Abstract. To quantify the contribution of new particle formation (NPF) to ultrafine
particle number and cloud condensation nuclei (CCN) budgets, one has to understand the mechanisms that
govern NPF in different environments and its temporal extent. Here, we study
NPF in Cyprus, an Eastern Mediterranean country located at the crossroads of
three continents and affected by diverse air masses originating from
continental, maritime, and desert-dust source areas. We performed 1-year
continuous measurements of aerosol particles down to ∼ 1 nm in
diameter for the first time in the Eastern Mediterranean and Middle East
(EMME) region. These measurements were complemented with trace gas data,
meteorological variables, and retroplume analysis. We show that NPF is a
very frequent phenomenon at this site and has higher frequencies of
occurrence during spring and autumn. NPF events were both of local and
regional origin, and the local events occurred frequently during the month
with the lowest NPF frequency. Some NPF events exhibited multiple onsets,
while others exhibited apparent particle shrinkage in size. Additionally,
NPF events were observed during the nighttime and during episodes of high
desert-dust loadings. Particle formation rates and growth rates were
comparable to those in urban environments, although our site is a rural one.
Meteorological variables and trace gases played a role in explaining the
intra-monthly variability of NPF events, but they did not explain why summer
months had the least NPF frequency. Similarly, pre-existing aerosol loading
did not explain the observed seasonality. The months with the least NPF
frequency were associated with higher H2SO4 concentrations but
lower NO2 concentrations, which is an indicator of anthropogenic
influence. Air masses arriving from the Middle East were not observed during
these months, which could suggest that precursor vapors important for
nucleation and growth are transported to our site from the Middle East.
Further comprehensive measurements of precursor vapors are required to prove
this hypothesis.