Plain Language Summary
Clouds play a critical role in Earth’s climate, both reflecting incoming
sunlight and trapping outgoing heat energy. Hence, even small errors in
the representation of clouds in climate models can lead to uncertainty
in predictions of, for example, sea ice extent. In the Arctic, clouds
often exist below 0°C and cloud water droplets can exist in a
supercooled liquid state. In the absence of a special class of particle
that can trigger ice formation in droplets, ice-nucleating particles
(INPs), supercooled water droplets can cool well below -35°C before
spontaneously freezing. Hence, the presence of INPs can reduce the
amount of supercooled water in clouds, making them less reflective with
a shorter lifetime. Based on our knowledge of INPs in other remote
oceans, we expected very low INP concentrations in the central Arctic.
However, we have shown that there are high concentrations of biological
INPs in the summertime North Pole. Furthermore, these INPs come from the
seas off the coast of Russia, a region already experiencing strong
climate change. It is possible that these sources may become even more
important as the Arctic becomes increasingly ice-free, causing changes
in Arctic clouds and further changes in climate.
1 Introduction
The Arctic climate is strongly influenced by ubiquitous low-level
mixed-phase clouds [Kay and L’Ecuyer, 2013; Tjernstrom et al. ,
2012; Vüllers et al. , 2021]. The radiative effect of these
clouds is influenced by the amount of ice and supercooled water they
contain, which depends on an intricate balance of dynamical and
microphysical processes [Morrison et al. , 2012]. Realistic
representation of these processes is needed to correct model biases in
the amount of supercooled liquid in mixed-phase clouds and reduce
uncertainty in feedbacks [Tan and Storelvmo, 2019].
A rare subset of the total aerosol particle population, ice-nucleating
particles (INPs), can induce primary ice production in Arctic
mixed-phase clouds when immersed in supercooled cloud droplets [Murrayet al. , 2012]. In the summertime, Arctic marine atmospheric
boundary layer temperatures are usually much warmer than those required
for homogeneous freezing (≲ –35 °C) [Herbert et al. , 2015],
hence heterogeneous nucleation on INPs determines the production of ice
in clouds, at least in the absence of ice precipitating from overlying
clouds [Vassel et al. , 2019]. Numerous INP types that can
induce nucleation over a large range of temperatures have been
identified [Hoose and Möhler, 2012; Kanji et al. , 2017; Murrayet al. , 2012]. However, the sources and ice-nucleating
properties of INPs in the Arctic, especially the central Arctic
(>80° N), are poorly defined.
INP measurements have been made around the periphery of the Arctic
circle from locations close to, or on, land, but relatively few
measurements have been made in the summertime central Arctic Ocean (see
compilations in [Welti et al. , 2020] and [Murray et
al. , 2021]. Recent research suggests that there are significant
terrestrial sources of Arctic INPs including glacial dust from Svalbard
[Tobo et al. , 2019] and Iceland [Sanchez-Marroquin et
al. , 2020], terrestrial biological aerosol from boreal forests
[Schneider et al. , 2021], and even particles released from
thawing permafrost [Creamean et al. , 2020]. There is also a
plethora of other high latitude dust sources that have not been
investigated in terms of their ice-nucleating ability [Bullardet al. , 2016]. Marine biogenic INPs emitted from the sea
surface through bubble bursting are also thought to contribute to the
INP population of the oceanic high-latitudes [Bigg, 1996; Bigg and
Leck, 2001; Hartmann et al. , 2020a; Hartmann et al. , 2021;
Ickes et al. , 2020; Irish et al. , 2017; Wilson et
al. , 2015]. Sea spray is thought to produce relatively low INP
concentrations, but in the absence of other INP types it can dominate
the INP population [McCluskey et al. , 2018a; Vergara-Tempradoet al. , 2017].
Ground level observations at several land-based sites around the Arctic
throughout the seasonal cycle showed the highest (but variable) INP
concentrations during spring, summer and autumn and the lowest
concentrations in winter [Wex et al. , 2019]. These
measurements suggest that there are marine and terrestrial INP sources
around the Arctic, but it is unclear how important these sources are for
clouds over the summertime central Arctic Ocean. Based on back
trajectory analysis of INP measurements in the central Arctic, Bigg
[1996] suggested that there was an open ocean source of INPs active
at –15 °C. Later, Bigg and Leck [2001] suggested the pack ice edge
and bubble bursting in local leads throughout the pack ice can serve as
a source of INPs. Indeed, it has been shown that there is a reservoir of
INPs in the seas around the Arctic [Creamean et al. , 2019;
Hartmann et al. , 2021; Irish et al. , 2017; Wilson et
al. , 2015] and INP concentrations in the central Arctic decrease
during the transition from Arctic summer to autumn, possibly due to the
reduced availability of ice-free marine sources [Bigg and Leck,
2001].
While it is clear that there are strong sources of INPs in the lower
Arctic environment (≲80°N), it is not clear if these INPs are
transported to the central Arctic. The prevailing view is that aerosol
within the summertime high Arctic boundary layer experiences little
effect from long-range transport [Kupiszewski et al. , 2013],
and with few sources of primary aerosol in the central Arctic Ocean,
sources such as local leads may be important [Bigg and Leck, 2001].
However, it has also been suggested that aerosol particles can be
transported from lower latitudes into the central Arctic boundary layer
either through boundary layer transport or entrainment from the free
troposphere [Igel et al. , 2017; Morrison et al. , 2012;
Schmale et al. , 2021].
The structure of the Arctic summertime boundary layer is complex (Figure
1). The boundary layer is typically several hundred meters to over a
kilometer deep, but often consists of two distinct layers: the surface
mixed layer and the cloud mixed layer. These two layers are each well
mixed, but separated by a decoupling layer at ~100 m to
300 m that prevents efficient transport between them [Brooks et
al. , 2017]. Hence, measurements at the surface are not necessarily
representative of those in the cloud mixed layer.
Here, we present measurements of INP concentrations close to the North
Pole both in and above the surface mixed layer. The measurements were
made during the Microbiology-Ocean-Cloud-Coupling in the High Arctic
(MOCCHA) campaign, which took place throughout August and September 2018
on the Swedish icebreaker Oden. Measurements took place while Oden was
on route to the North Pole as well as when it was moored to an ice floe
in the inner pack ice and drifting passively between 88-90°N. Samples
were collected for INP analysis at both ship level (in the surface mixed
layer) and using a balloon-borne sampler in the cloud mixed layer. We
use backward trajectories alongside other measurements to suggest that
the source of the most active INPs reaching the North Pole is outside of
the pack ice, and near the Arctic coast of Russia.