The provision of reliable observations of the chemical composition and physical properties of the atmosphere is a pillar for understanding atmospheric chemistry and climate change both in term of process investigation and detection of regional and global changes. To assure the reliability of observed data, it is crucial to standardize and increase compatibility of measurements and data: to these aims global (e.g., Global Atmosphere Watch by the World Meteorological Organization—GAW/WMO) and regional (e.g., Aerosol, Clouds, and Trace gases Infrastructure—ACTRIS) efforts were undertaken in the last decades (see [1
]). The standardization of measurement techniques is an urgent issue especially for atmospheric aerosols (hereinafter AEs) and reactive gases (hereinafter RGs).
AEs influence the energy budget of the atmosphere through direct and indirect radiative effects. Direct effects include the scattering and absorption of radiation in sensible wavelengths, while indirect effects involve the influence of AEs on cloud condensation nuclei (CCN) which in turn affects cloud albedo, lifetime and precipitation frequency. Absorbing AEs (both natural and anthropogenic) also affect climate by deposition on snow and ice surface, thus modifying the albedo of the cryosphere at regional and global scales. Stocker et al. [3
] estimated the AEs radiative forcing over 1750–2011 as −0.9 W/m2
(90% confidence interval: from −1.9 to −0.1 W/m2
). This radiative forcing encompasses a dominant cooling effect and a warming contribution mostly from black carbon (a well-known short-lived climate forcer, see [4
In addition to climate influence, AEs affect many aspects of human health and the environment. Toxic chemical components within AEs are known to have links to health problems like chronic respiratory and acute cardio-vascular diseases. AEs were also closely linked to visibility reduction, acid rain, and urban smog in many locations of the world. Dimming effect of solar radiation by AEs can induce a decrease of agriculture yields.
As a function of their diameters, two populations of particles, with different sources, sinks, sizes and chemical compositions, generally constitute the tropospheric AEs. The sub-micrometer fraction (“fine particles”) represents the sum of the nucleation (up to about 20 nm diameter), Aitken (20–100 nm) and accumulation modes (0.1–1.0 μm). The coarse fraction, encompasses AE particles with diameter greater than 1.0 μm. The fine fraction of AEs originates from condensation process and from atmospheric gas-to-particle conversion. It is removed by precipitation and dry deposition. On the other hand, the coarse particles are produced by mechanical processes (e.g., soil erosion by wind, dust mobilization by wind and seawater bubble-bursting) and is mainly removed by sedimentation and interaction with the Earth’s surface.
Accurate size-segregated or size-resolved measurements are needed to determine the AEs properties within each of these populations and to investigate their hence effects on climate and air quality. As an instance, the AE particles within the accumulation mode are the most important for direct radiative forcing while nucleation particle observations are useful for investigating particle formation processes. When present in large amount, also coarse AEs strongly affect the scattering of solar radiation. On the other side, the AE particles in the “fine” mode can potentially penetrate deep into the lung, thus playing important roles for human health and diseases.
RGs are a group of atmospheric molecules with short lifetimes (ranging from hours to a few months). Among them, important roles are played by tropospheric ozone (O3
), carbon monoxide (CO), reactive nitrogen gases (NOx
), volatile organic compounds (VOC) and reactive sulfur gases (SOx
). In particular, the RGs regulate the oxidative capacity of the atmosphere as they represent the key source of reactive atomic and free radical species (e.g., O*, OH, HO2
). In this way, they have indirect impact on the Earth’s atmosphere radiative forcing by affecting lifetime of greenhouse gas. O3
is a powerful greenhouse gas [4
], while NOx
, VOC and SO2
affect the formation of aerosol particles and clouds [5
]. RGs have direct effects on human health and ecosystem integrity; for these reasons they are the object of regulation emission directives in inhabited regions [6
]. Reliable long-term observations of RGs are also useful to assess the effectiveness of the adopted air-quality strategies. Indeed, as reported by [5
], information of RG spatial distribution and variability is essential for assessing the effectiveness of emission mitigation measures.
The accurate measurement of AEs and RGs represents a challenging task. To correctly determine AE properties it is paramount to minimize artifacts related with the air sampling. At background measurement sites RG mole fractions range from nmol/mol (i.e., ppb) to pmol/mol (i.e., ppt). Furthermore, RGs are often difficult to sample just because of their reactivity. Thus, a pre-requisite for reliable execution of atmospheric AE and RG observations is the adoption of suitable sampling systems able not only to guarantee the correct execution of air sampling (as an instance, at wet or windy environments) but also to minimize artifacts. In particular, the presence of wind gusts or high wind speed can make the air sampling not efficient, while the presence of water vapor condensation inside the inlet can produce interference in the AEs and RG measurements (i.e., hygroscopic growth of aerosol particle and interaction of gas molecules with water droplets).
GAW/WMO and ACTRIS provide standard operation procedures (SOP) and guidelines about the correct design and implementation of sampling systems for AEs and RGs. By the report [7
], GAW/WMO recommended that inlets for AEs provide undisturbed and representative aerosol sampling to measurement instrumentation. To this aim, sampled air should be brought into the station laboratory through a vertical stack with an inlet characterized by a sampling efficiency not depending with wind direction or wind speed. Along the stack the sample flow should be laminar to avoid losses of small particles by diffusion and turbulent inertial deposition. A Reynolds number (Re) of about 2000 is indicated as ideal by [7
]. Concerning materials, conductive and non-corrosive material must be used (e.g., stainless-steel) in order to not change the size distribution or chemical composition of the aerosol particles in the sampled air. The implementation of humidity control is suggested due to the strong influence of relative humidity (RH) on the size of atmospheric particles: GAW/WMO recommendations are to maintain the RH lower than 40%. For humid “like-tropical” locations as CGR, no agreed standard technique to dry the aerosols exists. In our case, we decided to implement a passive system based on the modest “sensible heating” of sunlight on the external coating of the air-intake without any direct temperature regulation. For AEs, it is also requested to exclude precipitation because of its interference with the measurements: to this aim, the air inlet should be equipped with a cover to avoid the sampling of drizzle and rain.
Concerning RGs, clear indications about air-inlet design have been provided by GAW/WMO (see [8
]) and ACTRIS (see [9
]) for the execution of NOx
measurements. In particular, the inner surface of the inlet line must be smooth, non-porous and inert: it is recommended to use PFA Teflon©
or similar material. To avoid vapors condensation, it is recommended to heat the inlet line. The temperature has to be chosen high enough that no condensation occurs but not too high that thermal decomposition of other substances (e.g., peroxide-acetyl-nitrates) become an artefact: controlled heating a few degrees (3–4°) above ambient temperature is suggested. Gas phase processes may also lead to changes in trace gas mixing ratios, because of different conditions in the inlet line compared to ambient. As an instance for NOx
observations an issue can be represented by the interaction with O3
in the sampling lines (NO titration). Thus, the residence time in the inlet line must be kept as short as possible (recommended is a residence time of less than 5 s, better below 2 s).
In this paper, we will describe two novel sampling systems for the near-surface observations of AEs and RGs that were developed in the framework of the PON/FESr I-AMICA Project (www.i-amica.it
). These systems make use of not-standard and low power consumption technical solutions for the management of wind variability (i.e., gusts and high wind speed) together with smart devices for the regulation of sampling flow rates and air temperature, as well as for the recording of diagnostic data. Firstly (Section 2
), we provide a technical description of the sampling systems with a particular emphasis on vertical stacks, used material and the adopted solution for the flow and temperature controls. Then (Section 3
), we provide evidences of the system performances by analyzing 1 full year of operation at the regional WMO/GAW station at Capo Granitola (Italy), located along the south-west coastline of Sicily. Finally, in Section 4
we summarize and discuss the results.
4. Summary and Discussion
In the framework of the I-AMICA Project (www.i-amica.eu
), we developed novel “smart” and relatively low-cost systems for the continuous sampling of ambient air devoted to the investigation of AE and RG variability, according with guidelines of leading internal programme for the investigation of air-composition variability (i.e., WMO/GAW and ACTRIS). These systems were designed to feed multiple instrumentation for the measurements of RG mixing ratios and AE physical properties (number size distribution, absorption and scattering coefficient, total particle number concentration). A particular attention was dedicated to the implementation of systems for the generation and control of the sampling flow rates and temperatures. In respect to “traditional” rotative or diaphragm pumps and turbo blowers, “low-cost” fans (typically used for computer cooling) are used to generate the sampling flow. These sampling fans can be easily accessed to perform preventive check or substitution in the case of damages. During the first period of system operation at CGR, a few sampling interruptions were experienced due to the damage of the wire welding of the fan motors. This problem was solved by apply an appropriate resin coverage to the electric contacts which prevent any oxidation due to the exposure of high level of sea-salt.
Moreover, in order to minimize “over-pressure” or “under-pressure” conditions, an innovative pneumatic design was applied to the air-intake (see Section 2.1
). As testified by one full year of operation at the WMO/GAW “CGR” coastal station, the proposed systems were effective in maintaining sampling flow rates and temperature to defined set-point values independently by external wind speed and external ambient temperature.
“Smart” and low-cost miniaturized computers (i.e., Raspberry PI) are used as control boards to monitor and record the system internal parameters (i.e., flow rates, sample temperature and relative humidity, power consumption, fan motor r.p.m.) as well as external conditions (wind speed, ambient temperature) which make easy the near-real time upload of recorded data as well as the remote control of the system (e.g., regulation of fan rotation velocity). Together with the very low electrical consumption (under 10 W), this makes the proposed systems very suitable for remote sites where the electrical and man powers are limited. Moreover, the following benefits can be obtained by using these smart technologies:
Low implementation costs
Easily obtain spare-parts or prepare “back-up” systems
Possibility to interact (as in the case of Raspberry© computers) with a wide user community for problem solving
Feasible maintenance even by not-experts
Possibility of remote diagnosis and setting-up
Further implementations are on-going to adapt these systems to polar or high-mountain environments. The density and viscosity variation with pressure, temperature and absolute humidity is of the order of 10%, and therefore negligible in Re calculations from the point of view of flow control needs. However, we will consider the possibility to include pressure and temperature readings in the calculation of the flow rate needed to maintain “laminar” conditions inside the AE vertical stack.