Small-Scale Diffusion Dryer on an Optical Particle Counter for High-Humidity Aerosol Measurements with an Uncrewed Aircraft System

Vasileios Savvakis aDepartment of Geosciences, Eberhard Karls Universität Tübingen, Tübingen, Germany

Search for other papers by Vasileios Savvakis in
Current site
Google Scholar
PubMed
Close
https://orcid.org/0000-0001-9568-1766
,
Martin Schön aDepartment of Geosciences, Eberhard Karls Universität Tübingen, Tübingen, Germany

Search for other papers by Martin Schön in
Current site
Google Scholar
PubMed
Close
,
Matteo Bramati aDepartment of Geosciences, Eberhard Karls Universität Tübingen, Tübingen, Germany

Search for other papers by Matteo Bramati in
Current site
Google Scholar
PubMed
Close
,
Jens Bange aDepartment of Geosciences, Eberhard Karls Universität Tübingen, Tübingen, Germany

Search for other papers by Jens Bange in
Current site
Google Scholar
PubMed
Close
, and
Andreas Platis aDepartment of Geosciences, Eberhard Karls Universität Tübingen, Tübingen, Germany

Search for other papers by Andreas Platis in
Current site
Google Scholar
PubMed
Close
Open access

Abstract

The negative effects of relative humidity to measurements of particulate matter (PM) due to hygroscopic growth are often not inherently handled by low-cost optical particle counters (OPCs). This study presents a new approach in constructing a miniaturized diffusion dryer, for use with an OPC mounted on an uncrewed aircraft system (UAS), namely, the DJI S900 (weight of 7.5 kg and flight endurance of 20 min) for short-term measurements under humid conditions. In this work, an OPC of type N3 (Alphasense) was employed alongside the dryer, with experiments both in the laboratory and outdoors. Evaluation of the dryer’s performance in a fog tank showed effective drying from almost saturated air to 41% relative humidity for 35 min, which is longer than the endurance of the UAS, and therefore sufficient. Changes in the flow rate through the OPC-N3 with the dryer showed a 17% reduction compared to an absent dryer, but the measured PM values remained unaffected. Airborne measurements were taken from four hovering flights near a governmental air pollution station (Mannheim-Nord, Germany) under humid conditions (88%–93%) where the system gave agreeable concentrations when the dryer was in place, but significantly overestimated all PM types without it. At a rural area near the Boundary Layer Field Site Falkenberg (Lindenberg, Germany), operated by the German Meteorological Service (DWD), vertical profiles inside a low-altitude cloud showed sharp increase in concentrations when the UAS entered the cloud layer, demonstrating its capability to accurately detect the layer base.

© 2024 American Meteorological Society. This published article is licensed under the terms of the default AMS reuse license. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Publisher's Note: This article was revised on 14 March 2024 to designate it as open access.

Corresponding author: Vasileios Savvakis, vasileios.savvakis@uni-tuebingen.de

Abstract

The negative effects of relative humidity to measurements of particulate matter (PM) due to hygroscopic growth are often not inherently handled by low-cost optical particle counters (OPCs). This study presents a new approach in constructing a miniaturized diffusion dryer, for use with an OPC mounted on an uncrewed aircraft system (UAS), namely, the DJI S900 (weight of 7.5 kg and flight endurance of 20 min) for short-term measurements under humid conditions. In this work, an OPC of type N3 (Alphasense) was employed alongside the dryer, with experiments both in the laboratory and outdoors. Evaluation of the dryer’s performance in a fog tank showed effective drying from almost saturated air to 41% relative humidity for 35 min, which is longer than the endurance of the UAS, and therefore sufficient. Changes in the flow rate through the OPC-N3 with the dryer showed a 17% reduction compared to an absent dryer, but the measured PM values remained unaffected. Airborne measurements were taken from four hovering flights near a governmental air pollution station (Mannheim-Nord, Germany) under humid conditions (88%–93%) where the system gave agreeable concentrations when the dryer was in place, but significantly overestimated all PM types without it. At a rural area near the Boundary Layer Field Site Falkenberg (Lindenberg, Germany), operated by the German Meteorological Service (DWD), vertical profiles inside a low-altitude cloud showed sharp increase in concentrations when the UAS entered the cloud layer, demonstrating its capability to accurately detect the layer base.

© 2024 American Meteorological Society. This published article is licensed under the terms of the default AMS reuse license. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Publisher's Note: This article was revised on 14 March 2024 to designate it as open access.

Corresponding author: Vasileios Savvakis, vasileios.savvakis@uni-tuebingen.de

1. Introduction

Being a vital constituent of the atmosphere, aerosol particles have been a scientific topic of interest for decades. It is well established that suspended particles in the atmospheric air have detrimental effects on human health (Davidson et al. 2005; Anderson et al. 2012). At the same time, particulate matter (PM) directly and indirectly interacts with the environment, affecting cloud formation and precipitation (Lohmann and Feichter 2005; Andreae and Rosenfeld 2008), the solar radiation budget (Charlson et al. 1992), the evolution of the atmospheric boundary layer (Li et al. 2017), and global warming (Chen et al. 2021). Monitoring of aerosol concentrations is usually attained with properly equipped measurement stations at key locations of urban centers or rural areas, but concentrations often vary on a much smaller spatial scale. This has led to the industrial development of mobile, cost-effective aerosol sensors (Rai et al. 2017), which can be employed in a more flexible manner, improving the spatial resolution of the measurements at desired areas.

As aerosol measurement systems get dimensionally downscaled, the possibility arises to install them on uncrewed aircraft systems (UASs) for more dynamic data acquisition in the horizontal and vertical direction. Fixed-winged aircraft systems have already been used for related research before, for example, measuring vertical profiles of aerosols and black carbon (Corrigan et al. 2008), ice nucleating particles in the lower part of the troposphere (Schrod et al. 2017), Saharan dust episodes (Mamali et al. 2018), or ultrafine particles within the atmospheric boundary layer (ABL) (Altstädter et al. 2015). However, such aircraft often require a large space for takeoff and landing and are generally more expensive than multirotor UAS, which can be employed easier for purposes related to, e.g., urban air quality. Alvarado et al. (2017) employed a small UAS with an OPC-N2 (predecessor of the N3) and also did indoor tests of the rotor effect to the placement of its sampling probe. Air pollution tracking with multirotor UAS is also under development, for example, in studies by Weber et al. (2017), Gu et al. (2018), and Bretschneider et al. (2022). Currently, there are multirotor UAS being used solely for meteorological measurements, for example, by Brosy et al. (2017).

When performing aerosol particle measurements, it is critical to consider the influence of relative humidity (RH), because of the hygroscopic growth effect, which describes the relative growth of a particle’s size as a result of water uptake from the environment (Swietlicki et al. 2008). It has also been observed that hygroscopic growth is affected by the mixing state of the aerosol particles (Cruz and Pandis 2000), and hygroscopic particles can produce different levels of light scattering based on their chemical composition (Tang 1996). Water uptake by the aerosol particle increases its apparent size, which translates to erroneous readings by a sensor like an OPC. Therefore, measurements of a dry airflow are necessary for accurate PM observations. This issue has been previously assessed either by applying thermal treatment to the OPC’s inlet (Irwin et al. 2013; Magi et al. 2020), or by introducing RH-related mathematical corrections to the raw data during the postprocessing of the analysis (Di Antonio et al. 2018; Crilley et al. 2018). Nevertheless, using heated inlets to the sensors is often energy demanding and would require a complicated arrangement to avoid heat losses, whereas the accuracy of the postprocessing corrections decreases when conditions are close to saturation, and they undergo certain sets of assumptions of particle chemical composition for the value of the hygroscopicity factor κ. Furthermore, since ambient RH conditions not only affect the size of the aerosols due to hygroscopic growth, but also their optical properties (Tang and Munkelwitz 1994), the usage of optical-based sensors such as OPCs becomes even more challenging, as demonstrated by Rosati et al. (2015). Since an OPC assumes a refractive index that is typically of a dry particle, this complexity shows how a drying method prior to sampling could prove more physically meaningful than a postprocessing correction based on inaccurate responses from the sensor’s optical detector.

On that account, increased relative humidity causes aerosol–water interactions that have consequences on PM measurements, and has been a recent topic of concern. In the laboratory as well as at ambient conditions, the performance of the low-cost sensor Plantower PMS1003A was evaluated and the results (Jayaratne et al. 2018) showed major increases in particle number concentrations (PNC) during fog events (28% rise of the total number of particles, and 50% rise for particles bigger than 2.5 μm), compared to a particle mass monitor, which featured a charcoal dryer at its inlet. A particle sensor system including the OPC-N3 also showed significant positive bias as RH increased toward 90% (Vogt et al. 2021). Later on, a first attempt to accommodate an inexpensive drying channel on the OPC-N3, which was based on applying voltage to provide thermal energy for moisture extraction, showed improved results for the sensor compared to the same instrument without the dryer (Samad et al. 2021).

Research on creating low-cost drying chambers for OPCs has taken its first steps, for example, by Chacón-Mateos et al. (2022), who constructed one heated drying chamber of 50 cm length for an OPC-R1 (Alphasense 2019). That study limited itself to laboratory testing and the chamber itself, apart from being energy consuming, would not be dimensionwise convenient for portable operation in a real environment. At the same time, such long extra compartments should be chosen wisely considering the strength of the OPC’s internal fan, which becomes less and less efficient the longer the extra tube is. PM2.5 average total bias of 30% was observed for the OPC sensor MASQ under conditions of mild and high pollution in Sarajevo, Bosnia and Herzegovina, over a period of six months between December 2019 and May 2020, which was partly due to hygroscopic growth (Masic et al. 2020). In that study, the spectrometer GRIMM 11-D operated with a self-made diffusion dryer (8 cm external diameter, with 1 kg of silica gel) and it performed much better compared to the reference instrument (an expensive beta attenuation monitor). Since measurements were meant for a longer period of time, the constructed dryer was bigger in dimensions and significantly heavier in weight, than a potential minidiffusion drying chamber for a low-cost OPC, which has not been thoroughly tested yet. On the implementation of drying component for airborne measurements, a recent study using a multirotor UAS with an unmodified OPC-N3 notes the sensitivity of the sensor at higher humidity levels during their measurements as an issue that needs to be addressed preferably prior to postprocessing (Samad et al. 2022). Another study by Platis et al. (2016), featured aerosol particle measurements with a large, fixed-wing UAS and drying-equipped instrumentation, yet the payload of the aircraft as well as the sensor system was significantly higher. This stresses the necessity of lightweight and less costly alternatives. Furthermore, increased strictness in aviation regulations when it comes to uncrewed operations raises the difficulty of using large platforms in certain areas and especially urban centers, which also calls for the development of miniaturized UAS and scientific payloads.

In this study, we present a fully self-constructed particle measurement system that includes meteorological sensors and an OPC with a diffusion drying chamber, designed with computer-aided design (CAD) software, and at the appropriate fitting dimensions for placement as an extension of the instrument’s inlet. The system operates as a scientific payload for a multirotor UAS, specifically the S900 by DJI (China). As commercial diffusion dryers are often fairly expensive, constructed at specific dimensions, as well as overly heavy for use on a low-end OPC, an easily reproduced, economical alternative is suggested here, which provides structural flexibility for experiment or sensor specific requirements. The construction supports an OPC-N3, which was used for this study, or another OPC with similar dimensions. A new approach for the drying chamber’s construction has been used, which consists of 3D printed internal and external tubes instead of commonly used material like copper or stainless steel, and blue silica gel in between to act as the desiccant for diffusion drying. The dryer is easily removable in case wet particles in humid conditions are of interest to be measured.

Using an airtight container with liquid water and an ultrasonic humidifier, fog was created in a controlled environment in the laboratory where the dryer’s capacity of moisture removal from the airflow was tested by measuring RH inside and outside of the fog tank, i.e., before and after the air sample has passed through the chamber that hosts the desiccant. Further experiments collecting airborne data were conducted outdoors, at two locations in Germany: in the city of Mannheim next to a governmental air pollution station under highly humid conditions, and at a field site near Lindenberg (Brandenburg), performing vertical profiles through a low-altitude stratus cloud. Consecutive flights with and without the dryer were performed, to analyze differences due to water uptake and to assess the accuracy of the system on the UAS against reference sensors. With these experiments, both in laboratory conditions as well as outdoors in a realistic environment, how well the drying chamber extracts water vapor from the airflow, as well as the performance of the low cost OPC system as a whole against the reference instrument, was examined. The duration of the drying has to be effective for a time period longer than the flight endurance of the UAS, i.e., at least 20 min long. The main goal of this particular drying channel in this configuration is to achieve a reduction of relative humidity to around 40%, for at least the duration of the UAS’s flight endurance, which can be described as a dry flow where the hygroscopic growth effect is negligible (Held and Mangold 2021). Such a dryer is therefore intended for short-term measurements on a UAS platform, instead of longer-term use on a stationary sampling station.

2. Methods

a. Measurement system

The OPC-N3 (cost of less than EUR 500 at the time of purchase) is used for aerosol particle measurements at a size range of 0.3–40 μm through 24 discrete size bins. The sensor’s weight is approximately 105 g including its fan and it has a typical sampling flow rate of 280 mL min−1, and total flow rate of 5.5 L min−1. As with all OPCs, the sampled airstream goes through a laser beam that is hosted inside the sensor, and scattered light intensity is used to determine particle size, a method that is theoretically described in physics with Mie scattering theory (Drake and Gordon 1985). The diode laser used in the instrument has a wavelength of 658 nm, and a spherical particle shape with a complex refractive index n = 1.5 + 0i and density ρ = 1.65 g cm−3 is assumed internally by the instrument. It should be noted that the sensor itself has an internal calculation for its sampling flow rate (SFR), based on the concept of time of flight of particles traveling through the volume illuminated by the detection laser. The item comes precalibrated by the manufacturer, using mono dispersed polystyrene latex particles, and does not have a built-in battery, but instead requires a power supply between 4.8 and 5.2 V for operation.

For the atmospheric conditions during the time of measurements, an array of 6 miniaturized meteorological sensors (SHT31, Sensirion) is on board the UAS for temperature and humidity measurements. Each sensor is placed below each rotor arm of the S900 (see section 2c for more details on the UAS platform), inside a cylindrical radiation shield with enough airflow through the front and back openings, but not in direct contact with sunlight. The OPC-N3 runs through the use of a companion computer (Raspberry Pi 3b) at a sampling rate of 1 s, with a real-time clock (RTC) for an accurate measurement time stamp and an independent power supply that provides an endurance of approximately 2 h, thus facilitating several UAS flights if needed. The total weight of all components including the dryer amounts to 450 g, which makes it a compact and lightweight payload for UAS operations with multirotor systems such as the S900.

b. Drying channel design and construction

The drying channel’s conceptual basis is diffusion drying and consists of two coaxial cylinders with different diameters, where the desiccant is placed in between and dries the airflow for as long as it passes through the chamber. Two different techniques of additive manufacturing were chosen for each tube, as the inner tube should be perforated, nonporous on its solid surface, and as smooth as possible to produce the least roughness induced turbulence and particle loss possible. For that, masked stereolithography apparatus (MSLA) printing was used (resin printer model Phrozen Sonic Mini), which provides products of high precision (XY-plane resolution of 35 μm, Z-axis resolution of 10 μm) with an LED-based printing technique. The result is significantly lighter than commonly used parts of other diffusion dryers, like stainless steel. The inner diameter of the tube was chosen to match the diameter of the OPC-N3’s own inlet, so that there are no sudden changes in diameter between the different tubings of the system, as that can be another source of particle loss (Muyshondt et al. 1996). Holes of 1.6 mm were designed first following a circular pattern on one side of the dryer, and then expanding the same pattern along its length, while keeping the distances of each hole with its neighboring constant, creating a perforated cylinder with a homogeneous distribution of openings. The outer tube was created with a Prusa I3 MK3, a fused deposition modeling (FDM) open source 3D printer, first released in 2020. Details of the drying chamber that was tested are shown in Table 1.

Table 1.

Dimensions and material used for the design of the drying chamber.

Table 1.

Commercially available blue silica gel was chosen as a desiccant (Wisedry). This silica consists of small almost spherical beads, which are color coded based on the amount of moisture they have absorbed, with deep blue showing completely dry silica and light pink/purple showing saturated silica. Reactivation can be achieved by heating up the beads for a short period of time, making the whole dryer reusable and sustainable. Including the blue silica, its weight is 50 g, considerably lighter than most current alternatives, which grants the possibility for short-term measurement use. A drawing of the drying chamber and its position on the OPC’s inlet is shown in Fig. 1. The printed inner tubes were coated with graphite paint spray to ensure conductivity of the material, which is necessary to avoid aerosol particle losses due to static (Liu et al. 1985).

Fig. 1.
Fig. 1.

Basic sketch of the diffusion drying chamber (arbitrary dimensions). The air goes through the chamber in the inner, perforated, UV resin printed tube, where blue silica gel extracts the water vapor from the flow. When exiting the chamber, the air is dry enough to be sampled at the OPC’s sensing area.

Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-23-0093.1

The presented construction approach accommodates a number of advantages: dimensions or number of products can be decided on demand, as the material cost per dryer is minimal. Its weight allows for usage on a UAS, which requires lightweight scientific payload. On the contrary to bigger, heavier commercial diffusion dryers, where the desiccant alone can often weigh more than 1–2 kg, this dryer stands as an equivalent, miniaturized version that utilizes the same physical process for drying but built more compact, in a way that is ideal for short-term UAS measurements. Its length of 12 cm takes the strength of the OPC-N3’s internal microfan into consideration, as a tubing part that is too long can nullify the provided aspiration and stall the sensor’s measurement process. While exchanging the fan with a stronger pump can be a way to circumvent this problem for cases of longer drying systems, the OPC-N3 was kept intact in this study to maintain the total weight as low as possible. A hyperbolic surfaced inlet is attached on the upper end of the drying chamber to accelerate the airflow inside the drying chamber, thus increasing the sample flow rate (SFR) closer to its typical value stated by the manufacturer. Its shape is based on a previous study by Crazzolara et al. (2019), who used the same UAS platform as in this work for measurements of pollen.

c. Multirotor UAS

The measurement system described in section 2a operates on a multirotor UAS, the S900 by DJI. The S900 is a multirotor UAS with 6 rotor arms and a span of 900 mm (hence the name) from rotor to rotor. Lithium-polymer batteries that grant an endurance from 15 to 25 min depending on the combination, are used for the flights. The platform, including the batteries and the scientific load, weighs about 7.5 kg and has a maximum cruising speed of 14 m s−1, while its ascending/descending speed is maintained at 1.5 m s−1. The OPC system is located in the middle of the platform of the UAS, as close as possible to its center of gravity for more stable flying. Two Styrofoam dome parts are placed on top and below the platform, leaving only the rotors on the outside. More details regarding all the information around the operation of the UAS have been noted by Bramati et al. (2024).

An opening at the middle point of the top Styrofoam dome at exactly the same diameter as the outer part of the bell-shaped inlet of the OPC system provides air samples from the environment for PM measurements. The opening is covered by circular cover on top at a distance of 4 cm, to counteract direct effects by the rotor-induced downwash of the UAS and for protection from dirt falling directly into the inlet (much like the cover shields in stationary air pollution measurement locations). A picture of the UAS with the measurement system fixed on it, along with the two Styrofoam domes, can be seen in Fig. 2.

Fig. 2.
Fig. 2.

The DJI S900, with the OPC measurement system installed on it. The dryer stands in the middle of the construction and aspiration is in the vertical downward direction, first through the dryer channel and then toward the sampling volume of the OPC-N3. Six SHT31 sensors are placed around the platform under each rotor arm. The OPC-N3 is located at the bottom of the supporting table, which is mounted on the main body of the UAS. Once the upper Styrofoam dome is in place, air sampling happens through the opening on top, which has a cover cup protection above it.

Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-23-0093.1

3. Experimental setup

a. Laboratory experiments

Initially, we investigated the effect of the bell-shaped inlet and the extra tube attached to the OPC-N3’s inlet on the SFR and, consequently, the PM values. This was achieved by doing short-duration measurements in laboratory conditions, where the air is generally ventilated and low on aerosol particles, consecutively with and without the drying channel, but also with nothing attached to the sensor at all. Differences were first observed between the three cases and how that translated to the output particle concentrations and the recorded SFR by the system. Following that, the same procedure was followed but this time by removing and replacing only the bell inlet, to see the changes on the same parameters. In this way, it was possible to discover if major measuring discrepancies arise just by the drying channel itself, and how much the output would deviate compared to the same sensor, unmodified.

The drying channel’s performance was then evaluated in a self-constructed fog tank. The fog tank consists of an airtight plastic container box, half full of water and an ultrasonic humidifier dipped inside it. In essence, a conversion of electricity to high-frequency signals produces bubbles at the top of the humidifier, which are then hurled toward the surface and produce a strong mechanical oscillation that decouples liquid droplets from the water, resulting in the creation of fog that goes along the airflow above. Small circular openings were made for ventilation around the upper part of the container, as well as a hole at the appropriate size to fit the OPC inlet/drying channel inner tube diameter, from where fog samples were extracted for testing the channel’s performance under heavily humid conditions (above 95%).

While having an enclosed fog tank at saturated conditions, RH was observed before and after the dryer. One end of the dryer was placed in the tank’s opening of the same diameter, and the other end was in a second, smaller box of ambient air levels of moisture. On that side, an external micropump was placed to ensure a steady airflow from the fog tank through the dryer and into the additional box, similar in power to the strength of the OPC-N3’s fan. Two SHT31 sensors were placed on each side of the dryer, one inside the tank and one in the extra compartment, and RH was measured with a tank full of produced fog from the humidifier. In that way, the difference between the humidity levels of the air between a saturated fog environment and the same air after it has passed through the dryer could be inspected.

Figure 3 shows a sketch of the experimental process. As the flight time of the S900 is limited to 25 min, the effect of the dryer was tested for about 35 min to assure whether or not drying will be effective for a whole flight duration. At the same time, the same setup was also tested without a dryer in place, for obtaining the humidity differences in the ambient air box between the two cases and quantify the drying efficiency.

Fig. 3.
Fig. 3.

Description of the drying chamber testing in a self-constructed fog tank. Fog is produced with an ultrasonic humidifier in an enclosed container, raising the relative humidity levels close to saturation. An SHT31 sensor measures the humidity in the fog tank, and the air is then drawn using a small pump, first through the diffusion dryer, into a second box with ambient air where a second SHT31 sensor is located, which measures the humidity after the humid air passes along the drying chamber.

Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-23-0093.1

b. High-humidity measurements

On 14 February 2023, an OPC system as described in section 2 was employed on board the S900 at urban conditions (Mannheim, Germany) next to a governmental air pollution station that collects data, among others, of aerosol particles. The name of the station is Mannheim-Nord (WGS84 east: 8°27′55.01″, WGS84 north: 49°32′38.68″) and is operated by the State Department of Environment Baden-Württemberg: Landesanstalt für Umwelt Baden-Württemberg (LUBW). On the station, a Fidas 200 (Palas) is installed, which is a high-end optical aerosol spectrometer (OAS) that provides information on PM1, PM2.5, PM4, and PM10. The Fidas 200 is an established measurement system that has also been used as a reference before (Chakraborty et al. 2020; Bílek et al. 2021; Vogt et al. 2021) For taking care of hygroscopic growth at higher humidity levels, the sensor also features its own drying procedure installed by the manufacturer, namely, the Intelligent Aerosol Drying System (IADS).

Two sets of hovering flights were performed next to the station (Fig. 4) and at an altitude of approximately 5 m (to match the height of the station’s sensor inlet at the roof) in the morning (between 1029 and 1119 UTC) and in the afternoon (between 1253 and 1329 UTC) of that day: with the OPC-N3 and a drying chamber, and then without it, consecutively. Specific details of the short four flights can be found in Table 2.

Fig. 4.
Fig. 4.

The S900 in flight (right) next to Mannheim-Nord station (left). The sensors, including the aerosol particle measurements from the Fidas 200, are located at the roof of the station. The distance between the S900 and the station was maintained between 7 and 12 m for safety reasons, and for the UAS downwash to leave the station’s measurements unaffected.

Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-23-0093.1

Table 2.

Takeoff and landing times during the flights on 14 Feb 2023 near Mannheim-Nord station. The RH conditions at the time were collected by the SHT31 sensors on board and the flights for measurements with or without the dryer are also noted.

Table 2.

Humidity conditions at the time of the comparison was high (above 85% in both cases—third column in Table 2) and the relative difference between dried/nondried measurements was analyzed, along with the accuracy of the system with its dryer and while in flight, against the reference instrumentation. Data from the station were available for every 10 s. For the UAS data, an initial resampling to 10 s was done from measurements every 2 s, and then further resampling to 1 min was performed as that is a commonly used averaging time for PM sensors, and to get a clearer trend of the concentrations compared to the Fidas 200.

Mean values with standard deviations of PM1, PM2.5, and PM10 were calculated for the data from the flights, and then compared with the reference, as well as the root-mean-square error (RMSE) differences between the two for each flight case (with a dryer, referred to as “dry,” and without a dryer, referred to as “ambient”). Particle number and volume concentration size distributions were then calculated to determine size-specific effects of hygroscopic growth within the range that amounts up to PM10, i.e., for aerosol particles up to 10 μm.

c. Cloud measurements

Prior to that, the UAS had been used for further measurements at high-humidity conditions at the Boundary Layer Field Site (GM) Falkenberg (WGS84 east: 14°7, WGS84 north: 52°10′, elevation of 73 m above sea level), in Lindenberg, Germany, operated by the German Meteorological Service [Deutscher Wetterdienst (DWD)]. Among a series of operations at the field site, there is a 99-m-high meteorological tower taking measurements of quantities like temperature, relative humidity, air pressure, wind speed and direction, at different heights. Four vertical profiles up to an altitude of 180 m above ground level were performed on 22 November 2022 at the site, in the same philosophy as in the comparison with Mannheim-Nord, two sets of data collection with and without a dryer on the OPC. Table 3 shows the flight specifics.

Table 3.

Takeoff and landing times during the flights on 22 Nov 2022, at the Boundary Layer Field Site Falkenberg. Flights for measurements with or without the dryer are noted.

Table 3.

Temperature and RH profiles at heights 10, 40, 60, and 80 m were taken from Vaisala HMP45D sensors in an aspirated radiation shield (Young 43408), which are the sensors installed on the meteorological tower. During the flights 1 and 2 (first and second rows in Table 3) in the morning, a low altitude stratus cloud was present and visible at an altitude of approximately 65 m above ground. At the location, there is also an operating ceilometer (CHM 15k “NIMBUS,” Lufft GmbH, Germany) that captures cloud altitude heights and extent, which was used to verify that altitude during the time of the early day operations. The UAS performed a vertical profile below it and through the layer until about 180 m above ground level. For flights 3 and 4 (third and fourth rows in Table 3), the cloud layer had disappeared but conditions were still very humid and close to saturation. A similar vertical profile was performed to observe the PM measurement of the OPC-N3 with and without a drying chamber, inside the cloud layer and afterward in its absence. As the UAS ascends at a vertical speed of 1.5 m s−1 to 180 m, the vertical profiles only lasted a few minutes, which are enough for it to rise at that altitude and return on the ground.

4. Results

a. Effect of the dryer on SFR and PM

The outcome of the sample flow rate SFR difference measured by the OPC in case of present or absent drying channel can be seen in Fig. 5. When nothing is attached to the OPC, the SFR is higher and closer to its typical value as stated by the manufacturer, here measured at 4.76 ± 0.25 mL s−1, the mean value among the three measurement periods under this setup. SFR is reduced when either a tube or a dryer of same length is attached by about 0.8 mL s−1, measured at 3.92 ± 0.17 mL s−1 and 3.97 ± 0.13 mL s−1, respectively. This change in SFR has a small impact on the output of PM1, where a value of 0.72 ± 0.18 μg m−3 was measured without anything attached, but 0.79 ± 0.18 μg m−3 with a tube and 0.78 ± 0.18 μg m−3 with a dryer. A similar change is observed for PM2.5, where 1.87 ± 1.0 μg m−3 was measured without anything attached, but 1.98 ± 1.1 μg m−3 with a tube and 1.98 ± 1.1 μg m−3 with a dryer. No significant effect of the tubing/dryer on PM10 can be identified in Fig. 5 (mean values of 6.0, 5.85 and 6.56 μg m−3, respectively).

Fig. 5.
Fig. 5.

Recorded SFR and the three PM values for a time of 43 min in laboratory conditions for the dryer test case. Each colored area at the top graph indicate measurement times where there was no tube attached to the OPC-N3’s inlet (area in gray), a tube but not a dryer was attached (area in brown), or the dryer was attached (area in cyan). Small gaps in between the different measurement stages indicate the time for adjusting or removing the dryer, and were excluded from the analysis.

Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-23-0093.1

A similar approach was followed for looking at the differences between a bell-shaped inlet in place, and an absent one. The results are seen in Fig. 6. The difference between the case of an attached bell and absent bell for the SFR is 0.55 mL s−1, with measured 3.18 ± 0.25 mL s−1 and 3.91 ± 0.25 mL s−1 for each case, respectively. The PM values were comparable: 0.69 ± 0.18 μg m−3 and 0.64 ± 0.17 μg m−3 for PM1, 1.84 ± 1.2 μg m−3 and 1.82 ± 1.0 μg m−3 for PM2.5, and 7.21 μg m−3 and 6.86 μg m−3 for PM10. It can also be clearly seen that the bell-shaped inlet indeed increases the flow through the dryer, reaching a value that is closer to what the sensor would have while in operation without any extra component. Differences on the PM values with or without the dryer fall on the second decimal of the mean value, yet the standard deviation is at least one order of magnitude bigger. This makes the effect of the dryer to the flow through the OPC negligible and can therefore be assumed that there is no significant alteration, caused by the extra channel and inlet attached to it, to the operation with its parent microfan.

Fig. 6.
Fig. 6.

Recorded SFR and the three PM values for a time of 37 min in laboratory conditions for the bell-shaped inlet test case. Each colored area at the top graph indicate measurement times where there was no bell inlet attached to the OPC-N3’s extra tube (area in red), or it was attached (area in gray). Units for the three PM values are μg m−3.

Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-23-0093.1

b. Evaluation in the fog tank

The experiments in the fog tank included RH measurements of the airflow first from inside the saturated container, and then after the airflow has passed through the dryer in an ambient air box. When the humidifier is on, RH levels in the tank (measured by the SHT31 inside it) reach 95%–100%, and condensation is visible in the form of fog which provides an environment similar to highly humid or foggy conditions outdoors.

Experiments in the fog tank consists of two distinct cases:

  1. humid conditions using the dryer (95% humidity inside the tank, present fog but no active humidifier)

  2. humid conditions (as above) without the dryer

As it can be observed in Fig. 7a, highly humid air of 95% RH is present in the fog tank, which is then dried by the drying chamber down to 41%, which shows moisture being removed and the airflow being sufficiently dry for an aerosol measurement with the OPC. A short kick is recognized the moment the pump is turned on, which is balanced by the active blue silica that constantly retains the RH level at the ambient air box low. At the same time, RH in the fog tank is almost constantly at 95% with very low variance. At exactly the same conditions but with no drying chamber connecting the tank and the dryer box (Fig. 7b), the humidity, as expected, does not reduce significantly, but remains the same in the two measurement boxes. A 4% difference in relative humidity is observed after 8 min of run time even in the case of no drying chamber used (Fig. 7b). This is a result of the operation of the micropump itself, which slowly raises the temperature in the ambient air box, which in turn reduces the humidity locally.

Fig. 7.
Fig. 7.

Two different examples of relative humidity measurements at the fog tank, using the constructed drying channels. (a) Fog is present in the fog tank creating high-humidity conditions, but the humidifier is then turned off. The dashed vertical black line indicates the moment when the pump is turned on, beginning the suctioning of wet air through the drying chamber. (b) The conditions are the same as in the first case, but this time there is no drying chamber but a simple copper tube connecting the two boxes.

Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-23-0093.1

c. Airborne measurements

1) Comparison with reference urban station

Results from the comparison between Mannheim-Nord station and the UAS hovering flights in the station, explained in section 3b, are shown in Fig. 8. A distinct difference can be seen between measurements with the drying chamber and without it. It is clear that under such high-humidity conditions, the OPC with its installed drying channel lowers the concentrations and is in acceptable statistical agreement with the reference station for all three PM types. At times with no present dryer, the overestimation is significant. Table 4 contains the mean values and standard deviations calculated for each flight, along with mean values from the reference Fidas 200.

Fig. 8.
Fig. 8.

Measurements at Mannheim-Nord station during the morning (flights 1–2) and afternoon (flights 3–4) of 14 Feb 2023. Two consecutive flights with (i.e. “dry” case) a dryer on the OPC system and without (i.e. “ambient” case) it, for the cases of the three PM types. Measured RH from the SHT31 is shown at the bottom plot, and the solid black line shows the reference measurements from the Fidas 200 at the station. The solid S900 lines indicate 1 min resampled averages, and the less opaque lines of the same colors indicate the nonaveraged data.

Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-23-0093.1

Table 4.

Mean values and standard deviations for each PM type, and for each flight. Flights 1 and 3 include the dryer and flights 2 and 4 do not. Near the measurements from the S900 are shown the PM levels measured by the Fidas 200 at the Mannheim-Nord station. All PM values have units of μg m−3.

Table 4.

During the morning, and at an average RH level of 93.6%, the nondrying OPC on the UAS heavily overestimates the concentrations, as PM1 is 83%, PM2.5 is 397%, and PM10 is 481% higher than the equivalent values of the reference instrument. On the contrary, during the UAS flight with the drying OPC, results are on the same order of magnitude. Specifically, the airborne OPC underestimates PM1 by 13%, slightly overestimates PM2.5 by 3% and measures slightly higher PM10 concentrations too, with an 11% difference from the reference instrument. The standard deviations are higher when no dryer is present as well, and the same behavior remains during the afternoon with RH levels at 86.5%, with significant overestimations during the “ambient” flights (41%, 141%, and 144% for PM1, PM2.5, and PM10). Measurements during the drying OPC flight show similarity with the morning case, as PM1 is underestimated by 8%, but PM2.5 and PM10 are overestimated by 10% and 20%. The results from these four flights show that the dryer provides effective drying at such humid conditions. The average RMSE for the “dry” cases was 3.7 μg m−3 for PM1, compared to 25.95 μg m−3 during the “ambient” cases. The same calculation was 3.4 μg m−3 compared to 107 μg m−3 for PM2.5, and 9.5 μg m−3 compared to 138.2 μg m−3 for PM10.

Figure 9 shows the size distributions of the number and volume concentrations from the flights. Both distributions show a larger difference between “dry” and “ambient” case between 1 and 4 μm, while the same is less obvious for smaller or bigger aerosols in this size range. The distributions also do not follow a completely similar trend because of this, as the aforementioned size ranges seem to get more affected by water uptake due to hygroscopic growth.

Fig. 9.
Fig. 9.

Calculated (left) aerosol particle number and (right) volume size distributions between 0.5 and 9 μm.

Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-23-0093.1

During the airborne measurements, the flow rate through the OPC was determined from the sensor output, as in section 4a. For the first flight, the mean SFR was 3.21 mL s−1, while it was slightly lower during the second flight, at 2.95 mL s−1. This indicates that the flight conditions affect the flow rate through the OPC system further than on the ground, which is expected due to the rotor movement in the vicinity. Nevertheless, calculated PM values are still not negatively affected from this flow rate decrease during the “dry” measurements, as seen from Fig. 8 and Table 4.

2) Vertical profiles inside a cloud

Figure 10 contains the results from consecutive vertical profiling flights, with and without a dryer on the OPC-N3, taken on 22 November 2022 at Falkenberg tower. All flights show measurements only during the ascent of the UAS, to avoid potential downwash effects from its rotors on the PM data of the sensor. A sharp increase in concentrations can be noticed exactly at the bottom altitude level of the cloud layer (measured by the CHM-15k ceilometer), for PM2.5 and PM10, during the morning flights, in both the dried and nondried measurements. For PM1, concentrations do not indicate the cloud layer, which is expressed in the bigger sizes with the larger PM types. During the afternoon (flights 3 and 4 in Fig. 10), all concentrations are considerably lower than in the morning as there is no cloud present at the covered altitude range.

Fig. 10.
Fig. 10.

Measurements inside a low-stratus cloud at Falkenberg tower. The first two flights took place in the morning when the cloud layer was quite low, starting at an altitude of 65 m, as recorded by the installed CHM-15k at the site. (top) Its level is indicated with the black horizontal line for flights 1 and 2. (bottom) In the afternoon (flights 3 and 4), the cloud was not anymore at the same altitude range. The three PM types are depicted under the two cases of using a dryer (i.e., dry) and not using one (i.e., ambient). Units for the three PM values are μg m−3.

Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-23-0093.1

Temperature and humidity up to 99 m from the Falkenberg tower at the location are shown in the upper panel of Fig. 11. The temperature closer to the ground was slightly below 0°C in the morning, and the RH levels were also higher than in the afternoon. It can be seen that as early as 0750 UTC, measured RH at the tower was at saturation at three altitudes (20, 60, and 99 m), while later at 1400 UTC, RH was generally high at all altitudes, with a value below 96% only at 40 m and at saturation at 99 m. At these extremely humid conditions during these flights, a comparison between “dry” and “ambient” measurements from the OPC system on the UAS show, apart from an evident indication of the vertical extent of the cloud layer, the general hygroscopic growth effect on the aerosol particles that have not yet condensed into cloud droplets.

Fig. 11.
Fig. 11.

Humidity and temperature during the time of measurement flights of the UAS, taken from the sensor instrumentation at Falkenberg tower. (top) The two lines correspond to the two points during that day when the vertical profiles where performed, one in the morning and one in the afternoon. (bottom) The cloud layer height, observed by the ceilometer, is shown for a time period of 1 h. The time of flights of the UAS are also indicated in shaded blue, starting from the first vertical profile at 0750 UTC.

Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-23-0093.1

5. Discussion

a. Experiments in the laboratory

The effect of the dryer, at these dimensions and configuration, on the flow rate of the OPC-N3, was first examined in the laboratory by measuring SFR and PM levels while having the extra components in place or not. From Figs. 5 and 6, there is a recognizable difference in SFR levels when either a tube, a dryer, or a bell-shaped inlet is in place, compared to when the sensor is running without anything fixed to it. As the microfan preinstalled in the instrument was built to provide a certain pressure drop, the change is expected, yet from the results it seems that it does not negatively affect the final output of the sensor. Differences between the PM levels due to the dryer alone lie on an order of magnitude that is smaller than the accuracy of the sensors, and SFR levels do not drop drastically in a way that they would completely negate the aspiration of the OPC-N3. Chacón-Mateos et al. (2022) noted that the inner diameter of the low-cost 50 cm dryer of that study for an OPC-R1 was chosen so that the flow rate through the sensor does not have a deviation bigger than 2% compared to the case of an absent dryer. In our case, the deviation is higher on the ground measurements, at about 17%, but it is demonstrated that the PM readings are essentially the same under the different conditions. From this outcome, it is considered that the specific dryer does not interfere with the sensor’s operation to an extent that it would result in false PM measurements.

The dryer was employed in a container where fog was produced through an ultrasonic humidifier dipped in water. Miniaturized meteorological sensors (SHT31) recorded the relative humidity inside the fog tank and in the air after the drying chamber to compare differences between the water vapor levels of the airflow before and after it passed through the blue silica tube (as depicted in Figs. 3 and 7). Figure 7a shows the performance of the drying chamber when in contact with air that has an average level of 95% relative humidity. It seems that after the microfan (to assure an airflow through the system) is turned on, relative humidity after passing the dryer is lowered down to approximately 40%, which is an acceptable level to consider the air as dried, and have also been considered a safe standard for other aerosol measurement studies of the same topic, for example, in the drying chamber by Bezantakos et al. (2018). Without any drying chamber, there is no significant RH drop (this case is the equivalent of using the OPC at a humid environment without a drying method). The duration of the fog tank test amounts for a time of more than 30 min, longer than the flight endurance of the UAS on which the OPC system is intended to be used. As the dryer is easily reproducible, multiple units can be used for multiple flights and it becomes vital that the dryer is effective for at least one flight, which is achieved here as shown in Fig. 7.

b. Outdoor measurements

For the urban experiments near Mannheim-Nord, the OPC-N3 with a dryer on the UAS agree well with the monitoring station using a certified Fidas 200. The underestimation of PM1 is an anticipated result as the Fidas 200 has a lowest detectable size boundary of 0.18 μm, which is lower than the one of the OPC-N3 at 0.35 μm. This means that the Fidas 200 is able to detect even smaller aerosol particles, which are all eventually amounted for the final PM1 value it records. Thus, the difference between the PM1 values between the two systems could be explained by the fact that particles between 0.18 and 0.35 μm were detected by the Fidas 200 but not by the OPC-N3. This discrepancy is most evident in that PM type but not in the bigger sizes, as the OPC-N3 PM2.5 outputs were in agreement with the Fidas 200. PM10 was overestimated even with in the case of the “dry” measurements, which is a result that has been noted before for the OPC-N3 (Vogt et al. 2021). In that evaluation study, which compared a few low-cost sensors and an OPC-N3, relative humidity problems were also identified for that specific instrument and a difficulty to measure larger particles was also pointed out. The authors noted how all the low-cost sensors did not reach a high agreement with their reference instrumentation for bigger sizes. However, our study focuses on the relative difference between “dry” and “ambient” measurements and from that perspective, the installed dryer evidently dried the airflow effectively for the duration of the flights. Noteworthy is that the overestimation of “dry” measurements increased from the morning to the afternoon, despite the decrease in RH, which may be due to the lowering efficiency of the desiccant, exposed to such high-humidity conditions, as hours passed by.

Numerous studies have already pointed out the measurement overestimation of low-cost OPCs at high-humidity conditions. For example, Samad et al. (2021) found that an OPC-N3 without a drying chamber measured almost double the PM amounts compared to the same sensor with its installed drying chamber, at a laboratory test in a climate chamber at high-humidity conditions. Badura et al. (2018) arrived at similar conclusions were drawn in a longer-term experiment of PM2.5 measurements in real-life conditions for an OPC-N2 (which suffers from the same issue as the OPC-N3 when it comes to humidity), when measurement conditions were under RH of 80% or more. Samad et al. (2022) noted PM2.5 overestimation at certain times during their UAS flights compared to the reference, which were attributed to the RH conditions at the time. In the case of the Mannheim experiments, also with airborne measurements, there is significant overestimation from the system when no dryer was attached, as seen in Fig. 8 and at the PM mean value calculations at Table 4. This can be attributed to the same reason, as the flights with the dryer show concentrations much closer to the reference values to the Fidas 200. There is a lower level of overestimation in the “ambient” afternoon flights compared to the flights in the morning, which shows how much a reduction of RH from close to 95% to a bit less than 90% can affect PM measurements on an OPC without a component to remove the water vapor from the airflow.

Size distributions of number and volume concentrations in Fig. 9 also show different water uptake at different sizes, which is likely a result of the specific particle chemical composition at the given time of measurements. Such a result supports the notion of not applying postprocessing solutions to collected nondried measurements, as hygroscopic growth can be different throughout different aerosol size modes, but also chemical composition and type. Since such information cannot be retracted solely from an optical sensor such as the OPC-N3, it is difficult to assess the reasons behind size-specific hygroscopic growth differences. A previous study including crewed aircraft flights was done by Hegg et al. (2007), where different hygroscopic growth factors at different altitudes, but also at different sizes between 0.5 and 3.2 μm, were observed. The authors of that work stated that particles differed in composition across the covered size range, which is presumably the case for our measurements too. Another hygroscopicity study for particles up to 10 μm in size revealed higher hygroscopic growth factors in the micron range and up to 5 μm for RH = 91% (Eichler et al. 2008), which was a comparable result to the depicted distribution in Fig. 9. It has been shown that there is a proportionality of hygroscopic growth factors for continental submicron- and micron-sized aerosols (Zhang et al. 2014), which may imply the presence of particles from other sources during the Mannheim measurements, which were more hygroscopic in the micron range. This emphasizes the effect of particle mixing and composition on the resulting growth due to water uptake.

During the vertical profiling flights at Falkenberg tower, a low-altitude cloud was over the area during the morning of 22 November 2022, but later dissipated during the day and was gone in the afternoon, while RH conditions remained high (Fig. 11). The ceilometer recorded a cloud layer altitude between 62 m (at 0810 UTC) and 65 m (at 0800 UTC) in the area, and Fig. 10 clearly shows a significant rise in PM10 as the UAS entered the cloud at that altitude. At the same time, the RH sensors on the tower seem to be at saturation at their highest altitude of 99 m. This indicates that measuring at such conditions with two identical OPCs that only differ on the presence of a dryer could accurately show the altitude of a cloud ceiling when standard RH sensors have slower response time or get saturated. From Fig. 10, it can also be observed that the PM concentrations under the cloud level were relatively low, and on the same order of magnitude as the concentrations during the afternoon flights for the whole vertical profile, showcasing how the PM difference above 65 m between the two sets of flights, comes from the cloud only present in the morning. It is important to note that the SFR during the vertical profiles was lower than in the laboratory tests (section 4a) as well as the hovering flights at Mannheim (section 4c), ranging from 1 to 1.5 mL s−1, indicating how the UAS and atmospheric conditions affected the flow rate through the sensor. This SFR change however did not decrease the measurement accuracy, as the OPC data also showed generally low raw bin counts, i.e., sampling less volume of air, but also less particles than what it would, at a higher SFR. This can be realized from the high PM levels especially for PM10 during the morning flights (Fig. 10), where peaks reached 250 μg m−3 during the flight with a dryer, and 750 μg m−3 during the flight without it.

As water droplets have different optical properties than dry aerosol particles, the implications regarding their interaction with an optical sensor such as an OPC-N3 were not within the objectives of this work, which only assesses the drying procedure. However, a recent study by Nurowska et al. (2023) examined the usage of an OPC-N3 specifically for fog measurements in detail, and the authors concluded that the sensor measures fog droplets, as well as aerosol particles that are wetted from water vapor in the environment (i.e., hygroscopically grown). In our morning vertical profiles (flights 1 and 2 in Fig. 10), both “dry” and “ambient” measurements rise above the 65 m mark for PM2.5 and PM10, which leads to the conclusion that the constructed dryer was able to only partially dry the sampled air when liquid water was present. Hence, it is more appropriate for measurements at high-humidity conditions, when condensation has not occurred yet. Still, the base height of the cloud was accurately detected, and the difference between the two profiles can give an indication of the amount of noncondensed aerosol particles across the layer. Later on in the day (flights 3 and 4 in Fig. 10), concentrations are significantly lower and more comparable to how they were during the morning below the cloud layer, i.e., showing that the cloud has passed. We assume that the vertical extent of the cloud was in the lower part of the atmosphere and just some tens of meters above the highest point of the UAS at that time, when considering the potentially major decrease that started evolving at the very top of the PM2.5 and PM10 vertical profiles during the morning flights. However, ascent was not continued due to considerations about rotor icing at such cold temperatures (already 0°C at ground level), which would severely compromise the flying operations.

6. Conclusions

A compact, lightweight diffusion drying channel was built and tested with an OPC-N3 on a UAS for short-term measurements in highly humid conditions. Its dimensions and weight provide for a flexible and sustainable solution for UAS data collection, which requires components smaller and lighter than their ground-based counterparts. The effect of the dryer was tested in a fog tank where drying from almost saturated to 40% RH was achieved for at least half an hour. The dryer’s length (12 cm) and attached bell-shaped inlet on top also did not significantly interfere with the sensor’s aspiration process, which stems from the use of a microfan. More meticulous analysis of the flow inside the dryer and potential particle loss through its inner chamber will be the focus of a future, separate study.

In this work, the question was approached by looking at the final SFR levels through the OPC-N3 and the corresponding PM values, which remained reasonable. This was also further evaluated during hovering flights near the reference station Mannheim-Nord at conditions of high humidity, where effective drying was demonstrated by comparing measurements with and without the dryer, against the tower’s instrumentation as reference. The rotor-induced downwash of this specific UAS platform (S900 + Styrofoam encasing) is also another point of a follow-up study, which will indicate more clearly the ideal placement of the sensor or dryer. However, the adjustability of such a system is high, as it is self-designed and can be modified further for future experiments. While the effect of the rotor flow to the resulting sampling flow through the inlet and dryer was not directly inspected, the resulting SFR values during flight still gave reasonable values for the PM concentrations, as it is apparent from the comparison with the reference, Fidas 200.

Vertical profiles inside a low-altitude cloud revealed that the OPC-N3 can accurately detect the cloud layer base, which was aligned with the indication from the ceilometer. When used in combination with a dual measurement with and without a dryer, it can give further information on the amount of fog droplets and humidified aerosols. An OPC can potentially record a sharp change in concentrations when the cloud was encountered, and therefore pinpoint cloud ceiling with high accuracy. For future work, a more compact UAS, suited for measurements in cold and icy conditions, with two mounted OPCs for capturing ambient and dry concentrations simultaneously, is being developed and will be further used with a drying channel as described currently. Furthermore, replacement of the parent microfan with a steady pump can accommodate for less flow rate undulations, but with careful consideration of the weight limits related to UAS operations. This small dryer showcased that it can provide adequate drying for short-term measurements even in highly humid air, and could prove useful in many future UAS activities that include miniaturized aerosol particle sensors such as OPCs.

Acknowledgments.

We thank the Baden-Württemberg State Institute for the Environment (LUBW) for providing the data from Mannheim-Nord station for the comparisons conducted in the paper. Furthermore, we appreciate data provided by DWD from the Falkenberg site and specifically the help of Dr. Frank Beyrich during our flight activities and the hospitality during our stay in Lindenberg for the experimental campaign. This work is partly funded by the European Union Horizon 2020 research and innovation program under Grant Agreement 861291 as part of the Train2Wind Marie Skłodowska-Curie Innovation Training Network (https://www.train2wind.eu/). Partial funding was also acquired from the Federal Ministry for Education and Research for Project Funding (NABF) by the German Weather Service (DWD) in Germany, under grant agreement with a reference: 4819EMF01. The authors of the manuscript declare no conflicts of interest.

Data availability statement.

Raw sensor outputs, processing and data acquisition scripts can be shared from the article’s lead author, along with necessary clarifications, upon reasonable request.

REFERENCES

  • Alphasense, 2019: User manual: OPC-R1 optical particle counter. Alphasense Doc., 33 pp.

  • Altstädter, B., and Coauthors, 2015: ALADINA—An unmanned research aircraft for observing vertical and horizontal distributions of ultrafine particles within the atmospheric boundary layer. Atmos. Meas. Tech., 8, 16271639, https://doi.org/10.5194/amt-8-1627-2015.

    • Search Google Scholar
    • Export Citation
  • Alvarado, M., F. Gonzalez, P. Erskine, D. Cliff, and D. Heuff, 2017: A methodology to monitor airborne PM10 dust particles using a small unmanned aerial vehicle. Sensors, 17, 343, https://doi.org/10.3390/s17020343.

    • Search Google Scholar
    • Export Citation
  • Anderson, J. O., J. G. Thundiyil, and A. Stolbach, 2012: Clearing the air: A review of the effects of particulate matter air pollution on human health. Med. Toxicol., 8, 166175, https://doi.org/10.1007/s13181-011-0203-1.

    • Search Google Scholar
    • Export Citation
  • Andreae, M. O., and D. Rosenfeld, 2008: Aerosol–cloud–precipitation interactions. Part 1. The nature and sources of cloud-active aerosols. Earth-Sci. Rev., 89, 1341, https://doi.org/10.1016/j.earscirev.2008.03.001.

    • Search Google Scholar
    • Export Citation
  • Badura, M., P. Batog, A. Drzeniecka-Osiadacz, and P. Modzel, 2018: Evaluation of low-cost sensors for ambient PM2.5 monitoring. J. Sens., 2018, 5096540, https://doi.org/10.1155/2018/5096540.

    • Search Google Scholar
    • Export Citation
  • Bezantakos, S., F. Schmidt-Ott, and G. Biskos, 2018: Performance evaluation of the cost-effective and lightweight Alphasense optical particle counter for use onboard unmanned aerial vehicles. Aerosol Sci. Technol., 52, 385392, https://doi.org/10.1080/02786826.2017.1412394.

    • Search Google Scholar
    • Export Citation
  • Bílek, J., O. Bílek, P. Maršolek, and P. Buček, 2021: Ambient air quality measurement with low-cost optical and electrochemical sensors: An evaluation of continuous year-long operation. Environments, 8, 114, https://doi.org/10.3390/environments8110114.

    • Search Google Scholar
    • Export Citation
  • Bramati, M., M. Schön, D. Schulz, V. Savvakis, Y. Wang, J. Bange, and A. Platis, 2024: A versatile calibration method for rotary-wing UAS as wind measurement systems. J. Atmos. Ocean. Technol., 41, 2543, https://doi.org/10.1175/JTECH-D-23-0010.1.

    • Search Google Scholar
    • Export Citation
  • Bretschneider, L., and Coauthors, 2022: MesSBAR—Multicopter and instrumentation for air quality research. Atmosphere, 13, 629, https://doi.org/10.3390/atmos13040629.

    • Search Google Scholar
    • Export Citation
  • Brosy, C., K. Krampf, M. Zeeman, B. Wolf, W. Junkermann, K. Schäfer, S. Emeis, and H. Kunstmann, 2017: Simultaneous multicopter-based air sampling and sensing of meteorological variables. Atmos. Meas. Tech., 10, 27732784, https://doi.org/10.5194/amt-10-2773-2017.

    • Search Google Scholar
    • Export Citation
  • Chacón-Mateos, M., B. Laquai, U. Vogt, and C. Stubenrauch, 2022: Evaluation of a low-cost dryer for a low-cost optical particle counter. Atmos. Meas. Tech., 15, 73957410, https://doi.org/10.5194/amt-15-7395-2022.

    • Search Google Scholar
    • Export Citation
  • Chakraborty, R., J. Heydon, M. Mayfield, and L. Mihaylova, 2020: Indoor air pollution from residential stoves: Examining the flooding of particulate matter into homes during real-world use. Atmosphere, 11, 1326, https://doi.org/10.3390/atmos11121326.

    • Search Google Scholar
    • Export Citation
  • Charlson, R. J., S. E. Schwartz, J. M. Hales, R. D. Cess, J. A. Coakley Jr., J. E. Hansen, and D. J. Hofmann, 1992: Climate forcing by anthropogenic aerosols. Science, 255, 423430, https://doi.org/10.1126/science.255.5043.423.

    • Search Google Scholar
    • Export Citation
  • Chen, S.-L., S.-W. Chang, Y.-J. Chen, and H.-L. Chen, 2021: Possible warming effect of fine particulate matter in the atmosphere. Commun. Earth Environ., 2, 208, https://doi.org/10.1038/s43247-021-00278-5.

    • Search Google Scholar
    • Export Citation
  • Corrigan, C. E., G. C. Roberts, M. V. Ramana, D. Kim, and V. Ramanathan, 2008: Capturing vertical profiles of aerosols and black carbon over the Indian Ocean using autonomous unmanned aerial vehicles. Atmos. Chem. Phys., 8, 737747, https://doi.org/10.5194/acp-8-737-2008.

    • Search Google Scholar
    • Export Citation
  • Crazzolara, C., M. Ebner, A. Platis, T. Miranda, J. Bange, and A. Junginger, 2019: A new multicopter-based unmanned aerial system for pollen and spores collection in the atmospheric boundary layer. Atmos. Meas. Tech., 12, 15811598, https://doi.org/10.5194/amt-12-1581-2019.

    • Search Google Scholar
    • Export Citation
  • Crilley, L. R., M. Shaw, R. Pound, L. J. Kramer, R. Price, S. Young, A. C. Lewis, and F. D. Pope, 2018: Evaluation of a low-cost optical particle counter (Alphasense OPC-N2) for ambient air monitoring. Atmos. Meas. Tech., 11, 709720, https://doi.org/10.5194/amt-11-709-2018.

    • Search Google Scholar
    • Export Citation
  • Cruz, C. N., and S. N. Pandis, 2000: Deliquescence and hygroscopic growth of mixed inorganic–organic atmospheric aerosol. Environ. Sci. Technol., 34, 43134319, https://doi.org/10.1021/es9907109.

    • Search Google Scholar
    • Export Citation
  • Davidson, C. I., R. F. Phalen, and P. A. Solomon, 2005: Airborne particulate matter and human health: A review. Aerosol Sci. Technol., 39, 737749, https://doi.org/10.1080/02786820500191348.

    • Search Google Scholar
    • Export Citation
  • Di Antonio, A., O. A. M. Popoola, B. Ouyang, J. Saffell, and R. L. Jones, 2018: Developing a relative humidity correction for low-cost sensors measuring ambient particulate matter. Sensors, 18, 2790, https://doi.org/10.3390/s18092790.

    • Search Google Scholar
    • Export Citation
  • Drake, R. M., and J. E. Gordon, 1985: Mie scattering. Amer. J. Phys., 53, 955962, https://doi.org/10.1119/1.14011.

  • Eichler, H., and Coauthors, 2008: Hygroscopic properties and extinction of aerosol particles at ambient relative humidity in south-eastern China. Atmos. Environ., 42, 63216334, https://doi.org/10.1016/j.atmosenv.2008.05.007.

    • Search Google Scholar
    • Export Citation
  • Gu, Q., D. R. Michanowicz, and C. Jia, 2018: Developing a modular unmanned aerial vehicle (UAV) platform for air pollution profiling. Sensors, 18, 4363, https://doi.org/10.3390/s18124363.

    • Search Google Scholar
    • Export Citation
  • Hegg, D. A., D. S. Covert, H. Jonsson, and P. A. Covert, 2007: An instrument for measuring size-resolved aerosol hygroscopicity at both sub- and super-micron sizes. Aerosol Sci. Technol., 41, 873883, https://doi.org/10.1080/02786820701506955.

    • Search Google Scholar
    • Export Citation
  • Held, A., and A. Mangold, 2021: Measurement of fundamental aerosol physical properties. Springer Handbook of Atmospheric Measurements, T. Foken, Ed., Springer, 535–565.

  • Irwin, M., Y. Kondo, N. Moteki, and T. Miyakawa, 2013: Evaluation of a heated-inlet for calibration of the SP2. Aerosol Sci. Technol., 47, 895905, https://doi.org/10.1080/02786826.2013.800187.

    • Search Google Scholar
    • Export Citation
  • Jayaratne, R., X. Liu, P. Thai, M. Dunbabin, and L. Morawska, 2018: The influence of humidity on the performance of a low-cost air particle mass sensor and the effect of atmospheric fog. Atmos. Meas. Tech., 11, 48834890, https://doi.org/10.5194/amt-11-4883-2018.

    • Search Google Scholar
    • Export Citation
  • Li, Z., and Coauthors, 2017: Aerosol and boundary-layer interactions and impact on air quality. Natl. Sci. Rev., 4, 810833, https://doi.org/10.1093/nsr/nwx117.

    • Search Google Scholar
    • Export Citation
  • Liu, B. Y., D. Y. Pui, K. L. Rubow, and W. W. Szymanski, 1985: Electrostatic effects in aerosol sampling and filtration. Ann. Occup. Hyg., 29, 251269, https://doi.org/10.1093/annhyg/29.2.251.

    • Search Google Scholar
    • Export Citation
  • Lohmann, U., and J. Feichter, 2005: Global indirect aerosol effects: A review. Atmos. Chem. Phys., 5, 715737, https://doi.org/10.5194/acp-5-715-2005.

    • Search Google Scholar
    • Export Citation
  • Magi, B. I., C. Cupini, J. Francis, M. Green, and C. Hauser, 2020: Evaluation of PM2.5 measured in an urban setting using a low-cost optical particle counter and a federal equivalent method beta attenuation monitor. Aerosol Sci. Technol., 54, 147159, https://doi.org/10.1080/02786826.2019.1619915.

    • Search Google Scholar
    • Export Citation
  • Mamali, D., and Coauthors, 2018: Vertical profiles of aerosol mass concentration derived by unmanned airborne in situ and remote sensing instruments during dust events. Atmos. Meas. Tech., 11, 28972910, https://doi.org/10.5194/amt-11-2897-2018.

    • Search Google Scholar
    • Export Citation
  • Masic, A., D. Bibic, B. Pikula, A. Blazevic, J. Huremovic, and S. Zero, 2020: Evaluation of optical particulate matter sensors under realistic conditions of strong and mild urban pollution. Atmos. Meas. Tech., 13, 64276443, https://doi.org/10.5194/amt-13-6427-2020.

    • Search Google Scholar
    • Export Citation
  • Muyshondt, A., A. R. McFarland, and N. K. Anand, 1996: Deposition of aerosol particles in contraction fittings. Aerosol Sci. Technol., 24, 205216, https://doi.org/10.1080/02786829608965364.

    • Search Google Scholar
    • Export Citation
  • Nurowska, K., M. Mohammadi, S. Malinowski, and K. Markowicz, 2023: Applicability of the low-cost OPC-N3 optical particle counter for microphysical measurements of fog. Atmos. Meas. Tech., 16, 24152430, https://doi.org/10.5194/amt-16-2415-2023.

    • Search Google Scholar
    • Export Citation
  • Platis, A., B. Altstädter, B. Wehner, N. Wildmann, A. Lampert, M. Hermann, W. Birmili, and J. Bange, 2016: An observational case study on the influence of atmospheric boundary-layer dynamics on new particle formation. Bound.-Layer Meteor., 158, 6792, https://doi.org/10.1007/s10546-015-0084-y.

    • Search Google Scholar
    • Export Citation
  • Rai, A. C., P. Kumar, F. Pilla, A. N. Skouloudis, S. Di Sabatino, C. Ratti, A. Yasar, and D. Rickerby, 2017: End-user perspective of low-cost sensors for outdoor air pollution monitoring. Sci. Total Environ., 607–608, 691705, https://doi.org/10.1016/j.scitotenv.2017.06.266.

    • Search Google Scholar
    • Export Citation
  • Rosati, B., G. Wehrle, M. Gysel, P. Zieger, U. Baltensperger, and E. Weingartner, 2015: The white-light humidified optical particle spectrometer (WHOPS)—A novel airborne system to characterize aerosol hygroscopicity. Atmos. Meas. Tech., 8, 921939, https://doi.org/10.5194/amt-8-921-2015.

    • Search Google Scholar
    • Export Citation
  • Samad, A., F. E. Melchor Mimiaga, B. Laquai, and U. Vogt, 2021: Investigating a low-cost dryer designed for low-cost PM sensors measuring ambient air quality. Sensors, 21, 804, https://doi.org/10.3390/s21030804.

    • Search Google Scholar
    • Export Citation
  • Samad, A., D. Alvarez Florez, I. Chourdakis, and U. Vogt, 2022: Concept of using an unmanned aerial vehicle (UAV) for 3D investigation of air quality in the atmosphere—Example of measurements near a roadside. Atmosphere, 13, 663, https://doi.org/10.3390/atmos13050663.

    • Search Google Scholar
    • Export Citation
  • Schrod, J., and Coauthors, 2017: Ice nucleating particles over the eastern Mediterranean measured by unmanned aircraft systems. Atmos. Chem. Phys., 17, 48174835, https://doi.org/10.5194/acp-17-4817-2017.

    • Search Google Scholar
    • Export Citation
  • Swietlicki, E., and Coauthors, 2008: Hygroscopic properties of submicrometer atmospheric aerosol particles measured with H-TDMA instruments in various environments—A review. Tellus, 60B, 432469, https://doi.org/10.1111/j.1600-0889.2008.00350.x.

    • Search Google Scholar
    • Export Citation
  • Tang, I. N., 1996: Chemical and size effects of hygroscopic aerosols on light scattering coefficients. J. Geophys. Res., 101, 19 24519 250, https://doi.org/10.1029/96JD03003.

    • Search Google Scholar
    • Export Citation
  • Tang, I. N., and H. R. Munkelwitz, 1994: Water activities, densities, and refractive indices of aqueous sulfates and sodium nitrate droplets of atmospheric importance. J. Geophys. Res., 99, 18 80118 808, https://doi.org/10.1029/94JD01345.

    • Search Google Scholar
    • Export Citation
  • Vogt, M., P. Schneider, N. Castell, and P. Hamer, 2021: Assessment of low-cost particulate matter sensor systems against optical and gravimetric methods in a field co-location in Norway. Atmosphere, 12, 961, https://doi.org/10.3390/atmos12080961.

    • Search Google Scholar
    • Export Citation
  • Weber, K., G. Heweling, C. Fischer, and M. Lange, 2017: The use of an octocopter UAV for the determination of air pollutants—A case study of the traffic induced pollution plume around a river bridge in Duesseldorf, Germany. Int. J. Educ., 2, 6366.

    • Search Google Scholar
    • Export Citation
  • Zhang, X., P. Massoli, P. K. Quinn, T. S. Bates, and C. D. Cappa, 2014: Hygroscopic growth of submicron and supermicron aerosols in the marine boundary layer. J. Geophys. Res. Atmos., 119, 83848399, https://doi.org/10.1002/2013JD021213.

    • Search Google Scholar
    • Export Citation
Save
  • Alphasense, 2019: User manual: OPC-R1 optical particle counter. Alphasense Doc., 33 pp.

  • Altstädter, B., and Coauthors, 2015: ALADINA—An unmanned research aircraft for observing vertical and horizontal distributions of ultrafine particles within the atmospheric boundary layer. Atmos. Meas. Tech., 8, 16271639, https://doi.org/10.5194/amt-8-1627-2015.

    • Search Google Scholar
    • Export Citation
  • Alvarado, M., F. Gonzalez, P. Erskine, D. Cliff, and D. Heuff, 2017: A methodology to monitor airborne PM10 dust particles using a small unmanned aerial vehicle. Sensors, 17, 343, https://doi.org/10.3390/s17020343.

    • Search Google Scholar
    • Export Citation
  • Anderson, J. O., J. G. Thundiyil, and A. Stolbach, 2012: Clearing the air: A review of the effects of particulate matter air pollution on human health. Med. Toxicol., 8, 166175, https://doi.org/10.1007/s13181-011-0203-1.

    • Search Google Scholar
    • Export Citation
  • Andreae, M. O., and D. Rosenfeld, 2008: Aerosol–cloud–precipitation interactions. Part 1. The nature and sources of cloud-active aerosols. Earth-Sci. Rev., 89, 1341, https://doi.org/10.1016/j.earscirev.2008.03.001.

    • Search Google Scholar
    • Export Citation
  • Badura, M., P. Batog, A. Drzeniecka-Osiadacz, and P. Modzel, 2018: Evaluation of low-cost sensors for ambient PM2.5 monitoring. J. Sens., 2018, 5096540, https://doi.org/10.1155/2018/5096540.

    • Search Google Scholar
    • Export Citation
  • Bezantakos, S., F. Schmidt-Ott, and G. Biskos, 2018: Performance evaluation of the cost-effective and lightweight Alphasense optical particle counter for use onboard unmanned aerial vehicles. Aerosol Sci. Technol., 52, 385392, https://doi.org/10.1080/02786826.2017.1412394.

    • Search Google Scholar
    • Export Citation
  • Bílek, J., O. Bílek, P. Maršolek, and P. Buček, 2021: Ambient air quality measurement with low-cost optical and electrochemical sensors: An evaluation of continuous year-long operation. Environments, 8, 114, https://doi.org/10.3390/environments8110114.

    • Search Google Scholar
    • Export Citation
  • Bramati, M., M. Schön, D. Schulz, V. Savvakis, Y. Wang, J. Bange, and A. Platis, 2024: A versatile calibration method for rotary-wing UAS as wind measurement systems. J. Atmos. Ocean. Technol., 41, 2543, https://doi.org/10.1175/JTECH-D-23-0010.1.

    • Search Google Scholar
    • Export Citation
  • Bretschneider, L., and Coauthors, 2022: MesSBAR—Multicopter and instrumentation for air quality research. Atmosphere, 13, 629, https://doi.org/10.3390/atmos13040629.

    • Search Google Scholar
    • Export Citation
  • Brosy, C., K. Krampf, M. Zeeman, B. Wolf, W. Junkermann, K. Schäfer, S. Emeis, and H. Kunstmann, 2017: Simultaneous multicopter-based air sampling and sensing of meteorological variables. Atmos. Meas. Tech., 10, 27732784, https://doi.org/10.5194/amt-10-2773-2017.

    • Search Google Scholar
    • Export Citation
  • Chacón-Mateos, M., B. Laquai, U. Vogt, and C. Stubenrauch, 2022: Evaluation of a low-cost dryer for a low-cost optical particle counter. Atmos. Meas. Tech., 15, 73957410, https://doi.org/10.5194/amt-15-7395-2022.

    • Search Google Scholar
    • Export Citation
  • Chakraborty, R., J. Heydon, M. Mayfield, and L. Mihaylova, 2020: Indoor air pollution from residential stoves: Examining the flooding of particulate matter into homes during real-world use. Atmosphere, 11, 1326, https://doi.org/10.3390/atmos11121326.

    • Search Google Scholar
    • Export Citation
  • Charlson, R. J., S. E. Schwartz, J. M. Hales, R. D. Cess, J. A. Coakley Jr., J. E. Hansen, and D. J. Hofmann, 1992: Climate forcing by anthropogenic aerosols. Science, 255, 423430, https://doi.org/10.1126/science.255.5043.423.

    • Search Google Scholar
    • Export Citation
  • Chen, S.-L., S.-W. Chang, Y.-J. Chen, and H.-L. Chen, 2021: Possible warming effect of fine particulate matter in the atmosphere. Commun. Earth Environ., 2, 208, https://doi.org/10.1038/s43247-021-00278-5.

    • Search Google Scholar
    • Export Citation
  • Corrigan, C. E., G. C. Roberts, M. V. Ramana, D. Kim, and V. Ramanathan, 2008: Capturing vertical profiles of aerosols and black carbon over the Indian Ocean using autonomous unmanned aerial vehicles. Atmos. Chem. Phys., 8, 737747, https://doi.org/10.5194/acp-8-737-2008.

    • Search Google Scholar
    • Export Citation
  • Crazzolara, C., M. Ebner, A. Platis, T. Miranda, J. Bange, and A. Junginger, 2019: A new multicopter-based unmanned aerial system for pollen and spores collection in the atmospheric boundary layer. Atmos. Meas. Tech., 12, 15811598, https://doi.org/10.5194/amt-12-1581-2019.

    • Search Google Scholar
    • Export Citation
  • Crilley, L. R., M. Shaw, R. Pound, L. J. Kramer, R. Price, S. Young, A. C. Lewis, and F. D. Pope, 2018: Evaluation of a low-cost optical particle counter (Alphasense OPC-N2) for ambient air monitoring. Atmos. Meas. Tech., 11, 709720, https://doi.org/10.5194/amt-11-709-2018.

    • Search Google Scholar
    • Export Citation
  • Cruz, C. N., and S. N. Pandis, 2000: Deliquescence and hygroscopic growth of mixed inorganic–organic atmospheric aerosol. Environ. Sci. Technol., 34, 43134319, https://doi.org/10.1021/es9907109.

    • Search Google Scholar
    • Export Citation
  • Davidson, C. I., R. F. Phalen, and P. A. Solomon, 2005: Airborne particulate matter and human health: A review. Aerosol Sci. Technol., 39, 737749, https://doi.org/10.1080/02786820500191348.

    • Search Google Scholar
    • Export Citation
  • Di Antonio, A., O. A. M. Popoola, B. Ouyang, J. Saffell, and R. L. Jones, 2018: Developing a relative humidity correction for low-cost sensors measuring ambient particulate matter. Sensors, 18, 2790, https://doi.org/10.3390/s18092790.

    • Search Google Scholar
    • Export Citation
  • Drake, R. M., and J. E. Gordon, 1985: Mie scattering. Amer. J. Phys., 53, 955962, https://doi.org/10.1119/1.14011.

  • Eichler, H., and Coauthors, 2008: Hygroscopic properties and extinction of aerosol particles at ambient relative humidity in south-eastern China. Atmos. Environ., 42, 63216334, https://doi.org/10.1016/j.atmosenv.2008.05.007.

    • Search Google Scholar
    • Export Citation
  • Gu, Q., D. R. Michanowicz, and C. Jia, 2018: Developing a modular unmanned aerial vehicle (UAV) platform for air pollution profiling. Sensors, 18, 4363, https://doi.org/10.3390/s18124363.

    • Search Google Scholar
    • Export Citation
  • Hegg, D. A., D. S. Covert, H. Jonsson, and P. A. Covert, 2007: An instrument for measuring size-resolved aerosol hygroscopicity at both sub- and super-micron sizes. Aerosol Sci. Technol., 41, 873883, https://doi.org/10.1080/02786820701506955.

    • Search Google Scholar
    • Export Citation
  • Held, A., and A. Mangold, 2021: Measurement of fundamental aerosol physical properties. Springer Handbook of Atmospheric Measurements, T. Foken, Ed., Springer, 535–565.

  • Irwin, M., Y. Kondo, N. Moteki, and T. Miyakawa, 2013: Evaluation of a heated-inlet for calibration of the SP2. Aerosol Sci. Technol., 47, 895905, https://doi.org/10.1080/02786826.2013.800187.

    • Search Google Scholar
    • Export Citation
  • Jayaratne, R., X. Liu, P. Thai, M. Dunbabin, and L. Morawska, 2018: The influence of humidity on the performance of a low-cost air particle mass sensor and the effect of atmospheric fog. Atmos. Meas. Tech., 11, 48834890, https://doi.org/10.5194/amt-11-4883-2018.

    • Search Google Scholar
    • Export Citation
  • Li, Z., and Coauthors, 2017: Aerosol and boundary-layer interactions and impact on air quality. Natl. Sci. Rev., 4, 810833, https://doi.org/10.1093/nsr/nwx117.

    • Search Google Scholar
    • Export Citation
  • Liu, B. Y., D. Y. Pui, K. L. Rubow, and W. W. Szymanski, 1985: Electrostatic effects in aerosol sampling and filtration. Ann. Occup. Hyg., 29, 251269, https://doi.org/10.1093/annhyg/29.2.251.

    • Search Google Scholar
    • Export Citation
  • Lohmann, U., and J. Feichter, 2005: Global indirect aerosol effects: A review. Atmos. Chem. Phys., 5, 715737, https://doi.org/10.5194/acp-5-715-2005.

    • Search Google Scholar
    • Export Citation
  • Magi, B. I., C. Cupini, J. Francis, M. Green, and C. Hauser, 2020: Evaluation of PM2.5 measured in an urban setting using a low-cost optical particle counter and a federal equivalent method beta attenuation monitor. Aerosol Sci. Technol., 54, 147159, https://doi.org/10.1080/02786826.2019.1619915.

    • Search Google Scholar
    • Export Citation
  • Mamali, D., and Coauthors, 2018: Vertical profiles of aerosol mass concentration derived by unmanned airborne in situ and remote sensing instruments during dust events. Atmos. Meas. Tech., 11, 28972910, https://doi.org/10.5194/amt-11-2897-2018.

    • Search Google Scholar
    • Export Citation
  • Masic, A., D. Bibic, B. Pikula, A. Blazevic, J. Huremovic, and S. Zero, 2020: Evaluation of optical particulate matter sensors under realistic conditions of strong and mild urban pollution. Atmos. Meas. Tech., 13, 64276443, https://doi.org/10.5194/amt-13-6427-2020.

    • Search Google Scholar
    • Export Citation
  • Muyshondt, A., A. R. McFarland, and N. K. Anand, 1996: Deposition of aerosol particles in contraction fittings. Aerosol Sci. Technol., 24, 205216, https://doi.org/10.1080/02786829608965364.

    • Search Google Scholar
    • Export Citation
  • Nurowska, K., M. Mohammadi, S. Malinowski, and K. Markowicz, 2023: Applicability of the low-cost OPC-N3 optical particle counter for microphysical measurements of fog. Atmos. Meas. Tech., 16, 24152430, https://doi.org/10.5194/amt-16-2415-2023.

    • Search Google Scholar
    • Export Citation
  • Platis, A., B. Altstädter, B. Wehner, N. Wildmann, A. Lampert, M. Hermann, W. Birmili, and J. Bange, 2016: An observational case study on the influence of atmospheric boundary-layer dynamics on new particle formation. Bound.-Layer Meteor., 158, 6792, https://doi.org/10.1007/s10546-015-0084-y.

    • Search Google Scholar
    • Export Citation
  • Rai, A. C., P. Kumar, F. Pilla, A. N. Skouloudis, S. Di Sabatino, C. Ratti, A. Yasar, and D. Rickerby, 2017: End-user perspective of low-cost sensors for outdoor air pollution monitoring. Sci. Total Environ., 607–608, 691705, https://doi.org/10.1016/j.scitotenv.2017.06.266.

    • Search Google Scholar
    • Export Citation
  • Rosati, B., G. Wehrle, M. Gysel, P. Zieger, U. Baltensperger, and E. Weingartner, 2015: The white-light humidified optical particle spectrometer (WHOPS)—A novel airborne system to characterize aerosol hygroscopicity. Atmos. Meas. Tech., 8, 921939, https://doi.org/10.5194/amt-8-921-2015.

    • Search Google Scholar
    • Export Citation
  • Samad, A., F. E. Melchor Mimiaga, B. Laquai, and U. Vogt, 2021: Investigating a low-cost dryer designed for low-cost PM sensors measuring ambient air quality. Sensors, 21, 804, https://doi.org/10.3390/s21030804.

    • Search Google Scholar
    • Export Citation
  • Samad, A., D. Alvarez Florez, I. Chourdakis, and U. Vogt, 2022: Concept of using an unmanned aerial vehicle (UAV) for 3D investigation of air quality in the atmosphere—Example of measurements near a roadside. Atmosphere, 13, 663, https://doi.org/10.3390/atmos13050663.

    • Search Google Scholar
    • Export Citation
  • Schrod, J., and Coauthors, 2017: Ice nucleating particles over the eastern Mediterranean measured by unmanned aircraft systems. Atmos. Chem. Phys., 17, 48174835, https://doi.org/10.5194/acp-17-4817-2017.

    • Search Google Scholar
    • Export Citation
  • Swietlicki, E., and Coauthors, 2008: Hygroscopic properties of submicrometer atmospheric aerosol particles measured with H-TDMA instruments in various environments—A review. Tellus, 60B, 432469, https://doi.org/10.1111/j.1600-0889.2008.00350.x.

    • Search Google Scholar
    • Export Citation
  • Tang, I. N., 1996: Chemical and size effects of hygroscopic aerosols on light scattering coefficients. J. Geophys. Res., 101, 19 24519 250, https://doi.org/10.1029/96JD03003.

    • Search Google Scholar
    • Export Citation
  • Tang, I. N., and H. R. Munkelwitz, 1994: Water activities, densities, and refractive indices of aqueous sulfates and sodium nitrate droplets of atmospheric importance. J. Geophys. Res., 99, 18 80118 808, https://doi.org/10.1029/94JD01345.

    • Search Google Scholar
    • Export Citation
  • Vogt, M., P. Schneider, N. Castell, and P. Hamer, 2021: Assessment of low-cost particulate matter sensor systems against optical and gravimetric methods in a field co-location in Norway. Atmosphere, 12, 961, https://doi.org/10.3390/atmos12080961.

    • Search Google Scholar
    • Export Citation
  • Weber, K., G. Heweling, C. Fischer, and M. Lange, 2017: The use of an octocopter UAV for the determination of air pollutants—A case study of the traffic induced pollution plume around a river bridge in Duesseldorf, Germany. Int. J. Educ., 2, 6366.

    • Search Google Scholar
    • Export Citation
  • Zhang, X., P. Massoli, P. K. Quinn, T. S. Bates, and C. D. Cappa, 2014: Hygroscopic growth of submicron and supermicron aerosols in the marine boundary layer. J. Geophys. Res. Atmos., 119, 83848399, https://doi.org/10.1002/2013JD021213.

    • Search Google Scholar
    • Export Citation
  • Fig. 1.

    Basic sketch of the diffusion drying chamber (arbitrary dimensions). The air goes through the chamber in the inner, perforated, UV resin printed tube, where blue silica gel extracts the water vapor from the flow. When exiting the chamber, the air is dry enough to be sampled at the OPC’s sensing area.

  • Fig. 2.

    The DJI S900, with the OPC measurement system installed on it. The dryer stands in the middle of the construction and aspiration is in the vertical downward direction, first through the dryer channel and then toward the sampling volume of the OPC-N3. Six SHT31 sensors are placed around the platform under each rotor arm. The OPC-N3 is located at the bottom of the supporting table, which is mounted on the main body of the UAS. Once the upper Styrofoam dome is in place, air sampling happens through the opening on top, which has a cover cup protection above it.

  • Fig. 3.

    Description of the drying chamber testing in a self-constructed fog tank. Fog is produced with an ultrasonic humidifier in an enclosed container, raising the relative humidity levels close to saturation. An SHT31 sensor measures the humidity in the fog tank, and the air is then drawn using a small pump, first through the diffusion dryer, into a second box with ambient air where a second SHT31 sensor is located, which measures the humidity after the humid air passes along the drying chamber.

  • Fig. 4.

    The S900 in flight (right) next to Mannheim-Nord station (left). The sensors, including the aerosol particle measurements from the Fidas 200, are located at the roof of the station. The distance between the S900 and the station was maintained between 7 and 12 m for safety reasons, and for the UAS downwash to leave the station’s measurements unaffected.

  • Fig. 5.

    Recorded SFR and the three PM values for a time of 43 min in laboratory conditions for the dryer test case. Each colored area at the top graph indicate measurement times where there was no tube attached to the OPC-N3’s inlet (area in gray), a tube but not a dryer was attached (area in brown), or the dryer was attached (area in cyan). Small gaps in between the different measurement stages indicate the time for adjusting or removing the dryer, and were excluded from the analysis.

  • Fig. 6.

    Recorded SFR and the three PM values for a time of 37 min in laboratory conditions for the bell-shaped inlet test case. Each colored area at the top graph indicate measurement times where there was no bell inlet attached to the OPC-N3’s extra tube (area in red), or it was attached (area in gray). Units for the three PM values are μg m−3.

  • Fig. 7.

    Two different examples of relative humidity measurements at the fog tank, using the constructed drying channels. (a) Fog is present in the fog tank creating high-humidity conditions, but the humidifier is then turned off. The dashed vertical black line indicates the moment when the pump is turned on, beginning the suctioning of wet air through the drying chamber. (b) The conditions are the same as in the first case, but this time there is no drying chamber but a simple copper tube connecting the two boxes.

  • Fig. 8.

    Measurements at Mannheim-Nord station during the morning (flights 1–2) and afternoon (flights 3–4) of 14 Feb 2023. Two consecutive flights with (i.e. “dry” case) a dryer on the OPC system and without (i.e. “ambient” case) it, for the cases of the three PM types. Measured RH from the SHT31 is shown at the bottom plot, and the solid black line shows the reference measurements from the Fidas 200 at the station. The solid S900 lines indicate 1 min resampled averages, and the less opaque lines of the same colors indicate the nonaveraged data.

  • Fig. 9.

    Calculated (left) aerosol particle number and (right) volume size distributions between 0.5 and 9 μm.

  • Fig. 10.

    Measurements inside a low-stratus cloud at Falkenberg tower. The first two flights took place in the morning when the cloud layer was quite low, starting at an altitude of 65 m, as recorded by the installed CHM-15k at the site. (top) Its level is indicated with the black horizontal line for flights 1 and 2. (bottom) In the afternoon (flights 3 and 4), the cloud was not anymore at the same altitude range. The three PM types are depicted under the two cases of using a dryer (i.e., dry) and not using one (i.e., ambient). Units for the three PM values are μg m−3.

  • Fig. 11.

    Humidity and temperature during the time of measurement flights of the UAS, taken from the sensor instrumentation at Falkenberg tower. (top) The two lines correspond to the two points during that day when the vertical profiles where performed, one in the morning and one in the afternoon. (bottom) The cloud layer height, observed by the ceilometer, is shown for a time period of 1 h. The time of flights of the UAS are also indicated in shaded blue, starting from the first vertical profile at 0750 UTC.