Usefulness of UAV-Mounted Multi-Sensors System for In Situ Atmospheric Measurement: A Case Study from Wrocław, Poland ^†

Abstract

Air pollution, especially particulate matter (PM_x), is one of the most serious environmental threats worldwide. It is challenging in terms of both public health, impact on climate, and the reduction in visibility. The assessment of spatial variability of PM_x allows us to better understand the processes that cause smog episodes, and may also be an additional element for the validation of the results of dispersion models. This study presents the results of measurements of basic meteorological parameters and air pollution involving a multi-sensor system. A Matrice 600 hexacopter with an installed environmental head was used as the measurement platform. This system enables us to measure the concentrations of PM_2.5, PM₁₀, air temperature and humidity.

1. Introduction

Air pollution, especially particulate matter (PM), is considered as a major environmental threat worldwide [1,2]. Associations between ambient air pollution and adverse health effects are well documented, both for long-term [3,4] and short-term health impacts [5,6]. Particulate matter has a serious detrimental effect on the environment, damages the crop, causes climate change and reduces visibility [7,8,9]. High-resolution monitoring of air pollution concentration and meteorological conditions (temperature, humidity, wind, etc.) within the atmospheric boundary layer (ABL) [10] is crucial for various environmental applications [11]. It is important to understand the processes of surface–atmosphere interactions, which govern, e.g., high air pollution concertation events, and it is vital to measure the impacts of air pollution on human health [12] and the environment. Information about these parameters is usually limited to a few meters above the ground; moreover, these measurements are most frequently carried out in stationary mode [13], and sometimes are extended with remote sensing technology [14,15]. Thus, mobile measurement involving UAVs is an interesting supplement for ground-based measurement [16,17].

The development of unmanned aerial vehicles (UAVs) provides possibilities for atmospheric measurements within the ABL and seems to be useful for surveys in small areas [18]. Drones equipped with properly designed and constructed sensors are successfully used in measuring air pollution, greenhouse gases, and meteorological variables [19,20].

The main objective of the study is to present the ability of using UAVs in simultaneous research of air pollution concentration and meteorological parameters and to determine the variability of the PM concentration together with the ABL structure in the lower part of the atmosphere in various types of land use and with different emission structures.

2. Data and Methods

2.1. Design of the Hexacopter-Based Measuring System

The measuring head installed on the UAV is the original solution, configured to measure selected air pollutants (PM, O₃) and basic meteorological parameters (T, RH, P) during flight. In addition, the device records flight parameters such as geographic coordinates and altitude. The inlets of the sensors used in the measuring heads were placed above the plane of the rotors to minimize the influence of turbulence caused by the propellers. Low-cost PM (optical) and O₃ (electrochemical) sensors were installed in the head, the selection of which was preceded by research and tests. All data were validated against the higher-quality data from Meteorological Observatory during the test stage. Measurement data were saved in the logger’s memory placed on the UAV and continuously monitored on the monitor screen using radio transmission.

2.2. Field Companies

The measurements using the drone were carried out in Wrocław, SW Poland. Wrocław, as do many cities in Poland, suffers from poor air quality, especially during the winter period. This is the result of the widespread use of coal and wood for domestic heating. Due to this reason, Polish cities are placed among the regions with the worst air quality in the European Union [21].

In order to assess the spatial variability of the particulate matter concentration and meteorological conditions, measurement campaigns were performed in 3 locations representing urban, suburban (an area of detached houses and allotments) and rural locations. Measurements were carried out in horizontal and vertical modes. In the years 2019–2020, nine such sessions were conducted, and during each of them several flights were performed (Table S1 in Supplementary material). The effective monitoring time during each flight was about 10–20 min. Background data including one-minute PM_2.5 and PM₁₀ concentration, air temperature, humidity, wind speed and wind direction supplemented by sodar data were obtained from the Meteorological Observatory of the University of Wrocław.

2.2.1. Horizontal Measurements

Horizontal measurements were carried out automatically in a selected area and at a certain height above the roof layer (Figure 1). The data were then presented in the form of spatial distributions of the analysed parameter. This enabled the identification of the main sources of air pollution, as well as the assessment of the variability of air temperature in a given area.

2.2.2. Vertical Measurements

Vertical transects over different types of land-use within the lowest 350 m were carried out several times during each survey in order to obtain temperature, humidity, and PM profiles. This approach allowed us to assess the vertical distribution of these variables and compare the temperature profile with acoustic sodar data (Figure 2). Due to the distortion of air caused by the rotors, only ascending flights were used for further analysis.

3. Examples Results

To assess the variability of the structure of the atmospheric boundary layer and distribution of particulate matter, we investigated in detail three selected measurement sessions. The first one was characterized by stable conditions during the night with a well-developed ground-based temperature inversion. The second one, carried out in November 2020, was characterized by a multi-layered night boundary layer with quite high dynamics. The last one, from December 2020, was conducted during the morning transition period.

The development of a stable, nocturnal boundary layer during calm wind conditions provides a gradual increase in air pollution concentration from combustion sources within the surface layer (Figure 3). After sunset, the inversion layer develops because of the steadily dropping temperature caused by the radiative cooling of the ground. Such conditions also favour very strong gradients in the pollutant profile, and maximum concentrations occur below the inversion layer, falling to almost 0 µgm⁻³ in the zone above the ground-based layer. As indicated by the measurements carried out in the city centre (GS profiles), the increased roughness and the modification of radiation properties of the surface contribute to the reduction in the temperature gradients within the inversion layers to about 70 m a.g.l. and the occurrence of an isothermal layer in the lower part of the ABL.

During the next field study, the concentration of PM was lower because of the less stable conditions within the night ABL. As indicated by the sodar data, the inversion was characterized by a wavy structure caused by stronger mixing processes. The course of PM concentration in the vertical profile above the two selected areas was very similar, except for the last flight. The advection of the plume from the emission sources (also visible in horizontal measurements) increased the concentration by about 20 μg m⁻³ (Figure 4).

During the morning hours, the dynamics of PM concentrations vary from those of the evening or night conditions. First, the breaking of the night ABL leads to the formation of strong descending flows, which may cause an increase in PM concentrations close to the ground, below the elevated inversion (Figure 5). Moreover, the development of turbulence causes both the temperature and the PM concentration gradients in the vertical profile to be remarkably diverse.

4. Discussion and Conclusions

Research on the atmospheric boundary layer involving sensors mounted on drones is an interesting alternative to traditional measurement platforms, such as profile in situ measurements or remote sensing techniques. They can be used in various environments, e.g., in polar regions, in varying terrain, or in urban areas [16,19,21].

The studies indicate, first of all, the influence of the boundary layer structure on the concentrations of pollutant and the strongly vertical variability of the parameters analyzed in the vertical profile (Figure S1), which was also confirmed in other studies [22]. The key issue to be solved is the appropriate design of the measurement system for comprehensive atmospheric studies [18], and in the case of using low-cost sensors, their proper calibration [23] to obtain reliable data with high resolution. The use of miniature solutions also allows measurements to be made up to a height of 1000 m above ground level [22]; however, it seems that from the point of view of concentration dynamics and the structure of the ABL, the lowest 300–500 m above the ground is the most important, especially in urban areas [24].

The solution presented in this article, utilizing simultaneous measurements of meteorological variables carried out according to the standards and air quality, allows for a detailed analysis of the influence of the structure of the ABL on air quality. It also complements remote sensing measurements, such as ABL sodar research. Thanks to the modular construction of the measuring head, in addition to basic equipment for measuring temperature, humidity, particulate matter, and ozone concentration, selected devices could be added depending on the scientific needs.