Ground level ice nuclei particle measurements including Saharan dust events at a Po Valley rural site (San Pietro Capofiume, Italy)

Highlights

• Aerosol in the coarse range contributed significantly to the INP concentration.

• Particle number concentration was higher during SDE.

• Precipitation events influenced the aerosol concentration.

• There was mainly a low correlation between INP and total aerosol concentration.

Abstract

Filter-collected aerosol samples in the PM₁ and PM₁₀ fractions and particle number concentration were measured during experimental campaigns in a rural area near Bologna (Italy) in the periods 10–21 February 2014 and 19–30 May 2014. Ice nuclei particle (INP) concentrations measured off-line showed prevalently higher average values in the morning with respect to the afternoon, in the PM₁ fraction with respect to PM_1–10 (with the exception of the first campaign, at S_w = 1.01), and at water saturation ratio S_w = 1.01 with respect to S_w = 0.96.

The aerosol in the coarse size range (1–10 μm) contributed significantly to the total INP concentration. In the first campaign, the average INP concentration in the coarse fraction was 80% of the total in the morning and 74% in the afternoon, at S_w = 1.01. In the second campaign, the contribution of the coarse size fraction to the INP number concentration was lower. On the whole, the results showed that the freezing activity of aerosol diameters larger than 1 μm needs to be measured to obtain the entire INP population.

Sahara dust events (SDEs) took place during both campaigns, in the periods 17–20 February and 21–23 May 2014. Results show that the averaged particle number concentration was higher during SDE than during no-Saharan dust events. A low correlation between INP and total aerosol number concentration was generally measured, except for SDEs observed in February, in which the correlation coefficient between aerosol concentration in the coarse fraction and INP in the same range, at water supersaturation, was about 0.8.

Precipitation events influenced the aerosol concentration. In the February campaign, lower values of INP and particle concentrations were measured in case of heavy rain events. During the May campaign, an average number concentration of the aerosol in the range 0.5–10 μm was slightly higher than on days when no precipitation was measured, the rainfall intensity being low. Only in a few cases did we note a sharp drop in INP in the PM₁₀ fraction at S_w = 1.01 (26 May, 8 a.m. and 1 p.m.; 27 May, 1 p.m.).

Introduction

Insoluble aerosol particles that catalyze the formation of ice crystals in clouds are called ice nuclei particles (INP) (Vali et al., 2015) and can form ice through four different thermodynamic mechanisms: deposition, condensation-freezing, immersion-freezing, and contact-freezing.

Ice nuclei can be measured with a variety of techniques: mixing chambers (Langer, 1973, Bundke et al., 2008), expansion cloud chambers (Möhler et al., 2003, Tajiri et al., 2013), continuous flow diffusion chambers (Al-Naimi and Saunders, 1985, Rogers et al., 2001), and the filter method (Bigg, 1996, Klein et al., 2010, Langer and Rodgers, 1975). Each device has advantages and disadvantages. For instance, the continuous flow diffusion chamber (CFCD) cannot detect contact-freezing, and aerosol particles larger than about 2 μm need to be removed with an impactor to distinguish ice crystals reliably from background aerosols. In addition, the CFCD provides no information on the size distribution of INP. Besides cloud chambers and CFDC devices, many droplet-freezing techniques are available for immersion mode ice nucleation (Ardon-Dryer et al., 2011, Budke and Koop, 2015, Knopf and Alpert, 2013, Murray et al., 2011, Vali, 2008).

Membrane filter techniques have been used in different forms for a number of years. Particle sampling on filters can be convenient, because the samples can be processed later without nuclei degradation, and aerosol activation can be measured in all size ranges, and can simulate deposition, condensation and contact nucleation modes (Bigg, 1990a, Cooper, 1980, Stevenson, 1968). The temporal resolution is usually no better than 20–30 min. Early designs used a static vapour diffusion field, but several factors led to an underestimation of the ice nuclei (volume effect, chamber height effect, vapour depletion on the filter around growing crystals and hygroscopic particles, vapour competition between ice nuclei). However, Bigg (1990a) concluded that with suitable precautions the filter method is adequate for INP measurements. In order to circumvent some of the problems arising with the static chamber, a dynamic chamber filter processing chamber (DFPC) was introduced (Langer and Rodgers, 1975).

CFDC measurements sometimes yielded INP higher concentrations than the filter method (Al-Naimi and Saunders, 1985, Hussain and Saunders, 1984). In simultaneous comparison, Hussain and Kayani (1988) found that at water saturation their CFDC detected fourteen times as many nuclei than filters processed in a static chamber. However, in a similar comparison Saunders and Al-Juboory (1988) found nearly equivalent results between a CFDC and DFPC filters at − 16 °C and supersaturation with respect to water between − 3% and + 5%.

Plaude et al. (1996) performed simultaneous measurements on INP concentration using a cloud chamber and a filter technique on the territory of the former USSR over a five-year period. The agreement between the results obtained by the two techniques was found to depend on the overall pollution of the region, and is higher in a relatively clean atmosphere (INP_chamber/INP_filter ~ 1.9), and lower in an urban area which is more polluted (INP_chamber/INP_filter in the range 4 to 7).

Most INP measurements concern total INP number concentrations with less emphasis on determining their size distribution. In point of fact, information on airborne INP size distribution may be helpful to identify the predominant INP sources. For instance, primary biological aerosol particles can span physical dimensions of a few nanometers to hundred of micrometers (Huffman et al., 2012), whereas black carbon particles are mainly in the submicron range (Clarke et al., 2004, Schwarz et al., 2008). Therefore, size-resolved measurements of ice nucleating particles should be performed and even super-micron particles should be considered. Previous field size-resolved INP measurements are scarce (Berezinski et al., 1988, Huffman et al., 2013, Rosinski et al., 1986, Rosinski et al., 1988, Santachiara et al., 2010).

Mason et al. (2016) recently attempted to tackle this problem using a technique which combines aerosol particle collection by a cascade inertial impactor (Moudi MSP Corp., USA) and a microscope-based immersion-freezing apparatus. The results showed that coarse aerosol particles are a significant component of the ice nuclei population in many different ground-level environments (see Table 2a, Table 2b of the paper).

Our paper reports the results of two experimental campaigns performed with the filter method at a rural background site. This procedure activates all size ranges of aerosol in the deposition and condensation-freezing modes for the examination of fine and coarse particles. Even if the sampling site is a rural area, it is surrounded by highly populated industrialized regions.

Our measurements were performed at ground level. To date, few INP measurements at ground level in low polluted areas have been reported in the literature. Ground-based INP measurements are important as INP can originate from the surface and be transported to higher levels (Ardon-Dryer et al., 2011, Alizadeh-Choobari et al., 2015, Després et al., 2012, Mamouri and Ansmann, 2015). The aim of the campaigns was to characterize the considered site with respect to certain INP properties, i.e. concentrations, and their relationship with particle number concentration, diurnal variations and possible sources.