University of Nebraska UAS profiling during LAPSE-RATE

Abstract. This paper describes the data collected by the University of Nebraska-Lincoln (UNL) as part of the field deployment during the Lower Atmospheric Process Studies at Elevation — a Remotely-piloted Aircraft Team Experiment (LAPSE-RATE) flight campaign in July 2018. UNL deployed two multirotor unmanned aerial systems (UASs) at various sites in the San Luis Valley (Colorado, USA) for data collection in support of three science missions: convection-initiation, boundary layer transition, and 5 cold air drainage flow. We conducted 172 flights resulting in over 1300 minutes of cumulative flight time. Our novel design for the sensor housing onboard the UAS was employed in these flights to meet the aspiration and shielding requirements of the temperature/humidity sensors, and attempt to separate them from the mixed turbulent airflow from the propellers. Data presented in this paper include time-stamped temperature and humidity data collected from the sensors, along with the threedimensional position and velocity of the UAS. Data are quality controlled and time-synchronized using a zero-order-hold 10 interpolation without additional post processing. The full dataset is also made available for download at (https://doi.org/10. 5281/zenodo.4306086 (Islam et al. , 2020)).

position, velocity, and attitude, through a serial interface. Additionally, a mobile application allows a user to plan and deploy a flight trajectory, and the remote controller allows intervention from the user at any point.

50
Specifications of the temperature-humidity (TH) sensors recorded in the dataset are described in Table 1. Every UAS flight used one iMet XQ2 from InterMet Systems (Grand Rapids, MI, USA) as the primary TH sensor. The XQ2 is a self-contained sensor package designed for UASs to measure atmospheric pressure, temperature, and relative humidity. It is also equipped with built-in GPS, and an internal data logger along with a rechargeable battery. A serial interface provides access to the logs, or real-time observations produced by the sensor at 1Hz. The internal data-logger was only used as backup and is not part of 55 this dataset. Data included in the dataset are collected through the DAQ using the serial interface. Some UAS flights feature an older version of this sensor, called iMet XQ1, as the secondary backup sensor.
Some flights also use a nimbus-pth as the secondary sensor, which is a pressure, temperature and humidity sensor we designed and built. Several nimbus-pth can be stacked as nodes, and in some data files two of them might be present. In such cases, one of them is aspirated inside our sensor housing, and other one sits directly underneath the UAS in a traditional non-60 aspirated configuration. In the data files, first two sensors are shielded and aspirated inside the housing, and the third sensor (when available) is in traditional non-aspirated configuration.

Sensor Housing
The sensor housing is designed to meet or exceed sensor placement requirements, such as constant aspiration for the sensors, shielding from the solar radiation and other indirect heat sources. The housing draws air passively by exploiting the pressure differential between the region just above a propeller and the region just beyond the rotor wash. The airflow through the housing is always maintained as long as the propellers are spinning, and provides a consistent aspiration for the sensors. The inlet and outlet of the housing are shaped as a cone to provide high speed airflow across the housing tube with small pressure difference between the two ends. Additional design considerations are made to ensure that the flow is consistent, and provides airflow ≥ 5m/s across the sensors even at the lowest propeller speeds.

70
The housing is also designed to be modular, printed entirely using a 3D printer, and has an easy screw-in assembly. Impact of the housing on the UAS's stability and flight time is minimal. Further details and the full schematic of the housing and the evaluation can be found in our previous work (Islam et al. , 2019).

Data acquisition:
Data are collected using a data acquisition (DAQ) system made of a compact single-board computer, Odroid XU4 (Odroid ,75 2019) running a linux operating system. Odroid runs the robot operating system (ROS) (ROS , 2019) that communicates with the serial devices through the USB port of the Odroid. ROS facilitates collecting many different sensor data independently at their own output frequency; recording the timestamp for when data were generated and when they are received by ROS. ROS interfaces the collection of all available devices even in the case of a single device fail. Synchronization of the data can either be done at runtime or in the post-processing. In our case, it is done in the post-processing using MATLAB. 80 Odroid was connected with a ground computer using wireless 2.4 GHz XBee radios for operation of DAQ, debugging and periodic checks on the data. The data collected by the DAQ were retrieved using an ethernet connection.
Temperature-humidity sensors connect over serial with ROS to send periodic updates of the observations. UAS's autopilot also interface with ROS to provide updates of position, velocity, altitude, attitude etc. which are also recorded to spatially and temporally synchronize the observation.  Table 3 shows the distribution of UASs deployed by UNL by date and time and mission objectives.

110
On July 14, 2018, the mission objective was to compare both of the systems against a reference point, the MURC tower (de Boer et al. , 2020), to calibrate and validate the sensor observations. One flight for each system was conducted where the UAS ascended to the height of MURC tower (15.2m) and hovered for 10 minutes. After that, the UAS ascended to 120m at 1m/s, hovered for 30 seconds, and descended at the same speed to land. This mission was performed in collaboration with  On July 15, 2018, the mission objective of the day was convection initiation (CI). Vertical profiling flights were conducted up to 500m altitude at 1m/s ascent/descent speed in Golf and Gamma location. Flights were planned to be at every 30 minutes to allow recharge of the UAS batteries while cycling through multiple sets of batteries. At Golf, the weather was slightly cloudy 120 in the morning; clear throughout the day; very windy conditions existed for the last few flights. At Gamma, the weather was clear and windy in the morning, and slightly cloudy for the last half of the flights.
On July 16, 2018, the scheduled mission objective was also CI with flights at the same locations. Flights were limited to 120m altitude at 1.5m/s ascent/descent speed due to Notice to Airman (NOTAM) not being active for the day. Due to reduced altitude, more flights could be conducted with available batteries. As such, flights were conducted every 15 minutes. At Golf,  the weather started slightly cloudy, and then clear through out the day. At Gamma, the weather was clear throughout the day, with partly cloudy condition for the last few flights.
On July 17, 2018, the scheduled missions were for boundary layer transition (BLT). The early morning experiments, before sunrise, help validate the sensor housing reading since the measurement error from the downwash is more easily detected in stable versus well-mixed conditions. We conducted simultaneous flights for both UAS with six vertical profile and 3 horizontal 130 profile at various UAS movement speeds. The sky was cloudy throughout all the flights.
On July 18, 2018, the scheduled mission was for CI. Flights were conducted up to 500m altitude at 1.5m/s ascent/descent speed at both Golf and Gamma locations. Flights were generally conducted every 30 minutes. At the conclusion of the day, ten additional 150 m altitude flights were performed at the Golf location at various ascent/descent speed to study the effect of UAS movement speed on temperature and humidity observations. At both location, sky was clear for first half of the flights, 135 and partly cloudy for the second half.
On July 19, 2018, the mission objective was cold drainage flow. UASs were placed at the Charlie and India locations for this mission. Flights were performed starting before sunrise at maximum altitudes up to 350m at 1.5m/s ascent/descent speed.
Flights were scheduled for every 15 minutes. At both locations, the sky was cloudy before sunrise but clear afterwards. Data are recorded from individual sensors and UAS flight controller as they arrive to the DAQ as described earlier. The recorded data are then processed in MATLAB to synchronize using zero-order-hold method to create a single output file. We used a discrete sample time of 1 second for zero-order-hold to match the output rate of primary sensors. Invalid or missing data are replaced with -9999.9 wherever the sensor data are unavailable to the DAQ.
We note that the humidity sensor of the XQ2 on some flights for July 17, 2018 was saturated at 100% in one of the UAS 145 (M600P1) and are not usable; secondary sensor measurements should be used to replace these data. Also, humidity readings from nimbus-pth have sensitivity issues; although it displays a similar trend as the other sensors it does not capture the whole range of observation and will need further calibration.
No other processing was done on the data such as sensor response correction, bias correction, etc. File naming convention and explanation of the data fields can be found in the read-me file of Zenodo data repository.

Special topics of interest
The following are special topics of interest that can be studied from the dataset. Our analysis that focused on these topics can be found in our previous work (Islam et al. , 2019).

Calibration:
Data from July 14, 2018 can be used with MURC data available at Zenodo to obtain reference for calibration (de Boer et al. ,155 2020). Our previous paper (Islam et al. , 2019) discusses the deviation of our observations with MURC data over a period of 10 minutes. Other work (Barbieri et al. , 2019) compares all the different participating platforms along with ours against MURC tower data as well.
Effect of ascent/descent speed: To study the effect of ascent/descent speed on the sensor readings, 10 flights from M600P2 platform on July 18,2020 starting 160 at 20:21 UTC can be used. Flights were conducted up to 150m altitude with speeds ranging from 1-5m/s ascent speed, and 1-3m/s descent speed. Our analysis on these data can also be found in our paper (Islam et al. , 2019).

Detection of Inversion:
To study the sensor performance within an inversion layer, the first six flight from each platform can be used from July 17, 2020. The speed of flight through the inversion layer ranged from 0.5-5m/s for ascent, and 0.5-3m/s for descent. These data 165 could be used for comparison to the theoretical work of (Houston and Keeler , 2020).  Figure 3 shows examples of temperature and humidity profiles collected using the M600P1 platform's primary sensor. The top two panels illustrate a 500m profile taken through a well-mixed atmosphere. The bottom two panels in Figure 3 are an example of a profile taken before sunrise through a nocturnal inversion.