Impact of various air mass types on cloud condensation nuclei concentrations along coastal southeast Florida

Highlights

• African dust plumes do not perturb CCN concentrations in southeast Florida.

• CCN concentrations are greatly increased on days influenced by biomass burning.

• Heavy rainfall reduces CCN concentrations.

• The duration varies for CCN concentrations to restore after periods of heavy rain.

Abstract

Coastal southeast Florida experiences a wide range of aerosol conditions, including African dust, biomass burning (BB) aerosols, as well as sea salt and other locally-emitted aerosols. These aerosols are important sources of cloud condensation nuclei (CCN), which play an essential role in governing cloud radiative properties. As marine environments dominate the surface of Earth, CCN characteristics in coastal southeast Florida have broad implications for other regions with the added feature that this site is perturbed by both natural and anthropogenic emissions. This study investigates the influence of different air mass types on CCN concentrations at 0.2% (CCN_0.2%) and 1.0% (CCN_1.0%) supersaturation (SS) based on ground site measurements during selected months in 2013, 2017, and 2018. Average CCN_0.2% and CCN_1.0% concentrations were 373 ± 200 cm⁻³ and 584 ± 323 cm⁻³, respectively, for four selected days with minimal presence of African dust and BB (i.e., background days). CCN concentrations were not elevated on the four days with highest influence of African dust (289 ± 104 cm⁻³ [0.2% SS] and 591 ± 302 cm⁻³ [1.0% SS]), consistent with high dust mass concentrations comprised of coarse particles that are few in number. In contrast, CCN concentrations were substantially enhanced on the five days with the greatest impact from BB (1408 ± 976 cm⁻³ [0.2% SS] and 3337 ± 1252 cm⁻³ [1.0% SS]). Ratios of CCN_0.2%:CCN_1.0% were used to compare the hygroscopicity of the aerosols associated with African dust, BB, and background days. Average ratios were similar for days impacted by African dust and BB (0.54 ± 0.17 and 0.55 ± 0.17, respectively). A 29% higher average ratio was observed on background days (0.71 ± 0.14), owing in part to a strong presence of sea salt and reduced presence of more hydrophobic species such as those of a carbonaceous or mineral-dust nature. Finally, periods of heavy rainfall were shown to effectively decrease both CCN_0.2% and CCN_1.0% concentrations. However, the rate varied at which such concentrations increased after the rain. This work contributes knowledge on the nucleating ability of African dust and BB in a marine environment after varying periods of atmospheric transport (days to weeks). The results can be used to understand the hygroscopicity of these air mass types, predict how they may influence cloud properties, and provide a valuable model constraint when predicting CCN concentrations in comparable situations.

Introduction

The largest uncertainty in the anthropogenic contribution to climate change is the radiative forcing due to aerosol-cloud interactions (IPCC, 2013). Accurately quantifying concentrations of cloud condensation nuclei (CCN) is an integral part in reducing this uncertainty due to their role in determining cloud radiative properties and longevity (Albrecht, 1989; Rosenfeld et al., 2008; Twomey, 1974). Closure studies, in which in situ CCN concentrations are compared to those calculated based on measured input data (e.g., composition, size distribution), have been successful under background aerosol conditions (Chuang et al., 2000; Dusek et al., 2003; Gasparini et al., 2006; Rissler et al., 2004; Stroud et al., 2007; Vanreken, 2003). Urban environments present a greater challenge due to the increased complexity in the size and composition of the aerosols present (Crosbie et al., 2015; Cubison et al., 2008; Sotiropoulou et al., 2007). Sotiropoulou et al. (2007) reports that CCN model predictions in places impacted by urban pollution and biomass burning are most uncertain. With over half of the world's population living in urban environments, a number that is only growing (United Nations, 2018), there is increasing urgency to reduce error in model-predicted CCN concentrations in such settings to better forecast cloud properties, climate change, weather, and surface air quality (Andreae and Rosenfeld, 2008; Cubison et al., 2008; Molina et al., 2007).

Populated coastal areas create an environment in which anthropogenic and natural aerosols converge, but the resulting air masses and their properties remain unclear (Cruz et al., 2019; Cubison et al., 2008; Sotiropoulou et al., 2007). Kummu et al. (2016) reports that 39% of the global population in 2010 lived within 100 km of a coastline and that this number will continue to increase until at least 2050. With sea salt and dust comprising the majority of global aerosols by mass, the merging of these natural aerosol types and human-generated particles is inevitable (Huneeus et al., 2011; Middleton and Goudie, 2001).

Virginia Key, an island just east of Miami, Florida, provides an excellent location to study CCN since it hosts a complex mixture of aerosols, originating locally, regionally, and from distant continents. The resulting mixture of aerosols is ideal for challenging and improving models (Klejnowski et al., 2013). Virginia Key experiences periodic influence from Miami's urban emissions, such as those from shipping activity (Hu et al., 2021) and fossil fuel combustion (Klejnowski et al., 2013). Additionally, the island is subject to the constant presence of marine aerosol types (e.g., sea salt, organic matter, and those derived from dimethyl sulfide [DMS]; Quinn and Bates, 2014) characteristic of coastal areas (Gantt et al., 2015; Nolte et al., 2015). Of particular relevance to this location are episodic intrusions of transported African dust, which have been the subject of numerous past works (Aldhaif et al., 2020 and references therein; Prospero, 1999; Zuidema et al., 2019). African dust reaches southeast Florida year-round but peaks between June and August (Aldhaif et al., 2020; Prospero, 1999; Zuidema et al., 2019), typically including 5–8 episodes each summer (Kramer et al., 2020a). Studies have quantified the arriving mass concentration of African dust (Zuidema et al., 2019), investigated its influence on human health (Prospero et al., 2014) and air quality (Prospero, 1999), and characterized its size distribution (Kramer et al., 2020b; Li-Jones and Prospero, 1998) and chemical properties (Zamora et al., 2011, 2013). However, the influence of African dust on CCN concentrations in southern Florida has received scarce attention, and it is unknown if the dust perturbs CCN concentrations in the area. Many models (e.g., ECHAM5-HAM [Hoose et al., 2008], GLOMAP-bin [Manktelow et al., 2010], EMAC [Pringle et al., 2010], and Meso-NH [Bègue at al., 2015]) consider dust particles as efficient CCN, particularly if they acquire soluble material via internal mixing with other hydrophilic species (Karydis et al., 2011). To our knowledge, the hygroscopic properties of African dust reaching the southeastern U.S. have not been reported in the literature, which adds a degree of uncertainty to how CCN models currently parameterize dust plumes arriving to this region. Weinzierl et al. (2017) shows that CCN concentrations are elevated above background levels during summertime episodes of Saharan dust observed at Barbados, an island 2570 km southeast of Virginia Key. However, both the amount of African dust reaching Virginia Key is consistently less than that arriving in Barbados (Zuidema et al., 2019) and Virginia Key is periodically affected by a more complex, urban mixture of aerosols. Additionally, Denjean et al. (2015) and Kandler et al. (2018) each find the hygroscopicity of supermicrometer African dust particles to not change significantly as they are transported across the Atlantic Ocean to the Caribbean, further motivating interest in how African dust plumes affect the CCN budget of the southeastern U.S.

An additional complexity for the southern Florida coastal area is that it is a receptor site of BB emissions from different source regions. Jaffe et al. (2020) finds Florida to be the fourth highest state in the U.S. for annual area burned in prescribed fires, which occur most frequently in the boreal spring (March–May). One source responsible for these prescribed fires is the largest area for sugarcane production in the U.S., located ~130 km to the northwest of Virginia Key. Between mid-fall and spring (October–May), 50–80% of the plants’ foliage is burned before harvesting (Le Blond et al., 2017; Ma et al., 2014) and the resulting BB emissions have been shown to travel to eastern Florida where Virginia Key is located (Sevimoğlu and Rogge, 2015, 2016, 2019). Additionally, BB emissions from deforestation and crop residue fires in Central America (Wang and Christopher, 2006; Wang et al., 2006), Mexico (Kreidenweis et al., 2001), the Yucatan peninsula (Yokelson et al., 2009), Cuba (Brey et al., 2018), and the southeastern U.S. (Roy et al., 2018) have also been shown to reach Florida. Many studies have looked at the nucleating ability of BB aerosol particles (e.g., Bougiatioti et al., 2016; Hennigan et al., 2012; Lee et al., 2010; Pöhlker et al., 2018) and found CCN concentrations generally increase amid BB plumes. However, the exact effect of BB on CCN concentrations is dependent on the fuel source (Chen et al., 2019), background aerosol properties (Pöhlker et al., 2018), and aging of the emissions (Bougiatioti et al., 2016; Hennigan et al., 2012). Each of these conditions differs between Virginia Key and the locations investigated in previous studies, so it is not yet known how the different sources of fire mentioned above influence CCN concentrations in this region. Liu et al. (2013) predicts that the southeastern U.S. will experience a longer fire season in the future, and thus, understanding the effects of BB on CCN in this area is critical.

Another benefit of investigating CCN concentrations in southeast Florida is that aside from the complexity of aerosol sources, there is a wide range of meteorological conditions that both impact aerosol characteristics and are, in turn, potentially influenced by CCN themselves. In particular, Virginia Key experiences greater rainfall than most of the U.S. (Smith et al., 2019), and this precipitation peaks just before the influence of African dust is highest (Abiy et al., 2019; Zuidema et al., 2019). Many particles, including CCN, can be removed during rain episodes (Ohata et al., 2016). In order to improve climate models, not only does the nucleating ability of aerosols need to be understood, but the removal of CCN through wet scavenging needs to be as well.

The effect of episodic Saharan dust and BB emissions superimposed on background aerosol conditions, as well as the implications of wet scavenging, on CCN concentrations in this region is unknown but essential to study in order to improve models simulating aerosol particles, clouds, and precipitation. In response to this knowledge gap, this study aims to characterize the behavior of CCN concentrations in southeast Florida and to identify how they are impacted by the influence of different aerosol types and rain conditions. This work specifically examines in situ data for CCN concentrations measured at different supersaturations (SSs) on Virginia Key and uses complementary datasets to address the relative importance of dust, smoke, and rainfall on CCN properties.