The Multi-Purpose Airborne Sensor Carrier MASC-3 for Wind and Turbulence Measurements in the Atmospheric Boundary Layer

Abstract

For atmospheric boundary-layer (ABL) studies, unmanned aircraft systems (UAS) can provide new information in addition to traditional in-situ measurements, or by ground- or satellite-based remote sensing techniques. The ability of fixed-wing UAS to transect the ABL in short time supplement ground-based measurements and the ability to extent the data horizontally and vertically allows manifold investigations. Thus, the measurements can provide many new possibilities for investigating the ABL. This study presents the new mark of the Multi-Purpose Airborne Sensor Carrier (MASC-3) for wind and turbulence measurements and describes the subsystems designed to improve the wind measurement, to gain endurance and to allow operations under an enlarged range of environmental conditions. The airframe, the capabilities of the autopilot Pixhawk 2.1, the sensor system and the data acquisition software, as well as the post-processing software, provide the basis for flight experiments and are described in detail. Two flights in a stable boundary-layer and a close comparison to a measurement tower and a Sodar system depict the accuracy of the wind speed and direction measurements, as well as the turbulence measurements. Mean values, variances, covariance, turbulent kinetic energy and the integral length scale agree well with measurements from a meteorological measurement tower. MASC-3 performs valuable measurements of stable boundary layers with high temporal resolution and supplements the measurements of meteorological towers and sodar systems.

Keywords: 3D wind vector measurement; comparison with measurement tower; fixed-wing unmanned aircraft; remotely piloted aircraft (RPA); stable boundary layer; turbulence measurement; unmanned aircraft system (UAS).

Figure 1

Multi-Purpose Airborne Sensor Carrier (MASC-3) sketch (top) and pictures of the airframe with the sensor system (middle) and five-hole probe (bottom).

Figure 2

MASC-3 altitude profile during automatic landing procedure.

Figure 3

MASC-3 flight paths during the Hailuoto-II campaign. The island is indicated in white color and the grey area indicates water, which was completely frozen during the measurements allowing the installation of the indicated measurement tower. The sodar was installed on the island. The flight paths are plotted from the longitude and latitude readings of the inertial navigation system. The flight path section (leg) of Flight #10 and Flight #11 was used for the comparison between the tower, the Sodar and MASC-3. During Flight #10 and Flight #11 the mean wind direction at 100 m above ground level (AGL) is indicated. Map tiles by Stamen Design (http://stamen.com/) under CC BY 3.0 (http://creativecommons.org/licenses/by/3.0). Data by Open Street Map (http://openstreetmap.org/) under ODbL (http://www.openstreetmap.org/copyright).

Figure 4

Data and power flow diagram of the MASC-3 sensor system.

Figure 5

Sensor system hat (left) and mounted electronics inside the sensor hat (right).

Figure 6

Schematic software setup of the sensor system on-board MASC-3.

Figure 7

Meteorological measurement tower during the Hailuoto-II campaign. Viewing direction is north-north-east towards the harbour and the village Marjaniemi. The picture was taken by Kristine Flacké Haualand.

Figure 8

Results of the ogive test between vertical w and the horizontal $u_{rot}$ wind components using all observations from the three tower heights during the period 10 February 2018 14:30–22:00 UTC. Ogives are normalized by the value at the point closest to the frequency corresponding to 10 min, indicated by the first vertical dashed line from left. The second vertical dashed line represents the frequency corresponding to 60 s.

Figure 9

Wavenumber spectra (left) and structure functions (right) for the horizontal wind $v_h$ (top) and the vertical wind vector component w (bottom). The data of the tower at the 10.3 m level inherits a time series of ${\mathsf{\Delta }t}_{\mathrm{tower}}=165$ s, corresponding to the fetch of the MASC-3 flight leg with a duration of ${\mathsf{\Delta }t}_{\mathrm{UAS}}=77$ s. Flight #10 and the first leg at 11.7 m AGL is given.

Figure 10

Time series of the tower with the corresponding leg averages and standard deviation of MASC-3 at the lowest flight levels for the horizontal wind $v_h$ (top) and the wind direction $\varphi$ (bottom). The mean altitude and standard deviation of the individual flight leg is given for the MASC-3 data points. The data of the tower at 10.3 m is plotted as rolling (moving) average with standard deviation and a window length of ${\mathsf{\Delta }t}_{\mathrm{tower}}=170$ s corresponding to the fetch of the MASC-3 flight legs with an average duration of ${\mathsf{\Delta }t}_{\mathrm{UAS}}=80$ s. Furthermore the neighboring ten minute averages of the tower at all height levels are given.

Figure 11

Time series of the tower with the corresponding leg averages of MASC-3 at the lowest flight levels for the variance of the horizontal wind ${\mathrm{Var}}_{v_h}$ (top) and the vertical wind component ${\mathrm{Var}}_w$ (bottom). The mean altitude and standard deviation of the individual flight leg is given for the MASC-3 data points. The data of the tower at 10.3 m is calculated on a moving window with a width of ${\mathsf{\Delta }t}_{\mathrm{tower}}=170$ s corresponding to the fetch of the MASC-3 flight legs with an average duration of ${\mathsf{\Delta }t}_{\mathrm{UAS}}=80$ s. Furthermore the neighboring ten minute averages of the tower at all height levels are given. The ${\mathrm{Var}}_w$ for the first ten minute interval of the tower at 10.3 m and 4.5 m lie on top of each other.

Figure 12

Time series of the tower with the corresponding leg averages of MASC-3 at the lowest flight levels for the turbulent kinetic energy (TKE) (top) and the covariance of the vertical and horizontal wind component ${\mathrm{Cov}}_{w{u}_{rot}}$ (bottom). By 2D double rotation for the tower and by coordinate transformation with the mean wind direction of the individual MASC-3 flight legs, $u_{rot}$ was aligned with the mean wind direction. The mean altitude and standard deviation of the individual flight leg is given for the MASC-3 data points. The data of the tower at 10.3 m is calculated on a moving window with the length of ${\mathsf{\Delta }t}_{\mathrm{tower}}=170$ s corresponding to the fetch of the MASC-3 flight legs with an average duration of ${\mathsf{\Delta }t}_{\mathrm{UAS}}=80$ s. Furthermore the neighboring ten minute averages of the tower at all height levels are given.

Figure 13

Integral length scales of the horizontal wind $L\left(v_h\right)$ (top) and the the vertical wind component $L\left(w\right)$ (bottom). For the tower at 10.3 m, several fractions of the time series with a duration of ${\mathsf{\Delta }t}_{\mathrm{tower}}=170$ s, corresponding to the fetch of the MASC-3 flight legs with an average duration of ${\mathsf{\Delta }t}_{\mathrm{UAS}}=80$ s, were used to plot the length scales alongside the values for the individual MASC-3 flight legs. The mean altitude and standard deviation of the individual flight legs are indicated. Furthermore, the integral length scales of the 10 min time series of the tower at all height levels are given.

Figure 14

MASC-3 Flight #11 alongside the corresponding tower data and Sodar data as height profile for the potential temperature $\theta$ (left), the horizontal wind speed $v_h$ (middle) and the wind direction $\varphi$ (right). The time series of the tower data points have a duration of ${\mathsf{\Delta }t}_{\mathrm{tower}}=170$ s, corresponding to the fetch of the MASC-3 flight legs at the lowest levels with an average duration of ${\mathsf{\Delta }t}_{\mathrm{UAS}}=55$ s. The timestamps of the first measurement points of each profile and the timestamps of the Sodar profiles are given.

Figure 15

MASC-3 Flight #11 alongside the corresponding tower data as height profile for the variance of the horizontal wind speed ${\mathrm{Var}}_{v_h}$ (left) and the variance of the vertical wind speed ${\mathrm{Var}}_w$ (right). The time series of the tower data points have a duration of ${\mathsf{\Delta }t}_{\mathrm{tower}}=170$ s, corresponding to the fetch of the MASC-3 flight legs at the lowest levels with an average duration of ${\mathsf{\Delta }t}_{\mathrm{UAS}}=55$ s. The ${\mathrm{Var}}_w$ profile (right) inherits the scaling function.

Figure 16

MASC-3 Flight #11 alongside the corresponding tower data as height profile for the TKE (left) and the covariance ${\mathrm{Cov}}_{w{u}_{rot}}$ (right) of the vertical wind w and the transformed vector component $u_h$ which is aligned with the mean wind direction. The time series of the tower data points have a duration of ${\mathsf{\Delta }t}_{\mathrm{tower}}=170$ s, corresponding to the fetch of the MASC-3 flight legs at the lowest levels with an average duration of ${\mathsf{\Delta }t}_{\mathrm{UAS}}=55$ s. The ${\mathrm{Cov}}_{w{u}_{rot}}$ profile (right) inherits the scaling function.