 colors), a keyboard, and a mouse. The computer system also houses additional processor boards and interfaces for a GPS navigation receiver and other instruments, which share the operator display, keyboard, and mouse using a special multiplexer switch. The UPS supply was used to augment the aircraft power to avoid electrical transients as well as outages on switchover to ground power after flights, which was a requirement for the GPS navigation system for a related differential GPS experiment.
The front-end RF electronics are similar to those used in ground-based installations. A preamplifier is located near the antenna to boost the signal level from the antenna. This signal is fed through a low-loss, double-foil shielded cable (Belden 9311) to the receiver card located in the computer. A bandpass helical filter was used to attenuate out-of-band interference and was found to be useful in rejecting the relatively high noise levels found in the aircraft environment from other avionics, instrumentation, and computing equipment. Specifications for the system components are provided in Table 3.
b. Antenna development
The antenna is the most critical element in the design of an airborne APT system. The ideal design would operate omnidirectionally providing some gain over the full upper hemisphere for polar satellite passes that may track through any point above the horizon. The antenna should also provide circular polarization to receive signals in an optimal fashion since most weather satellites transmit with this type of polarization to reduce multipath effects. Antenna size is normally proportional to the operating wavelength that at 137.5 MHz is 2.18 m. Ground-based antennas in common use include the crossed dipole, quadrifilar helix, and volute designs, which are large and unwieldly and hence unsuitable for aircraft use. For aircraft use, an antenna should be as small as possible to minimize aerodynamic drag and complications due to structural, mechanical, and installation issues as well as related problems such as aircraft icing. No aircraft antennas specifically designed for APT reception were found to be commercially available. Also, the authors were not able to find reports of any existing designs in the literature, though we are aware of an installation on a NASA Ames aircraft using a pedestal antenna developed by the U.S. Air Force in the 1950s. This situation led us to study various alternative antenna designs and eventually to proceed with the design of an airborne APT antenna to meet the project requirements.
Initial studies were directed at considering the various alternatives to commonly used ground-based antennas. The first initiative involved looking at commercially available avionics and aircraft antennas. Most designs used in the VHF airband are optimized for air-to-ground navigation and communications using compact vertical stub–blade designs, which theoretically have a freespace gain independent of azimuth but proportional to the cosine of elevation angle. Ground tests of such an antenna using an aluminum sheet ground plane indicated that the expected nulling effect at high elevation angles for near overhead satellite passes was not a problem. The vertical antenna provided marginally adequate gain but was subject to signal fading, presumably because of its vertical polarization instead of the ideal circular polarization. In the fall of 1994, a prototype APT system using a vertical avionics antenna (Jordan et al. 1995) was installed on the NRC Convair 580 aircraft. A number of problems were experienced, including marginal gain, fading leading to loss of decoder lock, shading from an overhead HF longwire antenna, and interference from aircraft voice transmissions on the adjacent frequency band. As described in the next section, this system was used in the Beaufort and Arctic Storms Experiment (BASE), and good imagery was received on a number of passes, which proved useful to the experiment and helped to confirm the value of such a system. The vertical avionics antenna, though never intended for APT reception, is nonetheless commercially available, certified for aircraft use, and low in cost.
Subsequent studies of the antenna literature indicated that microstrip and cavity-backed slot antennas had been used for a number of space applications requiring circular polarization and omnidirectional gain characteristics. In particular, the microstrip antenna design seemed to be ideal for an airborne platform because of its low profile and the possibility of using it in a conformal design to match the curvature of the aircraft fuselage. A number of prototypes were developed for this application. In ground tests, these prototypes were found to perform very well, exceeding the performance of a reference helix antenna except at very low elevation angles. The main difficulty with these designs was their size (e.g., 100-cm square) and concerns with conformal mounting on an aircraft fuselage subject to expansion and contraction with pressurization. This led to the development of a novel “electrically small” patch antenna (40-cm square) using state-of-the-art microstrip technology.
The ground test results using the patch antenna, the avionics vertical antenna, and a reference helix antenna manufactured for use in ground receiving systems are shown in Fig. 2 and are indicative of the performance of these antennas. Note that all antennas were tested on a 1.2-m square sheet of aluminum acting as a ground plane to simulate the fuselage surface of the aircraft. The plots show signal strength versus elevation angle for typical near overhead satellite passes as read from the display of the weather satellite receiving system. The vertical avionics antenna provides barely adequate gain with no obvious nulling at zenith, while the effects of fading can be seen from the variations in signal strength of the plot. The patch antenna provides higher gain than the reference helix antenna at elevation angles over 25° but falls off rapidly below this angle. The data for the patch antenna were taken from a pass with a 60° maximum elevation angle, which was the best available during the limited time period for ground testing. Since the slant range for such a pass is greater than for a directly overhead pass, the signal strength plot shown is on the conservative side. Observations on the aircraft confirm that the antenna has a similar gain characteristic at higher elevation angles. The reference helix antenna is well known for its excellent gain characteristics and its ability to operate at low elevation angles, though this was not a critical requirement for airborne use.
The commercial vertical avionics antenna (Comant CI-119), illustrated in Fig. 3, is quite compact and aerodynamically streamlined. It or similar designs are commonly used on most aircraft. The patch antenna installed on the Convair is shown in Fig. 4 beside a 17-in. monitor to illustrate its size. This antenna has a base with a curvature to match the cylindrical shape of the fuselage for direct mounting. The antenna element is covered with a composite radome that is effectively transparent at VHF.
In summary, the patch antenna reported here provides a good compromise between performance and size. It provides circular polarization and excellent gain above 25° in elevation angle in comparison to a reference helix antenna, allowing the capture of imagery in the vicinity of the aircraft for passes offset in longitude by as much as half the scan width of the satellite scanner. At the same time, the antenna has a low profile to minimize aerodynamic drag and is small enough to permit easy mounting without major mechanical modifications to the airframe. The current version of the antenna is a research prototype, but it could be commercialized and certified for general aircraft use.
c. Software
The software used in the APT system was supplied by the manufacturer of the APT boards used in the airborne system and is DOS compatible. This software provides a number of modes of operation including Capture, Predict, and View, which need to be used one at a time, when used with a nonmultitasking operating system such as DOS. In the Capture mode, the APT receiver and decoder boards are configured and controlled by the software providing image acquisition once a signal is received and the decoder is locked. The Predict mode uses a software-based satellite prediction model to derive azimuth and elevation angles between the observer and the satellite and also the start and end times of satellite passes. On the Convair, this mode is used to start–stop image capture. Between passes, the View mode can be used to interactively enhance and display images. This software mode provides a number of capabilities including pan–zoom, interactive contrast–brightness adjustment, filtering, histogram, equalization, and enhancement tools. Also, the software provides geographic overlays, color contouring of infrared temperatures, and a readout of the position and temperature of a pixel selected interactively by the mouse. The geographic overlay is indispensable when land features are obscured by cloud. The distance–bearing and position readouts were found useful in flight to locate features on the image relative to the aircraft position.
The software worked well, though originally intended for operation from a fixed ground location. In initial experiments, the aircraft position was entered manually from the adjacent GPS receiver readout. For proper location of the geographic map overlays, accurate time of day is required as well as reasonably up-to-date position and ephemeris data. A software update supplied by the vendor now allows time and position information to be accepted by a serial port on the host PC workstation using the NMEA 0183 format data output provided by many GPS positioning systems. This overcomes the problem of manually entering this information, though ephemeris data must still be updated every week or so for accurate geographic overlays.
3. Results and discussion
The two antennas described in section 2 have been used in a total of four field experiments. The vertical avionics antenna was initially mounted on the aircraft and used during the BASE experiment (AES 1994) during September and October 1994. Subsequently, the patch antenna was mounted on the aircraft in February 1995 and used in the Canadian Freezing Drizzle Experiment (CFDE) (Cober et al. 1996) off the coast of Newfoundland, Canada, in February and March 1995; in a Synthetic Aperture Radar Experiment (UKSAR) in the United Kingdom in May 1995; and in the Radiation Aerosol and Cloud Experiment (RACE) in Nova Scotia, Canada, in August and September 1995. Images obtained during these experiments are provided here to illustrate the capabilities of the weather satellite imaging system.
The first image is shown in Fig. 5 from the IR channel of NOAA-9 from the BASE experiment on 29 September 1994. This shows a series of mesoscale vortices that moved into the BASE experiment area that were the subject of a series of aircraft measurement flights. The BASE experiment involved the investigation of storm evolution in the Beaufort Sea area during the late autumn when the area experiences a transition to winter conditions as the arctic airmass advances southward.
The second image is shown in Fig. 6 from the visible channel of NOAA-14 on 6 March 1995 during CFDE. This shows the development of convective cloud structure over the Labrador Sea in the vicinity of Newfoundland from cold arctic air flowing over the warm sea as a result of circulation around a polar low located to the east.
The third image is shown in Fig. 7 from the visible channel of NOAA-14 on 4 May 1995. This shows clear weather over much of Europe and the United Kingdom with many of the land features visible. This image was acquired in flight during a nonmeteorological experiment involving trials of an SAR radar system.
The fourth image is shown in Fig. 8 from the visible channel of NOAA-14 on 10 September 1995 during RACE in Nova Scotia on the east coast of Canada. This image shows the passage of a frontal system associated with a hurricane system that passed up the eastern seaboard of North America during the experiment period.
The system proved to be useful on a number of occasions, mainly during the BASE experiment because of the nature of the experiment and the frequent satellite passes at high latitudes. Near the beginning of the BASE experiment, a flight was undertaken on 4 September 1995 to investigate a weak mesoscale vortex that had developed along the coast of the Beaufort Sea in the MacKenzie River delta area. Imagery obtained in flight showed the development of a dry slot associated with the vortex, which was an unexpected development, but that confirmed other observations made in flight. On 24 September a mission was flown to document water vapor transport into the MacKenzie River Basin associated with an intense cyclonic system that was propagating into the BASE region from the Gulf of Alaska. Images obtained in flight showed that the secondary redevelopment of the system as it traversed the mountain ranges occurred faster than predicted by numerical modeling. On 29 September a low-level flight was undertaken to measure surface fluxes associated with the ice–sea interface offshore, when the imagery in Fig. 5 was obtained showing the remarkable development of a train of vortices over the Beaufort Sea area. This exceptional image allowed onboard mission scientists to plan a second mission later that day to document the baroclinic zone on which these vortices were developing. Without this image, the second mission would not have been scheduled and a very important opportunity to collect information on the structure of this baroclinic zone would have been missed. On 1 October another flight was scheduled to investigate the structure of one of these vortices observed 29 September. Imagery obtained in flight provided an update regarding the position of the mesoscale low so that measurements could be targeted to characterize the circulation near the center. On 6 October a mission was flown to document the structure of a mesoscale band of precipitation moving along the coast of the Beaufort Sea. The leading edge of the band was located using the images obtained in flight. A ground-based HRPT system was also available to forecasters and researchers during the BASE experiment. More details of the BASE experiment are described elsewhere (AES 1994).
Difficulties experienced with the APT system have mainly involved electromagnetic interference from other systems and instruments on the Convair 580. Voice transmissions from the pilot’s navigation–communication radio systems interfere with satellite reception since several watts of power are radiated only a few meters away from the APT antenna on an adjacent frequency. However, with some coordination, it is often possible to delay such transmissions until after the pass of a weather satellite. Interference from a GPS receiver was found to be a problem with the initial installation of the vertical avionics antenna but has not been a problem with the patch antenna. Signal shading and fading experienced by the vertical avionics antenna have been alleviated with the use of the patch antenna. In most situations, the patch antenna outperforms the vertical antenna since the vertical antenna provides marginal gain and is only linearly polarized. Note that the vertical antenna occasionally performs quite well for short periods of time at low elevation angles with the Meteor series of satellites, which use linear polarization. In this variable polarization situation, propagation may be optimal from time to time depending upon the orbital geometry and multipath conditions.
Experience with the airborne APT system indicates that the patch antenna can provide consistently good imagery when the elevation angle of the satellite is above 25°, allowing imaging in the vicinity of the aircraft when the satellite pass is offset in longitude by as much as half its scan width.
4. Concluding remarks
An APT weather satellite imaging system has been successfully installed on an aircraft and used in a number of meteorological field experiments with promising results. The most critical element of an airborne APT system is the antenna. The development of a low-profile“electrically small” patch antenna is reported here, which provided a good compromise between antenna performance and size for the airborne environment. Sample images from four field experiments were presented to demonstrate the capabilities of the system.
Acknowledgments
The authors acknowledge the work carried out by InfoMagnetics Technologies Corporation under the leadership of Dr. M. Barakat, in consultation with Dr. L. Shafai of the University of Manitoba Antenna Engineering Laboratory in the development of the patch antenna. The authors also acknowledge funding provided by Environment Canada, the National Research Council of Canada, the Office of Naval Research, and the Panel on Energy Research and Development.
REFERENCES
AES, cited 1994: The Beaufort and Arctic Storms Experiment (BASE) homepage. [Available on-line from http://www.tor.ec.gc.ca/BASE/base_homepage.html.].
Cober, S. G., G. A. Isaac, and J. W. Strapp, cited 1996: Characterization of the aircraft icing hazard associated with supercooled drizzle. Program and Abstracts of Canadian Meteorology and Oceanography Society Congress, Toronto, May 1996. [Available online from http://www.atmosp.physics.utoronto.ca/CMOS96/ abstracts/AvnMet/stewart_thu_08_45.html.].
Jordan, J. E., D. L. Marcotte, C. D. Hardwick, and J. W. Strapp, 1993:Instrumentation and computing facilities for the Canadian Atlantic Storms Program (CASP II). Preprints, Eighth Symp. on Meteorological Observations and Instrumentation, Anaheim, CA, Amer. Meteor. Soc., 181–185.
——, ——, and G. W. K. Moore, 1995: A prototype airborne APT weather satellite imaging system. Preprints, Ninth Symp. on Meteorological Observations and Instrumentation, Charlotte, NC, Amer. Meteor. Soc., 231–235.
McLain, E. P., W. G. Pichel, and C. C. Walton, 1985: Comparative performance of AVHRR-based multichannel sea-surface temperatures. J. Geophys. Res.,90, 11587–11601.
APT system block diagram.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0080:AAAWSI>2.0.CO;2
Intercomparison of antennas for near overhead NOAA passes: aircraft APT patch antenna, commercial avionics vertical antenna, and reference ground helix antenna (volute). Note that the aircraft patch antenna has more gain than the reference antenna above 20°–25° elevation angle. (Signal strength measurements are made with the aircraft APT receiver on a ground range using a 120-cm square ground plane underneath the antennas.)
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0080:AAAWSI>2.0.CO;2
Commercial avionics vertical antenna engineering drawing.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0080:AAAWSI>2.0.CO;2
NRC APT patch antenna beside 17-in. monitor.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0080:AAAWSI>2.0.CO;2
NOAA-9 IR (Channel 4) 29 September 1994 Beaufort and Arctic Storms Experiment (BASE): Vertical antenna.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0080:AAAWSI>2.0.CO;2
NOAA-14 VIZ (Channel 2) 6 March 1995: Canadian Freezing Drizzle Experiment (CDFE): Patch antenna.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0080:AAAWSI>2.0.CO;2
NOAA-14 VIZ (Channel 2) 4 May 1995: United Kingdom SAR Experiment (UKSAR): Patch antenna.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0080:AAAWSI>2.0.CO;2
NOAA-14: VIZ (Channel 2) 10 September 1995: Radiation Aerosol and Cloud Experiment (RACE): Patch antenna.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0080:AAAWSI>2.0.CO;2
Weather satellite transmissions.
AVHRR channels.
Specifications of electronic components.
