We will install a gust sensor system for measuring the three components of wind velocity relative to the aircraft. This is the heart of the UAV momentum flux measuring system and we have spent considerable time weighing the options. All available technology is based on arrays of pressure ports mounted either on a sting ahead of the nose of the aircraft or on the nose of the aircraft itself. In the former case the sting is necessary to be ahead of the free-stream flow distortion by the aircraft but comes with the penalty of additional weight (which is at a premium for small UAVs) and vibration of the sting, which requires additional accelerometers in the sting to account for it. A recent report of the use of a sting with a nine-port pressure system (5 dynamic pressure plus 4 static pressure) gave good measurements of horizontal velocities but concluded that integrated GPS/INS measurements were required for more accurate vertical velocity measurements (Spiess et al., 2007). We also consulted Aeroprobe Corporation (Blacksburg, VA), a leading maker of flow measuring probes. They confirmed our decision to use a 5-port array plus static pressure measurements, but the additional weight of their sting would take up approximately 1 lb. of otherwise usable payload, be considerably more expensive than a 5-port design integrated into the nose of the aircraft (see Figure 4), and potentially require corrections for the time lag between the pressure ports and the transducers. The last issue is avoided with flush-mounted transducers on the nose. It is worth noting that the NSF/NCAR C130 uses the 5-port design in the radome for atmospheric turbulence measurements.
The 5-port nose array plus static pressure ports along the side of the fuselage will give the three components of wind velocity relative to the UAV. The velocity field in an Earth frame is then determined by adding the motion of the aircraft from DGPS/INS measurements to the velocity field relative to the aircraft. We expect the (1 sigma) errors from the DGPS/INS velocity measurements using the C-Migits III (Table 2) to be no more than O(0.1) m/s. The C-Migits III is the state of the art in integrated DGPS/INS systems giving the same performance as those installed in the NSF/NCAR C130. (The standard DGPS-Omnistar system used for navigation in the Manta B is not sufficiently accurate for scientific motion correction.)
The differential pressure transducers to be used in the gust probe array will have an accuracy of mb = 1 Pascal at 100 Hz - 1 kHz. For incident velocities in the range 20-40 m/s, the airspeed range of the aircraft, this will permit resolution of the velocity relative to the aircraft with errors of O(0.01 - 0.001)U. This will permit resolution of pitch and yaw with errors of better than O(0.01 - 0.001) radians, or vertical and cross-track velocities with errors of O(0.01 - 0.001)U. Since is O(), and accounting for the DGPS/INS errors in velocity, this suggests that instantaneous Reynolds stress errors will be in the range of 5-10%. Based on the available literature on 5-port probes we expect that angles of pitch and yaw in the range degrees can be measured (Aeroprobe Corp.).
Figure 3: (a) UCSD wind tunnel setup for testing 5-port turbulence probe. (b) Schematic of the 5-port turbulence probe. (c) Directional response of the probe measured in the UCSD wind tunnel. (d) Set up of sonic anemometer and turbulence probe on the pickup truck. (e) Intercomparison of time series of horizontal and vertical velocities measured by both instruments.
The turbulence probe is the heart of the covariance flux measurement system for all variables since the vertical velocity measurement is common to all. Figure 3 shows the configuration of the hemispherical 5-port probe and the initial results of wind tunnel and field testing. The wind-tunnel testing shows the directional response of the probe, which when accompanied by the dynamic pressure measurement at the center port, gives the three components of velocity relative to the aircraft. The initial field testing has been accomplished by mounting the probe just behind the measuring volume of a CSAT3 sonic anemometer mounted above the cab of a pickup truck. The pickup truck was then driven at speeds up to 37 m/s along the road and data from both instruments recorded simultaneously. The figure shows direct intercomparisons of time series of streamwise (horizontal) velocities, , and vertical velocities, , along with corresponding spectra, and direct comparisons of Reynolds stress components and . The agreement is very good up to frequencies of 15-20 Hz, and we expect to be able to improve it even more with small improvements in the probe design. For the Aerosonde aircraft we plan to implement a 9-port turbulence probe which will provide improved directional performance and other benefits described below.
At low altitudes of less than 100 m, the height of the aircraft above the surface is measured by the laser altimeter, which, when combined with the aircraft GPS/IMU data, also measures the surface wave profile along the flight track of the aircraft with corrections for the orientation of the aircraft.